{"id": "9639f03b-1518-46f5-b184-2f4ccda88482", "input": {"query": "What is the overall extruder efficiency in polymer processing?", "source_url": "https://ajmcmillan.co.uk/AcademicPublications/EnergyInExtrusion_AcceptedVersion.pdf", "document_text": "Energy efficiency in extrusion-related polymer processing: a review of state of\nthe art and potential efficiency improvements\nChamil Abeykoon\u00aa,*, Alison McMillan, Bao Kha Nguyen\n\"North West Composites Centre and Aerospace Research Institute, Department of Materials, Faculty of Science and Engineering, University of\nManchester, Oxford Road, Manchester, M13 9PL, UK\nb Faculty of Arts, Science and Technology, Wrexham Glyndwr University, Wrexham, LL11 2AW, UK\n\"School of Engineering and Informatics, University of Sussex, Brighton, BN1 9QT, UK\nAbstract\nEnergy saving and industrial pollution have become increasingly important issues, therefore the identification and\nadoption of more energy efficient machines and industrial processes are now industrial priorities, and worthy topics\nfor further development through academic research. Polymeric materials are a major raw material, finding widespread\napplication to a range of current industrial machine components as well as multiple products and packaging found\nin our daily life. Polymer extrusion serves as a particular example of polymer processing techniques, representative\nof others in as much as there are analogous intermediate stages in the processing. Processing techniques which re-\nquire such intermediate stages include the manufacture of blown film, blow moulding, thermo-forming, and injection\nmoulding. Hence, the study of polymer extrusion is a representative paradigm for a wider range of processing tech-\nniques. Since polymer processing is an energy intensive process and accounts for a huge share (maybe more than 1/3)\nof the materials processing sector, any improvement to the process would contribute significantly to global energy\nsavings. This work presents a review of studies, which focus on, or appertain to, the energy consumption of extrusion\nrelated polymer processing applications. Typical energy demand and losses during processing are considered, and\npossible approaches for improving the process energy efficiency while maintaining the required end product quality\nare considered. Overall, this work provides a detailed discussion about how and where energy is utilized; how, where\nand why energy losses occur; and sets out approaches for optimizing the process energy efficiency.\nKeywords:\nEnergy consumption, Energy losses, Energy savings, Polymer extrusion, Process monitoring, Process control,\nMaterials processing, Energy efficiency, Industry 4.0, Circular economy, Dynamical systems\n1. Introduction\n1.1. Market demand for polymers\nAs the number of applications for polymer materials in high volume manufacturing sectors, such as packaging, con-\ntinues to grow, it is timely to consider manufacturing process optimisation from the energy efficiency point of view.\nAt the present time, the increasing adoption of thermoplastics for use in high performance component applications,\nsuch as automotive and aerospace, has meant that product quality has been the prime focus of process optimisation.\nAs the manufacturing processes associated with polymer processing have become more mature, there has been a cor-\nrespondingly greater utilisation of in-line sensors, and adoption of Industry 4.0 protocols. This has enabled greater\nunderstanding of the material performance under processing temperatures and pressures, thereby providing the neces-\nsary input data needed for high fidelity computational modelling. In the field of computational optimisation, there have\nbeen signficant advances in algorithms development, with the result that a much bigger class of multi-variable and\n*Corresponding author. +441613062540\nEmail address: chamil. abeykoon@manchester.ac.uk (Chamil Abeykoon)\nPreprint submitted to Elsevier\nMay 20, 2021\nmulti-objective problems can be addressed. As a result, the possibility to broaden the scope of process optimisation\ncan now be grasped.\nClearly, for both high volume and high performance applications, energy efficient manufacture is a desirable\ngoal, which not only leads to reduced manufacturing costs but also addresses National and International energy and\nCO2 reduction targets. As a result, scrutiny of the energy required in an energy intensive manufacturing process\nsuch as the extrusion process is driven by both business and environmental imperatives. The payment of energy\nbills for unnecessary usage reduces profit margins and hence increases the end product/service prices for customers.\nMeanwhile, because CO2 emissions are now very clearly understood to be detrimental to the environment, energy\nusage will be increasingly subject to disincentives such as high fuel commodity pricing and taxation.\nThe scale of the plastics industry is internationally huge, and expanding. For example, in 2015 in the UK [1], there\nwere of the order of 6,200 plastics companies, employing nearly 170,000 people, and with a combined annual sales\nturnover of over \u00a323.5 bn, of which one third represented exports. According to the reports of PlasticsEurope [2], by\nthe year 2016 the European plastics industry comprised of more than 60,000 companies, employing more than 1.5\nmillion people, and with total sales exceeding 350 EUR bn. Globally, plastics production has grown from 204 to 335\nmillion tonnes between 2002 and 2016. The statistics presented in Figures 1 and 2 illustrate this growing demand.\nMillion tonnes\n350\n300\n250\n200\n150\n100\nPE\nPP\nPVC\nPS-EPS\nABS-SAN\n50\n01\n2005\n2011\n2012\n2013\n2017\n2020\n2025\nFigure 1: Major thermoplastics: World demand distribution, by polymer between years 2005-2025 [3]\n%/year\n6\n2005-2012\n2012-2017\n5\n4\n3\n2\n1\n0\nPE\nPP\nPVC\nPS-EPS\nABS-SAN\nTOTAL WORLD'\nFigure 2: Major thermoplastics: World consumption growth rate, by polymer (2005-2012 and 2012-2017)[3]\nGiven this level is sustained, the level of growth in demand, and the development and accessibility of new poly-\nmer processing capability, it is clear that improvements in process energy efficiency could have a significant impact\non global energy savings [4, 5]. Furthermore, the European Best Practice Guide [6] claims, \u201cPlastics are the material\nfor the 21st century\u201d, explaining that a 3 Megatonne CO2 emission reduction could be achieved in Europe by a 10%\nreduction in the plastics industry energy consumption. With the current capacity of polymers and plastics manufac-\nturing sector, it is one of the largest energy consumers in industrial manufacturing and also a major source of global\nwaste generation. Meantime, the energy savings/optimization in the manufacturing sector is considered as one of the\nmain pillars of modern circular economy concept and both manufactures and consumers have been forced to re-think\nthe current take-make-waste extractive industrial model for reusing materials form end-of-life components/devices,\nwhere polymers/plastics industry is one of the major focuses of this concept [7].\n1.2. The polymer extrusion process\nA polymer \"extruder\u201d machine processes materials by forcing them through a set of processing stages. The\noperation and basic processing stages are described in Figure 3 below. The screw passes material through a cylindrical\nHopper\nBarrel\nBand type\nheaters\nControl unit\nDie\nGear box\nDrive\nmotor\nCooling fans\nScrew\nSolids conveying\nMelting Melt conveying \u00a6\nFigure 3: Operational schematic of a single screw extruder\nbarrel, around which heaters are wrapped, to provide the necessary heat for material melting. In addition to this\nexternally provided heat, a significant amount of heat is generated internally, inside the barrel, as a result of the\nmechanical work of the screw (i.e. the work done against viscous and frictional forces). The feed material absorbs\nheat as it is conveyed along the screw and is expected to be in the fully molten state at the point that the molten\nmaterial is forced into a die to form into the desired shape.\nCurrently, different types of extruders (e.g. single screw, multi screw, and disc/drum types) are available in in-\ndustry, while screws with different geometrical designs are commercially available. Moreover, a number of process\nmonitoring devices are used, to observe process functionality, and for diagnosing possible processing problems.\nFrankland [8], President of Frankland Plastics Consulting, LLC, explains this in detail in his on-line article about\nestimating extrusion melt temperature. The most significant points are that the mechanical energy feed into the drive\nis converted by the screw action on the material to create heat, and thus melting of the polymer. The energy share\nrequired for material conveying is relatively smaller, as is the energy supplied to the barrel heaters. He also lists energy\nlosses and their sources. More details on the polymer extrusion process and its operation can be found in the literature\n[9, 10, 11].\nManufacturing process stability is a key concern, and variation in the material temperature presents a challenge to\nthe end product quality control. For this reason most commercial polymer producers avoid operating their extruders at\nat the higher screw speeds. This is unfortunate since at higher speeds, and therefore at higher workpiece temperatures,\nthere is more potential for process energy efficiency improvement because the material viscosity is reduced and thus\nthe forming forces required are lower. Moreover, this undesirable cost is repeated, since many thermoplastic polymers\nare extruded more than once before their final products are manufactured [12]. Better concatenation of extrusion\nprocess steps would lead to greater energy reduction, by maintaining or controlling the heat in the workpiece during\nprocessing, thereby avoiding the need to re-heat.\n1.2.1. Basic processing mechanisms\nZones within the polymer processing screw can be broadly designated, based on the functional activity taking place\nwithin that zone, see Figure 3. The points of transition between zones are not generally well defined, as they depend\non the processing conditions and the materials.\na. Solids conveying\nIn this zone, the polymer is preheated before passing into the subsequent zones. While flowing along this zone,\nmaterial starts to absorb heat from the barrel heaters, but the mechanical heat generated by frictional and viscous\n3\nmechanisms is dominant in this zone [13, 14, 10, 15]. Generally, the screw channel depth is maintained constant in\norder to provide a constant material feed to the subsequent zones.\nThe first comprehensive theory for the action of solids conveying was developed by Darnell and Mol [16] in the\n1950s and this quantitative description still remains as the widely accepted model for solids conveying in extrusion.\nb. Melting or Plastication\nExperiments for studying the polymer extrusion melting mechanism were first carried out by Maddock and Street\nin 1959 [17]. The melting mechanism proposed by Maddock for single screw extruders still remains as the most\nwidely accepted melting mechanism in polymer extrusion. The Maddock melting mechanism is only a qualitative\ndescription of melting which occurs in single screw extruders. Maddock used a visual inspection method to investigate\nthe melting process by stopping the screw rotation suddenly during the process and 'freezing' the polymer by cooling\nthe barrel and screw rapidly. Later, Tadmor also extended the understanding of melting mechanism of extrusion\nprocesses [18, 19, 20, 21].\nAs the material reaches the \"end\" of the solids conveying zone, it begins to melt, and as such is considered to\nhave entered into the melting zone. As the material becomes soft, further heat will be added to the process by means\nof viscous dissipation of the material (i.e. work done against the viscoelastic nature of the material). Both solid\nand molten polymers co-exist in this zone. The solid bed would comprise both compacted solid polymer abutting\nthe \"trailing flight\u201d, and the melt pool pushing against the \u201cpushing flight\", as shown in Figure 4. As the material\nFlow direction\nSolid/Melt\ninterface\nPushing\nflight\nCirculating\nmelt pool\nExtruder barrel\nTrailing\nSolid bed\nflight\nScrew\nFigure 4: An illustration of the typical arrangement of the solid bed and melt pool inside a screw channel for a single-flighted conventional screw\nproceeds along the screw, the proportion of melt pool to solid bed increases. The screw channel depth is therefore\ndesigned to become smaller, which influences flow rate and mixing; however, the actual screw length at which melting\noccurs depends on a range of parameters such as screw geometry, operating conditions and physical properties of the\npolymer [20].\nAs was claimed by Severs [22], the plastication or melting process has a direct impact on the quality of the material\nproperties of final product, and thus must be carefully controlled. Tadmor, Klein and Gogos [18, 19, 21] proposed an\nequation for calculating the rate of melting, (Q), in a screw channel, and is given by Eq. (1).\nQ2 =\n-\n[Pm \u00d7 Vbx { km (Tb \u2212 Tmelt) +\u014b\n2 {Cp (Tm-Ts) + 1}\n11/2\n(1)\nWhere Pm is the melt density, Vbx is the transverse component of the barrel velocity, km is the thermal conductivity\nof the molten material, Tmelt is the melt temperature, T is the barrel temperature, \u014b is the melt viscosity, V; is the\nresultant relative velocity, Cp is the polymer specific heat capacity, and \u03bb is the temperature of the solid bed.\nThis equation clearly demonstrates that the melting rate can be increased by increasing the screw rotational speed\n[14]; however, for higher speeds, the polymer passes through more quickly, giving less time for temperature stabilisa-\ntion, and thus more variation in melt viscosity [10, 14]. As a result, to ensure controlled plastication, it is necessary to\ncontrol the melting rate, and this in turn depends on the material being proceeded, process set conditions, and nature\nof the processing unit/machine [23].\nc. Melt conveying\nMelt conveying starts as complete melting is achieved. The screw channel depth is constant along the zone and\nis shallower than in the other two zones. During this stage further heating and mixing of the melt takes place as the\npolymer is smeared by the tip of the screw flight against the barrel wall. Material has to be moved towards the die\nwith enough force to overcome the head pressure generated at the die - this is known as the \u201cdie head pressure\".\nMelt output rate from this zone depends on a combination of two main factors: the rate of the rotation of the screw\nand the screw channel pressure gradient [24]. Proper mixing of material is another requirement for the flow through\nthis zone. The melt conveying zone of some of the new screw designs is fitted with efficient mixer units to ensure\ngood mixing performance (e.g. the barrier flighted screw with a Maddock mixer).\nStudies on melt conveying operation of extrusion were reported very much earlier than in the other two zones.\nOne of the initial studies was carried out in 1920s [25, 26], which proposed the calculation of the melt conveying rate\nby considering the melt flow as a laminar fully developed flow. This is still a widely accepted model.\nIn addition to the above mentioned mechanism/theories, several other works have been reported later on improving\nthe understanding of these three main mechanisms and more details can be found in the literature [14, 9, 10, 11, 27].\n2. Energy required for materials processing\nThe assessment of energy requirements is not straight forward. The overall extrusion process can be broken down\ninto smaller activities, but even then, the power demands at each stage depend in a complex way on a large number of\nprocessing parameters. A useful energy flow model was developed by Severs [22], as presented in Figure 5.\nEquipment cooling\nLosses\nit\nCooling\n\u2191\nPolymer\nsolid\n\u2192 Melting\n\u2192 Forming\nSolidification\nPolymer\nproduct\nMotor power (mechanical energy)\nHeating system (thermal energy)\nElectric power\nFigure 5: Typical energy flow diagram for an extrusion process\nThe energy, Eu, used by an extruder for useful work in material melting and forming, [28], is given by Eq. (2):\n=\nEu Ein Elosses\n(2)\nwhere Ein is the energy input to the extruder and Elosses is the energy expended that does not contribute to the extrusion\nprocess. Thus, the energy efficiency can be given by Eq. (3):\nnextruder =\nEin - Elosses\nEin\n\u00d7 100%\n(3)\nIn these equations, the energy inputs (Ein) should be related to the energy consumed by the electrical components\nsuch as drive motor, barrel/die heaters, barrel/motor cooling fans, water pump/s, instrumentation in the control unit,\netc.. The energy losses are always associated with all the components and also occur due to forced cooling and via\nnatural convection and radiation, which can be accounted under Elosses. In general, the drive motor and the barrel and\ndie heaters are the source of the highest energy losses. In typical polymer extrusion processes, recovery of such lost\nenergy is impractical, as this is largely released as heat energy to water or air. More details concerning the energy\nrequired for polymer processing and the thermodynamic efficiency of an extruder have been discussed by the authors\npreviously [28].\n3. Prior art in extruder energy evaluation, monitoring and modelling\n3.1. Energy Consumption studies\nIn considering the energy consumption in any industrial process, the first step is to review the process capability\nof the existing or available plant machinery, and the power consumption. On that basis, potential modifications or\n5\nenhancements to machinery could be identified, where an economic case could be made on the basis of energy cost\nsaving. Some examples are given below.\nIn the late 1970s, Chung et al. [12] found that for a 63.5 mm diameter extruder mechanical energy efficiency of\n62% was typical, and for larger extruders the energy efficiency was lower. In 1981, Kruder and Nunn [29] claimed\nthat energy efficiency of extruders can range from 45%-75%. It was noted that the energy efficiency depended on the\ntransmission mechanism, screw design, product geometry, nature of polymer feedstock and the production rate, while\nthe major energy losses of an extruder occur as a result of the forced cooling process step, and the losses associated\nwith the drive and transmission unit. At low screw speeds, barrel heaters consume a considerably higher portion of\nenergy than at higher speeds, and significant energy savings could be made by running the processes at the highest\npossible power factor. Additionally, this work presented information on energy demand and losses of each individual\ncomponent of an extruder.\nSubsequently through the 1980s, most research into energy efficiency was focussed on the screw efficiency and\nmass flow rate. A reduction in the overall power requirement for an extruder can be achieved through the use of a\ngeared pump at the end of the extruder to increase the mass flow rate (McKelvey [30]). In 1985, Strauch et al. [31]\ncarried out an energy consumption study on a 63.5 mm diameter single screw extruder, and observed that most of the\nenergy was consumed by the mechanical parts, with less significant levels of consumption in process heating. The\nenergy conversion was then assessed and it was found that heating the water in the cooling system accounted for more\nthan half of the energy supplied.\nDuring the late 1980s and 1990s, the manufacturing sector was making many changes, with a view to improving\nproductivity and quality. Driven by the advances made in Japanese manufacturing, the main focus during that time\nwas on management methods, such as Total Quality Management, LEAN, and Six Sigma. These efforts initially\naddressed cost and time issues, where the biggest economic benefits were to be found. Latterly, interest in energy\nefficiency began to be seen as not only cost reduction opportunity but also as an environmental imperative.\nIn the context of power consumption in the extruder, in 1997 Anderson et al. [32] recognised that for the processing\nof most plastics, from room temperature, the specific energy consumption (SEC) of the extruder motor should be in\nthe\nrange of 0.0822 to 0.1644 kW.hr/kg.\nAt around the same time, a study by Falkner in 1997 [33] showed that motor operations accounted for over 65%\nof the 1994 UK industrial electricity usage. Asserting that more than 10% of this energy could be attributed to\ninefficiency, Falkner argued that this represented a loss of about \u00a30.5 billion to the annual UK economy. These values\naccounted for motor energy utilisation across multiple industrial sectors, but it should be recognised that the electric\nmotors in plastics industry processing machines are a major power consumers.\nA more detailed study by Rosato et al. in 2001 [34] observed that energy losses of between 3 and 20% can arise in\nthe transmissions and control systems. Despite this, a conclusion was made that because plastics have lower specific\nenergy requirements compared with most conventional raw materials, they are still highly competitive.\nFive years later, Womer et al. [35] considered the energy efficiency of extruder cooling. The results demonstrated\nthat water cooling systems consume more energy compared with air cooling, irrespective of the particular plastic being\nprocessed. As a result, a recommendation was made to use air only cooling unless extensive cooling was expressly\nrequired.\nIn 2010 [36], the plastics industry was recognised to be one of the major UK industries with a similar trend\napplying globally. On that basis any improvement in process energy efficiency would lead to a considerable reduction\nin global energy requirement. Also in 2010, Cantor [37] presented measurements of SEC, where the impact of the\nmotor and of each individual heater zone, with respect to the overall specific energy consumption, was separately\nrecorded. It was observed that the heaters account for over 95% of the supplied energy. In a slightly later study\nby Heur and Verheijen [38], the authors studied differences from one plant to another, and recommended the use of\nfrequency controllers to enable more precise process control.\nThe earliest mention of an Industry 4.0 implementation to energy efficiency control was by Jing et al. (2014) [39]\nwhich proposed the use of real-time monitoring. The rationale was to render unnecessary the installation of power\nmeters or the development of data-driven models. A fuzzy logic controller controlled the high melt quality in a single\nscrew extruder, and was shown to be a cheaper alternative to using a gear pump. This also paved the way for achieving\ngreater extruder energy efficiency by optimising the temperature settings.\nA number of other works [35, 40, 41, 42, 43, 44, 45] consider the drive motor efficiency compared with other\ndevices. The conclusion to be drawn is that the drive motor should be the primary design consideration for process\n6\nengineering the energy efficiency of the whole extrusion plant.\n3.2. Influence of process set parameters\nIn 2001, Rauwendaal [10] recognised the significance of process settings, and presented an account of a procedure\nto minimise power consumption. A little later, in 2003, Rasid and Wood [46] investigated the influence of individual\nbarrel zone temperatures and found that the solids conveying zone temperature had the greatest influence on overall\npower consumption.\nStudies carried out between 2004 and 2012, [47, 48, 49, 50], examined various process parameters and their\ninfluence on SEC. In addition to noting the effect of material viscosity, variation in energy consumption was also seen\nfor different designs of screw, and there were greater melt temperature fluctuations at higher screw speeds: another\nexample of the ever-present tension between cost and quality. The simultaneous need to achieve both energy efficient\noperation and finished part quality remains a challenge.\nStudies carried out by Abeykoon et al. [51, 28, 52, 53] between 2009 and 2016, focussed on the relationship\nbetween the process energy demands of the motor and barrel heating and melt thermal stability. The effects of the\nsettings for these processes, the screw geometry and choice of material were explored.\n3.3. Modelling\nFollowing a thorough trawl of the published scientific literature, it has become clear that relatively little work has been\nundertaken to model extruder energy consumption.\nThe earliest work in this area was by Mallouk and Mckelvey [54] in 1953, where a mathematical equation was\ndeveloped, based on assumptions of isothermal, Newtonian flow, in a screw channel with constant section. In 1996,\nWilczynski [55] developed a computer model where the five zones of the extruder plus the die were considered\nseparately. Subsequently, in 2000, Lai and Yu [56] also proposed a mathematical model for the calculation of energy\nconsumption based on screw speed, and including viscosity. In Abyekoon et al.'s [57, 52] studies of a single screw\nextruder, the data collected was analysed using static nonlinear polynomial models. The conclusion of the analysis\nwas that choosing energy efficient process settings would also lead to thermal stability.\nObviously, the availability of advanced modelling methods for predicting energy consumption, based on process\nparameters, would enable process operators to select optimum operating conditions. In particular, models which\nincorporate both energy consumption and melt thermal quality would be preferred but the development of such models\nis quite challenging. Melt thermal quality and energy efficiency present opposite behaviours with respect to the\nprocessing speed: the thermal quality deteriorates while the energy efficiency improves. Since the industrial sector\nhas to meet strict environmental regulations to minimize the carbon footprint, any reduction in the energy demands\nfor polymer processing would support future sustainability.\n3.4. General considerations in energy usage\nAccording to basic electricity principles, the typical power consumption of a DC and an AC device (PDC and PAC) is\ngiven by equations (4) and (5), respectively [58, 59],\nPDC = V XI\nPAC = VXIX cos\n(4)\n(5)\nwith I being the supply current, V voltage, and cos & the \"displacement power factor\u201d. From these, the power demand\nof any device in an extrusion plant can be evaluated; however, the energy losses related to each device might vary\nfrom component to component.\nThe power factor is an important consideration in the assessment of the energy usage of an electrical machine or\nprocess, and is defined as either the \"displacement power factor\" which is the cos o in Eq. (5) or the \"true power\nfactor\" which is given by Eq. (6).\nTrue\npower factor\n=\nTrue (or active) power\n(6)\n7\nApparent power\nImpedance\nApparent power (S)\n(units: VA)\nphase angle\n(units: VAR)\nReactive power (Q)\nActive power (P)\n(units: W)\nFigure 6: Power triangle showing the relationship between active, apparent and reactive powers\nThe true power factor lies in the range 0 \u2013 1, for which the running of the machine or process with true power factor\nequal to one would be the best possible energy efficient operating condition (when the impedance phase angle shown in\nFigure 6 is equal to zero). For a true power factor of less than one, the energy supplied to the load is not used optimally.\nIn such a case, a higher current must be drawn to compensate for the phase shift, o. Where industrial customers operate\nwith power factors below around 0.95, [60], this represents unbalanced additional power demands from the power\nsupplier, and hence additional infrastructure demand leading to additional costs. Furthermore, electrical devices are\nattributed with a 1\u00b2\u00d7R heat loss (R is the electrical resistance), so that increasing the required current while reducing\nthe power factor results in an increase of power loss as heat.\nMeasurements of power factor and total power consumption, for a DC motor driven 63.5 mm in diameter single\nscrew extruder operated at different screw speeds, are shown in Figure 7.\nPower factor\n30\n0.2\n+\n0.8\n0.6\nTotal power (kW)\nSS (rpm)\n\u00a6(a)\n0\n35\n\u00a6(b)\n30\n20\n10\n100\n(c)\n80\n60\n40\n20\n0\n0\n50\n150\n250\n320\nTime (s)\nFigure 7: (a). Power factor, (b). Total extruder power, (c). Screw speed [52]\nFrom this, it can be seen that both the power factor and the total power required are greater for greater processing\nspeeds; however, for higher processing speeds, the temperature uniformity of the process melt output deteriorates\nsignificantly, leading to poor product quality, [28, 52, 53]. Hence, the running of these processes at higher speeds and\n8\nwith the highest possible power factor is problematic, despite being desirable for energy efficiency.\n4. Potential for energy efficiency improvements\nEnergy demands and losses are illustrated in the form of an energy flow diagram, Figure 8. This diagram may be\nEnergy content in\nthe feed material\nDrive motor\nEnergy used\nfor material\nmelting and\nExternal\nheating/cooling\nOther losses\nEnergy for\nother auxiliary\ndevices\nForced\nDrive motor\nlosses\nTransmission cooling\nlosses (gear\nlosses\nbox)\nNatural\ncooling\nlosses\nforming\nFigure 8: A typical energy flow diagram for an extruder [52]\nextended for any auxiliary devices connected with the plant.\n4.1. Drive motor and gear box\nThe key component of any extrusion machine is the screw, which can be driven by a controllable direct current\n(DC) or an alternating current (AC) motor, or indeed by a hydraulic drive [10, 61]. The screw and the motor are\nconnected through a gear box with fixed or adjustable transmission ratio, as shown in Figure 3. For the case of\nan extruder with a DC motor drive, see the schematic, presented in Figure 9. Additionally, the machine can have\nsensing and control devices related to its operation, for example, PID temperature controllers to control set barrel/die\ntemperatures. Extruders with AC motor drives are essentially similar (with no rectifier), and may have additional\ncomponents depending on the type of the motor.\n(@set\nDactual)\nSet screw speed\n(@set)\n+\nPID Motor\nspeed controller\nActual screw\nspeed (actual)\nDC motor\nGear box\nScrew\nArmature voltage (Va)\nchanges to adjust the\nmotor speed\n@actual\nTachometer\ngenerator\nFigure 9: A schematic of an extruder drive mechanism\nFigure 10 shows measured motor power and total power consumptions, for the case of a 63.5 mm diameter single\nscrew extruder with a DC motor, driven at different screw speeds. The contribution of heaters to the total power\ndemand is also indicated. As marked on Figure 10, all the heaters were turned off at around 330 s and this has led to\nsmooth out the total power signal which were fluctuating due to the on-off action of the barrel/die heaters.\nThe drive motor is one of the major energy consuming components of an extruder [31, 34, 40], and, along with the\ngear box, is also responsible for significant energy losses, typically accounting for around 20% of the power supplied\nto an extruder [29]. In particular, DC motors are inefficient when operated at below the rated speed. Commercially\navailable DC motors fall into three main categories: \u201cpermanent magnet\u201d, \u201cseparately excited\u201d and \u201cself-excited\".\nThe first two are more commonly used. A block diagram for a polymer processing extruder with a \u201cseparately excited\ndirect current\" (SEDC) motor is shown in Figure 11.\n9\nPower (kW)\nSS (rpm)\n100\n60\n(a)\n100\n200\n300\n400\n500\n30\n600\n(b)\nTotal power\n20\n0\nMotor power\nAll the heaters turned-off\n-10\n0\n100\n200\n300\nTime (S)\n400\n500\n600\nFigure 10: (a). Screw speed (SS), (b). Motor power and total power signals over the time [52]\nTL\nElectrical dynamics\nGear Box\nVa\n1\nTm\n1\n@m\n@sc\n+\nK\u2081\n+\nLmS + Rm\nJmS+Bm\nN\nV\u2081\nEb\nMechanical dynamics\nKf\nKm\nSpeed controller\nFigure 11: Block diagram of an extruder with a variable field DC motor [62]\n10\n\n\nIn this figure, T is the motor torque, T is the load torque on the screw, Ra and La are the armature resistance and\narmature inductance respectively, K, and K, are the torque constants related to the field and motor, respectively, Ia is\nthe armature current, V+ is the field voltage and Bm and Jm are the damping constant and the steady-state inertia of the\nloaded screw, respectively. For extruders with a permanent magnet motor, the same block diagram is valid without\nthe branch related to the separately excited field (with V\u0192 and K\u0192 block).\nReasons for the popularity of DC motor drives in the polymer processing industry [63, 64, 52] include:\n\u2022 Smooth operation over a wide speed range,\n\u2022 Simplicity in speed control,\n\u2022 Production of a constant/consistent torque from zero to base speed,\nRelatively low power/energy consumption,\n\u2022 Relatively smaller size compared to other drive types with the same capacity,\n\u2022 Compact and simple power circuit, engaged with Silicon-controlled rectifiers,\n\u2022\nEasy installation,\n\u2022 High reliability,\n\u2022 Low intial capital cost, and\n\u2022 Less noisy than AC motors.\nDrawbacks of DC motor drives include the need for maintenance of brushes and commutator as well as the energy\nloss known as \"brush loss\u201d [59]. Green [63] observed that the best power factor that can be achieved by a DC motor\noperated at its top speed is of approximately 0.87, whereas for the best possible efficiency the power factor should be\nclose to 1.\nCurrently, AC motor drives are increasing in popularity since the power factor can be maintained constant across\nthe entire speed range. As a result, companies can avoid paying penalty charges for lagging power factor conditions.\nThe most significant drawback of AC motors is that they require a constant current to produce a constant torque, hence\ndemanding constant cooling regardless of the motor speed [63], and this results in an additional energy cost. Other\nissues include the fact that AC motors are generally larger, by a factor of 1.5-2.1, in volume, as well as being more\ncomplex than DC motors. These issues are becoming less critical nowadays, thanks to advances in electronics such\nas large scale chips and micro-processors. For both types of motors, while under operation at the rated speed, the\nmaximum energy efficiency can be achieved, industrial extruders are typically operated at lower speeds in order to\navoid undesirable fluctuations, particularly of the melt thermal quality.\nBarlow [40] argues that because the displacement power factor of a DC motor drive is proportional to the speed,\nthe power factor reduces as the motor slows down (see Figure 7). Further, it is pointed out that because the diode\nbridge of the input section of a pulse-width-modulated AC vector control drive rectifies the AC into DC, and that the\nenergy is stored in capacitors, the current and voltage waveforms are mutually in phase, and hence the motor operates\nat a power factor, in the range of 0.90 to 0.98. More information on these motor drives can be found in the literature\n[65].\nIt seems that a significant amount of electrical energy may be lost simply as a result of the low power factor\noperation of motor drives [44]. Here, Eickelberg [66] suggests that use of capacitors may be one of the solutions to\nthis problem, to smoothen the power supply. Several examples of the use of capacitors by commercial processes are\nprovided, but it is stated that this is unlikely to be a practical solution for polymer extrusion processes because of\nthe variability that occurs in the load. The installation of a more appropriate form of power factor correction would\nrequire investigation of the relevant issues [67].\nKent [41] observes that in extrusion plant energy usage assessments the energy requirements of motors in equip-\nment such as extruders and injection moulding machines is often over-looked. Other authors [68, 69] report that\nconsiderable energy savings can be achieved by replacing DC motors with AC motors. Lounsbury and Karafilidis\n[70] present factors to be considered in the selection of a drive motor.\n4.2. Barrel and die heaters\nNormally, three different types of heating method can be identified in extrusion. These are known as \"resistance\nheating\u201d, \u201cinduction heating\u201d and \u201cfluid heating\u201d [61].\n11\nResistance heating: This is also called electrical heating, and is the type most frequently used in extruders.\nUsually, electric heaters offer several advantages over fluid and steam heating, such as the possibility of covering\na broader temperature range, cleanliness, easy maintenance, low cost, and better efficiency. As a result of these\nadvantages, fluid and steam heaters have been replaced by electric heaters in modern applications. Currently extruders\ntypically have between two and ten heating zones depending on the size of the extruder.\nIn the most common conventional resistance heaters, the heat from the resistance wire is transferred to ceramic\nsegments that surround the outer surface of the barrel. This heats the barrel up until its inner surface is hot: the heat\nis then transferred to the plastic in the machine so it can be processed. With this heating method, much of the heat\ngenerated is wasted.\nInduction heating: In this case, an AC current is passed through the primary coil surrounding the extruder barrel.\nThis gives rise to an eddy current. Where the material being processed has significant relative permeability, heat may\nalso be generated by magnetic hysteresis. A high energy density can be achieved quickly with induction heating. The\nfrequency of the AC is selected depending on the size and material type and the heat penetration depth.\nWith the advantage of rapid heating and energy efficiency, induction heating has been used in many industrial\napplications [71]. The key advantage of induction heating over resistance heating is that the extruder barrel itself\nbecomes the heating element. This eliminates the conduction problem that exists with conventional heaters. As there\nare no ceramic layers, clamping bands, or water jackets to heat up, the heat is direct and instantaneous. Furthermore,\nthe induction heating generates a very even and precise heat profile, ensuring consistent heating of the polymer melt\nand leading to improved product quality. The application of induction heating to polymer processing has been devel-\noped by the Nordson Xaloy Company, which claims a reduction in heating related energy consumption of up to 50%\ncompared with typical band electric heaters [72, 73].\nFluid heating: Fluid heating uses hot liquid or steam passing through pipes/tubes. This can ensure even temper-\nature distribution but has significant disadvantages, such as demanding high levels of maintenance, the possibility of\nleaking or corrosion, and system complexity.\nAn alternative all three heating methods, is to supply the energy to the material being processed as drive power\nrather than as heater power [44, 74]. In this way shearing of the material leads to both heating and mixing. Of\ncourse, direct heating and cooling can be essential, to maintain the process thermal stability, but as guiding principles\nit would be preferable to minimize the direct heating and to avoid exceeding the melt temperature, in order to avoid\nunnecessary energy costs.\n4.3. Control electronics and monitoring devices\nA number of control and monitoring devices are used in extrusion lines, such as speed and temperature controllers and\nindicators; pressure indicators and gauges; dimension scanners; feed monitoring devices; current indicators; relays;\nswitches; alarms; etc.. All of these use some power for their operation; however, this is insignificant compared to the\ntotal energy demand. This is evident from the experimental data presented in Figure 10-(b), the amount consumed by\ncontrol electronics and monitoring devices are shown by the difference between blue and red lines (i.e., the difference\nbetween to total and motor powers) after turning off all the heaters, or by the blue line roughly between 540-570 s\nafter turning off all heaters and the motor.\n4.4. Auxiliary equipment\nAuxiliary equipment, such as pelletizers, gear pumps and screen changes, might also consume enough power to be\nconsidered in energy evaluations. Rice [75] suggests that improved energy efficiency for the entire operation of the\nplant can be achieved by combining processes.\n4.5. Process cooling\nProcess cooling is required where there is a need to remove excess heat in order to maintain the process thermal\nstability. This can be achieved by fan coolers attached along the barrel, or by cooling the screw core or the barrel wall\ninternally, using a cooling fluid such as water or oil. Air cooling provides slower changes in temperature compared\nwith liquid cooling.\n12\nAlthough cooling helps to ensure stable process operation, improper cooling, particularly cooling of the screw,\ncan lead to the generation of undesirable process fluctuations. Strauch [31] argues that excessive screw core cooling\ncan lead to a reduction in throughput rates, while at the same time affecting the melt temperature. This also impacts\non the pumping stability as a direct consequence of altering the viscosity of melt. Periodic or random temperature\nvariations of the extruder metal surfaces can arise as a result of cooling problems with the screw or barrel, and these\nmay lead to melt flow problems such as melt viscosity fluctuations [76].\nThe research of Womer et al. [35] into the effects of cooling indicated that, where water cooling is used rather\nthan air cooling, the extruder consumes more energy, irrespective of the material being processed. In consequence,\nit is recommended to use air cooling only, in conjunction with a properly designed screw, unless extensive cooling is\nrequired.\n5. Trends in polymer processing energy efficiency improvements\n5.1. Machine development and operational modifications\nAmong the current research and development \"hot topics\" for process energy efficiency, the three most likely to deliver\nsignificant benefits [63] are: (i) the design of direct drive extruders, (ii) improvements in heater and barrel design to\nreduce heat losses, and (iii) thedevelopment of \u201cadvanced vector control alternating current (AC) drives\u201d. In addition\nto these, there are reports which claim that direct drive machines offer further advantages including energy efficient\noperation, narrow footprint, quiet operation, and low maintenance requirements. The use of insulation blankets has\nalso becoming popular in energy saving of machines.\n\"Load management\u201d offers the possibility to save power in the near term. The idea is to achieve the required energy\ncost reduction by maintaining the load factor, rather than focusing on power consumption per se. Since commercial\nelectrical energy supply rates usually depend on the peak demand made by the customer, this can reduce the cost of the\nelectricity used over the duration of the charging period. To operate an effective load management plan, it is necessary\nto have a plant monitoring system to study the real-time plant power usage. In Kent's [41] discussion of energy saving\nin polymer processes, it is pointed out that the purchase of energy efficient capital equipment is profitable in long-run\ndespite initial capital costs.\n5.2. Waste heat energy recovery\nAlthough a significant amount of process heat is removed purposely to maintain the thermal stability in polymer\nextrusion, there has been insufficient attention on the recovery of waste heat for useful work.\nThe challenge [42, 77] is to find a means of re-use for that energy. Future research is required to explore such\nopportunities. For example, the recovered heat might be used for pre-heating the material prior to feeding into the\nhopper, or it could simply be used for space heating.\nFor some materials, resins need to be pre-heated prior to processing to remove moisture. In this case, part of the\nsupplied energy is lost through the evaporation of the moisture as water vapour, part is lost to heating of the surround-\nings, and the rest contributes to heating the resin. If the drying operation takes place remotely from the processing\nmachine, then the heat absorbed by resin will be lost during transit to the processing machine. Therefore, the re-design\nof the drying system, as an in-line step of the processing machine, should help to reduce overall processing energy\ncosts.\n5.3. Material choice, material recycling and disposal\nIt is usually the case that the manufacture and fabrication of plastic products makes lower energy demands than\nequivalent traditional metallic or glassware products. The development of new resins that can be processed at lower\ntemperatures, and hence for reduced energy, is part of the growing tide of interest to cut process energy expenses\neven further [78, 79, 80, 81]. On the other hand, polymeric materials based waste management has become a global\nconcern. A wide range of recycling techniques are available depending on the types of polymers and the manufacturing\ntechniques used [82]. In regard to process energy costs, the energy consumption for plastics is lower than for materials\nsuch as paper, glass, tin, and aluminium. What is more, Rosato et al. [34], the incineration of plastics as part of\nmunicipal waste yields much more energy than other material waste, such as food waste, paper and rubber, and waste\nvolume can be reduced by 90-98%.\n13\n6. Advanced process monitoring and control: adoption of Industry 4.0\nAdvanced process monitoring and control can play a vital role in achieving good product quality as well as in energy\noptimization. Some approaches are discussed in this section together with experimental results.\n6.1. Industry 4.0 and Internet of Things\nWidespread uptake of on-line or in-line monitoring and control in manufacturing processes has been enabled by\ncomputerised communications, earning it the epithet: \u201cThe fourth industrial revolution\u201d, or \u201cIndustry 4.0\u201d for short\n[83, 84]. The key requirements for an Industry 4.0 manufacturing process are: sensors - the means to observe the\nprocess as it currently stands; actuators - the means to modify the process; electronic communications - the means to\npass sensor or control information; and a decision-maker - to determine the course of action to be taken based on the\ninformation received.\nThe precise nature of the decision-maker is a point of some contention in the literature. For some, the decision-\nmaker is a computer-based Artificial Intelligence (AI), which would work completely autonomously, and steadily\nimproving its decision-making capability based on learned patterns of experience. There is still significant value in\nthe Industry 4.0 infrastructure even without an AI capability making the control decisions. For some applications, a\nrule- or model-based computer program would be effective, and many companies now boast of having an Industry\n4.0 implementation of this form. In some applications, having the sensor information fed to a control centre, means\nthat human decision-making can be facilitated and supported. Such control centres are valuable for the control of\nprocesses that are in remote or difficult to access locations, such as the health-monitoring of in-flight aircraft.\nIndustry 4.0 technology is also becoming an increasingly common tool in the home or in social care settings. Here,\nthe more common terminology is Internet of Things (IoT) [85], referring to communications connectivity between\n\"Smart\" devices. These smart devices are pieces of equipment which can send or receive information, in other words,\nthey are equipped with sensors or actuators. Thus, in the home, one might simply ask \u201cAlexa\u201d [86] to turn on the lights,\nbut in time it could easily be imagined that Alexa could modify the home heating to match the schedule inferred from\nfamily member diaries. In a patient care setting [87], the IoT system could be collecting valuable health-related\ninformation and relaying that to a control centre. Care staff or an AI system might detect anomalies and prompt a\ncheck of the patient. It is easy to see that these concepts are similar, whether applied to the home or to industry, and\nthe technology is pervasive and rapidly developing.\nIn the context of the polymer processing industry, it is clear that Industry 4.0 will not only enhance plant operation\nand its maintenance schedule, but also to energy efficiency [88, 89]. The lessons learned over the past half century,\nand reviewed in earlier sections of this paper, are ready to be applied. Wherever there are frequent variations in the\nfeedstock material, processing rate demand, or other factors, with an Industry 4.0 infrastructure in place, it becomes\npossible to monitor performance over time, and to make controlled changes. Systems health monitoring can be used\nto reduce life limiting loads on mechanical parts [90] as part of the maintenance strategy.\nGreater investment in IoT enabled devices will become an increasing imperative, for any polymer processing plant\nthat wishes to reduce its energy footprint, reduce processing costs, and sustain the mechanical plant more effectively.\nAs soon as the Industry 4.0 infrastructure is operational, development of the decision-making capability can be begun;\nwhether that is to be based on AI principles, control systems mathematics or other rule or model based paradigms.\n6.2. Ultrasound\nThe application of ultrasonic waves to reduce viscosity and thus energy consumption in polymer processing has been\ndiscussed in several recent reports [91, 92, 93, 94, 95, 96]. Maintaining a consistent melt viscosity enables improved\nprocess ability, leads to fewer product defects, reduces energy consumption and reduces materials wastage [97, 98, 99].\nChen et al [92] and Zhang and Li [93] report on the use of ultrasound vibration to influence polypropylene (PP)\nmelt, leading to non-Newtonian flow characteristics with reduced viscosity. Other authors have examined the rela-\ntionships between temperature, pressure, work-stuff throughput, the energy consumption and the ultrasonic intensity\n[95, 96]. The integration of ultrasound into a closed loop extruder control system has been developed and introduced\nin detailed by Nguyen et al. [100], who showed that controlling just the temperature, or just the ultrasonic output,\ngave a better time response and reduced energy consumption, than when trying to control both the temperature and\nultrasound together.\n14\n6.3. Closed loop melt temperature control\nMelt temperature can serve as a proxy parameter for melt viscosity and can thus be used to determine the melt quality.\nRecent works by Abeykoon et al [101, 102, 103] incorporate a fuzzy logic approach for the real-time closed loop\ncontrol of melt temperature, and demonstrated excellent performance in achieving desired set temperatures.\nAt present, in the majority of polymer processes, melt pressure and melt temperature are taken as the key param-\neters for process functionally and control. Screw speed, and barrel and die set temperatures are taken as the main\nprocess control parameters, but there is no actual feedback taken from the process melt for making process control\ndecisions, so no corrective actions can be taken to avoid product defects. Furthermore, this affects the production rate\nleading to wasted energy, labour and raw materials. Hence, combined process monitoring and control approaches,\nwhich can observe the melt quality and take control actions would represent a major development of polymer pro-\ncesses.\n7. Applying computational process simulation and dynamical systems to Control\nDevelopments in computational process simulation methods has the potential to revolutionise the design of manu-\nfacturing tooling and processes. Capabilities such as finite element method and computational fluid dynamics have\ndeveloped very significantly, with most commercial packages (e.g. [104, 105, 106, 107, 108, 109]) now offering\nmulti-physics simulation: simulation of static and transient solid mechanics, fluid dynamics, thermodynamics, elec-\ntricomagnetism, and in many cases, much more. Given the improvements in computer hardware, storage capacity\nand the development of parallel processing architecutures, computational analyses with high levels of geometric or\nmaterial modelling complexity can now be readily envisaged. This is a huge opportunity to grasp, and one for which\nmost industrial plants are not fully prepared.\n7.1. Computational mechanics methods\nThe key computational capabilities pertinent to polymer processing include a variety of modelling techniques for vis-\ncoelastic materials irrespectively of whether they should be treated as solids or liquids. Where previously, modelling\nbased on a Lagrangian Finite Element formulation, [110], would have been unable to capture the necessary defor-\nmation, Eulerian formulations, [111], and Smooth Particle Hydrodynamics (SPH), [112] and [113], now have that\npotential. Perhaps it is not enough to model using just one technique or another, but to capture one pertinent aspect\nof the physics in one region of the process, and another aspect in another region using such modelling techniques as\nco-simulation and subdomain modelling.\nNot only it is the type of analysis that is of importance to capture the process simulation requirement, but also the\nway in which the material properties are represented. The distinction between solid and liquid is no longer so clear-\ncut. There are very many material models, developed and applicable to different analysis types, that will capture the\nessential material property physics of any realizable material. Polymers offer particular challenges, but characteristics\nsuch as static stress-strain, strain rate dependence, creep, and temperature dependence are readily modelled. Chemical\nchanges, including exo- or endothermic reactions, and cure shrinkage still present a particular challenge to thermoset\npolymers, but for thermoplastic processing the challenge of representing materials properties at glass transition, and\nextent of crystallization, are, if not straight forward, at least more tractable [114, 115, 116]. For the characterisation\nof polymers undergoing extrusion processing Abeykoon et al, [117], have made significant investigations.\nWith the increase of computing power and the capability for higher model complexity there has been increased\ninterest in the application of computational mechanics methods using sophisticate material models to the simulation\nof the extrusion process. In the past decade authors such as Zairi et al, [118], have been able to model plastic flow,\ntaking account of viscoelastic properties, within a die of finite constrained section, and have been able to make some\nprediction of the microstructure of the end product. While the main focus has been of the effect of the die geometry\non the extruded product, another important consideration is the design of the die tooling, and the loads that must\nwithstand during processing [119].\n15\n7.2. Statistical process control\nThe biggest challenge for the modelling of the polymer extrusion process is to understand the conditions that give rise\nto large variations in product output quality. Where small changes in the processing parameters lead to only small\nchanges in output, even where the change is clearly non-linear, one might take a step-wise approach to linearize the\nparameters and have a measurable and definable set of limits for process control. This is the ideal, textbook, approach\nto manufacturing process excellence, LEAN manufacturing or Six Sigma, [120], [121]. In constrast, where small\nchanges in processing input lead to significant changes in the output quality, a manufacturing engineer would say that\nthe process is not in statistical control. The usual manufacturing engineer's approach for regaining statistical control\nis to monitor parameters to within ever tighter bounds. Clearly this would lead to increased manufacturing costs, and\ncould make the process financially unviable. The approach may ultimately even be completely unsuccessful.\nThe factors that would generally be considered for statistical process control would include the initial conditions:\nhow the process is started up, and the material condition on start-up, factors directly related to the geometry of the\nextruder and the die, environmental conditions and changes, and changes in the raw material batch.\nIn order to control the process, certain actions might be taken. The applied drive torque, a function of the power\nsupplied to the extruder, could be controlled in direct response to measurements relating to the process quality. In-\nevitably there is a delay between making the measurement, and making a change, primarily because the measurement\nis taken at some point down-stream in the process. There is a considerable level of research activity in this area, with\nvarious computational schemes for assessing these measurements and informing the choice of control action to be\ntaken. Clearly this data processing time must be minimised to minimise the feedback delay. Wagner et al, [122], and\nMcKay et al, [123], recognised that the fundamental requirement was to use the measurements to infer the local mate-\nrial viscosity within the process, using artificial intelligence methods such as neural nets. Chen et al, [124], employed\na power law model. Methods based on fuzzy logic were developed by McAfee, [125], McAfee and Thompson, [126],\nand later by Liu et al, [127] and Abeykoon [103]. More recent work includes the soft sensor technique of Deng et al,\n[128] and Abeykoon [102], and the multi-objective optimisation approaches presented by Carrano et al, [129].\n7.3. Non-linear system dynamics\nWhere the dynamics of the manufacturing process make statistical process control challenging, a radically different\napproach is needed. Non-linear systems dynamics [130] has been a topic of considerable research and development,\nmainly by researchers with a strong background in applied mathematics, and with applications including but by no\nmeans limited to manufacturing engineering applications. By developing a non-linear system dynamics representation\nof a manufacturing process, it is possible to explore how process parameters could influence the process dynamics,\nand from this pin-point the controling factors or initial conditions.\nIn the field of polymer extrusion modelling, McKinley, et al, [131], used laser doppler velociometry to visualise\nthe flow towards an abrupt contraction, and found that the flow near the tip of the contraction could show time periodic\nand aperiodic behaviour. Graham [132] modelled the fluid behavious regarding wall slip and was able to replicate the\nlarge amplitude periodic and aperiodic oscillations observed experiementally. Smith et al, [133, 134] demonstrate a\nsteady-state solution through modelling the polymer as a purely viscous non-Newtonian material, and optimising die\ngeometry.\nThe critial feature in the extrusion process is the variation in the material property of the polymer as it passes\nthrough the process. As the material is being worked mechanically, it becomes heated. As it gets hotter, the elastic\nstiffness and viscosity are reduced. The resulting thermal expansion gives rise to a localised increase of pressure.\nAs a result of both the reduced viscosity and increased pressure, the polymer would pass more readily through the\nprocess: and would require less working. Reduced working would result in reduced heating, with the result of inceased\nstiffness, increased viscosity and reduced pressure. It is clear that a cyclic response can be expected. Now consider\nhow the viscosity of the polymer changes with strain rate and the difference of time dependency in glass transition and\ncrystallization, it bceomes clear that there is a level of complexity between production rate, localized temperature and\nviscous response of the material: in the simplistic mass, spring and damper system it is the variability of the damper\nas well as the forcing that is key to understanding the dynamical system.\nA non-linear dynamical systems model of the polymer extrusion process might be modelled as follows. First,\nthe polymer has the properties of both an elastic solid, and a viscous fluid: it can be idealized as a mass, spring\nand damper system, as shown in Figure 12. The non-linear dynamics of such systems have been studied for many\n16\nidealized situations of varying applied force, F(t), and their modelling involves the selection of initial conditions, and\nthen computing the transient behaviour until a long term behaviour becomes apparent.\nk(t)\nx(t)\nc(t)\nm\nF(t)\nFigure 12: Idealised mass, spring and damper system\n-\nFigure 13 shows the time varying position, x(t) and velocity x(t) of the mass, driven by the time varying force\nF(t). The initial conditions for the system might be any point on the x \u2014 x plane, and one might think of particular\nexamples of such initial conditions being the tail end of the solid or dashed curved arrows. As time progresses, the\nsystem would move on from that initial condition, moving towards the head of the arrow. The dashed arrows indicate\nthe evolution away from an unstable limit cycle \u2013 the dashed closed loop. Any initial condition that begins on the\nunstable limit cycle will eventually decay inwards or outwards, and migrate towards a stable limit of some sort. The\nsolid arrows indicate evolution towards either a stable limit cycle the solid closed loop, or towards a steady-state\nsolution \u2015 the solid dot. The stable limit cycle represents a periodic motion. The area enclosed by the unstable limit\ncycle gives an indication of the relative level of damping in the system: it is important to note that increasing the\ndamping would increase the area within the unstable limit cycle, so that a larger proportion of initial conditions would\nultimately lead to a steady-state solution; however in the case of initial conditions sufficiently close to the stable limit\ncycle, periodic motion would still ensue. The notion that damping removes energy from a dynamic system is true, but\nthe energy required for periodic motion is renewed at every cycle by the time varying force, F(t).\nIn the context of polymer processing, periodic motion represents both unnecessary energy utilisation, and reduced\nprocess control. This explanation is more easily understood when described in the context of everyday experience.\nx(t)\nx(t)\nFigure 13: Time evolving position and velocity diagram\nConsider a friction driven vibration such as the vibration of a bowed violin string, brake squeal, or a singing wine glass.\nIn each case, the string, the brake disc and the glass have a stiffness, a density and some form of damping is present.\nUnder particular forcing conditions such as applied by a well-rosined violin bow, a certain braking pressure or being\nstroked with a wet finger, these structures vibrate at a well-defined pitch. The fact that the pitch is well-defined means\nthat the vibration is periodic. Under different forcing conditions: a bow without rosin, a different braking pressure\nor a dry finger, the contact slips with constant velocity. This is a steady-state solution: a force is being applied and\nso there is a displacement, but it is a constant force and a constant displacement. For a detailed account of such an\nanalysis, and the computational algorithms used [135].\n17\nThis description is necessarily rather simplistic, but it provides a basis for reviewing the pertinent features required\nfor modelling the extrusion system. In place of displacement and velocity one should be considering the extruded\nvolume of material as a function of time. In place of force, consider the power output of the machine and what\nresistance it meets from the material being processed. In an ideal process, one would wish for constant extruded\nvolume and constant levels of resistance, so that with an appropriate choice of initial conditions and damping, a steady-\nstate outcome can be achieved. The elastic response and the viscosity of the polymer depend on temperature and strain\nrate, and as such are the time dependent system stiffness, k(t), and time dependent damping, c(t), respectively. Recent\nwork in this vein, but applied to blown film extrusion, is presented in a detailed review paper by Pirkle et al, [136].\n8. Conclusions\nThere have been some very significant developments in computational modelling capability, which means that the\nsimulation of manufacturing processes such as polymer extrusion is now feasible. Multi-physics approaches, combin-\ning material flow and thermal behaviour of the material can, in theory, be modelled. Advances in computer hardware\nmean that models with very high levels of geometric complexity, and therefore high numbers of computational degrees\nof freedom are within reach. Co-simulation modelling, including the modelling of not only the material undergoing\nextrusion, but also of the working state of the dies and the extrusion machine itself, is also within reach, meaning\nthat optimal design of tooling for improved life and reduced machine maintenance demands are additional areas for\nfurther development.\nComplex models of the process and material property variation can also be used to create the simplified models\nrequired for a non-linear systems dynamics modelling approach. On that basis, measurable parameters that would\nhave an effect on the stability of the real production process can be examined, and warning limits found. Even where\nsuch an approach might not be able provide an accurate prediction of limits, it could provide insight that would lead\nto practical solutions or avoidance of particular operating regimes.\n9. List of Abbreviations\nTerm\nDefinition\nAC\nAlternating current\nDC\nDirect current\nIoT\nInternet of Things\nSEC\nSpecific energy consumption\nSEDC Separately excited direct current\n10. List of Symbols\n18\nTerm\nDefinition\nBm\nDamping constant\n\u0421\u0440\nSpecific heat capacity\nEin\nEnergy consumed\nElosses\nEnergy loss\nF(t)\nA time varying force\nI\nLine current\nIa\nArmature current\nIf\nField current\nJm\nSteady-state inertia of the loaded screw\nkm\nThermal conductivity\nKf\nTorque constant related to the field\nK\u2081\nTorque constant related to the motor\nLa\nArmature inductance\nN\nGear ratio\nP\nActive power\nQ\nReactive power\nR\nElectrical resistance\nS\nApparent power\nRa\nArmature resistance\nTb\nExtruder barrel set temperature\nTL\nLoad torque on the screw\nTmelt\nMelt temperature\nTm\nMotor torque\nx(t)\nA time varying position\nx(t)\nA time varying position velocity\nV\nLine voltage\nVa\nArmature voltage\nVbx\nThe component of the barrel velocity in the transverse direction\n19\nV\u0192\nField voltage\nVj\nThe resultant relative velocity\nTerm\nDefinition\nCOS\nThe displacement power factor\nPm\nMelt density\n\u03b7\nMelt viscosity\nWactual\nActual screw speed\nW set\n\u03bb\nC\u2081\n\u03a9\nSet screw speed\nTemperature of the solid bed\nRate of melting\nReferences\n[1] British plastics federation, about the british plastics industry, Available at: http://www.bpf.co.uk/Industry/Default.aspx., Last viewed: 27 of\nFebruary 2019.\n[2] Plasticseurope: Plastics the facts 2017 an analysis of european plasticsproduction, demand and waste data, Available at:\nhttps://www.plasticseurope.org/application/files/5715/1717/4180/Plastics the facts2017 FINAL for website ne page.pdf, Last viewed: 27 of\nFebruary 2019.\n[3] Market outlook:\nAvailable at:\nPolymers world turned on its head (icis chemical business),\nhttp://www.icis.com/resources/news/2013/10/25/9718765/market-outlook-polymers-world-turned-on-its-head/, Last viewed: 31 of\nAugust 2016.\n[4] Available at: http://www.plasticseurope.org/, Last viewed: 10 of August 2012.\n[5] J. Vlachopoulos, D. Strutt, Overview: Polymer processing, Materials Science and Technology 19 (2003) 1161\u20131169.\n[6] European best practice guide: Low energy plastics processing, Available at: http://www.tangram.co.uk/TI-Energy-LowLast viewed: 27 of\nFebruary 2019.\n[7] K. L. O. C. S. L. G. Kaur, K. Uisan, Recent trends in green and sustainable chemistry & waste valorisation: Rethinking plastics in a\ncirculareconomy, Current Opinion in Green and Sustainable Chemistry 9 (2018) 30\u201339.\n[8] J. Frankland, Mission (nearly) impossible: Estimating extrusion melt temperature, Plsatics TechnologyAugust Issue.\n[9] M. J. Stevens, J. A. Covas, Extruder principles and operation, 2nd Edition, Chapman & Hall, 1995.\n[10] C. Rauwendaal, Polymer Extrusion, 4th Edition, Hanser, 2001.\n[11] C. Abeykoon, Polymer Extrusion: A Study on Thermal Monitoring Techniques and Melting Issues, Lap Lambert Academic Publish-\ning: Verlag, 2012.\n[12] C. I. Chung, E. M. Mount, D. E. McClelland, Energy efficiency in plasticating screw extrusion, Proceedings of 176th Symposium in Energy\nConservation in Textile and Polymer Processing (1979) 21-36.\n[13] J. Agassant, Polymer Processing, Hanser-Munich, 1991.\n[14] A. N. Wilkinson, A. J. Ryan, Polymer Processing and Structure Development, Springer-Netherlands, 1998.\n[15] T. Osswald, J. Hernandez-Oritz, Polymer Processing, Hanser-Munich, 2006.\n[16] W. H. Darnell, E. A. J. Mol, Solids conveying in extruders, SPE Journal 12 (1956) 20-29.\n[17] B. H. Maddock, A visual analysis of flow and mixing in extruder screws, SPE Journal 15 (1959) 383-389.\n[18] Z. Tadmor, Fundamental of plasticating extrusion. i. a theoretical model for melting, Polymer Engineering and Science 6 (1966) 185\u2013190.\n[19] Z. Tadmor, I. Klein, The formulation of a mathematical model for plasticating screw extrusion, Rubber Chemistry and Technology 42 (1969)\n780-789.\n[20] Z. Tadmor, I. Klein, Engineering principles of plasticating extrusion, 1st Edition, Society of Plastics Engineers, 1970.\n[21] Z. Tadmor, C. Gogos, Principles of polymer processing, Wiley-New York, 1979.\n[22] E. Severs, Rheology of polymers, Van Nostrand Reinhold Inc, 1962.\n[23] C. Abeykoon, P. Perez, A. L. Kelly, he effect of materials' rheology on process energy consumption in polymer extrusion, Polymer Engi-\nneering and Science 60 (6) (2020) 1244-1265.\n[24] Y. Breaux, J.-Y. Charmeau, M. Moguedet, A simple model of throughput and pressure development for single screw, Journal of Materials\nProcessing Technology 209 (2009) 611-618.\n[25] H. S. Rowell, R. D. Finlayson, Screw viscosity pumps, Engineering 114 (1922) 606\u2013607.\n[26] H. S. Rowell, R. D. Finlayson, Screw viscosity pumps, Engineering 126 (1928) 249-387.\n[27] C. Abeykoon, Modelling and control of melt temperature in polymer extrusion, Ph.D. thesis, Queen's University Belfast, United Kingdom\n(2011).\n[28] C. Abeykoon, J. Vera-Sorroche, A. L. Kelly, E. C. Brown, P. D. Coates, J. Deng, K. Li, N. Karnachi, E. Harkin-Jones, M. Price, Process\nefficiency in polymer extrusion: Correlation between the energy demand and melt thermal stability, Applied Energy 135 (2014) 560-571.\n[29] G. A. Kruder, R. E. Nunn, Optimizing energy utilization in extrusion processing, SPE ANTEC Technical Papers (1981) 648-652.\n20\n\n\n[30] J. M. McKelvey, Energy utilisation in extrusion, Advances in Polymer Technology 3 (1983) 205-211.\n[31] T. H. Strauch, F. Gliese, G. Menges, Energy usage and conservation in extrusion plants, Plastics and Rubber Institute (Polymer Extrusion\nIII Conference), London, England 1 (1985) 2.1-2.10.\n[32] J. Anderson, C. Rauwendaal, M. del P. Noriega, Troubleshooting extrusion problems, SPE ANTEC Technical Papers 1 (1997) 127\u2013134.\n[33] H. Falkner, Energy cost savings in motive power, Water and Environment Journal 11 (1997) 225-227.\n[34] D. V. Rosato, N. R. Schott, D. V. Rosato, M. G. Rosato, Plastics engineering, manufacturing & data handbook: Fundamental and processes,\nVol. 2, Kluwer Academic Publishers, 2001.\n[35] T. W. Womer, W. S. Smith, R. P. Wheeler, Comparison of two different cooling methods for extrusion processes, SPE ANTEC Technical\nPapers (2006) 796-801.\n[36] About the industry, Available at: http://www.bpf.co.uk, Last viewed: 27 of February 2012.\n[37] K. M. Cantor, Analyzing extruder energy consumption, SPE ANTEC Technical Papers 2 (2010) 1300-1306.\n[38] R. V. Heur, M. Verheijen, Power quality & utilization guide: Plastics industry-energy efficiency, Available at: http: //www.leonardo-\nenergy.org/files/root/pdf/Application/Guide/Plastics.pdf, Last viewed: 15 of August 2012.\n[39] J. Deng, K. Li, E. Harkin-Jones, M. Price, N. Karnachi, A. Kelly, J. Vera-Sorroche, P. Coates, E. Brown, M. Fei, Energy monitoring and\nquality control of a single screw extruder, Applied Energy 113 (2014) 1775\u20131785.\n[40] S. Barlow, Reducing electrical energy costs for extrusion processes, SPE ANTEC Technical Papers (2009) 1157-1162.\n[41] R. Kent, Energy management in plastics processing - part 10: Motors and drives, Available at: http://www.tangram.co.uk, Last viewed: 22\nof february 2010.\n[42] B. Drury, The control techniques drives and controls handbook, The Institution of Electrical Engineers, London-UK, 2010, pp 308-312.\n[43] J. V. DeJuneas, How to reduce electrical energy in the extrusion process, TAPPI Pap Synth Conf (1975) 49\u201350.\n[44] J. M. Dealy, Energy conservation in plastics processing: A review, Polymer Engineering and Science 29 (1982) 528-535.\n[45] C. I. Chung, Energy efficinecy in single screw extrusion, SPE ANTEC Technical Papers (1983) 919-920.\n[46] R. Rasid, A. K. Wood, Effect of process variables on melt tempearture profiles in extrusion process using single screw plastics extruder,\nPlast. Rubb. Comp. 32 (5) (2003) 193\u2013198.\n[47] E. C. Brown, A. L. Kelly, P. D. Coates, Melt temperature homogeneity in single screw extrusion: effect of material type and screw geometry,\nSPE ANTEC Technical Papers (2004) 183\u2013187.\n[48] A. L. Kelly, E. C. Brown, P. D. Coates, The effect of screw geometry on melt temperature profile in single screw extrusion, Polymer\nEngineering and Science 46 (12) (2006) 1706-1714.\n[49] A. L. Kelly, J. V. Sorroche, E. C. Brown, P. D. Coates, Improving thermal efficiency of single screw extrusion, SPE ANTEC Technical\nPapers 2 (2012) 1080\u20131087.\n[50] J. V. Sorroche, A. L. Kelly, E. C. Brown, P. D. Coates, N. Karnachi, E. Harkin-Jones, K. Li, J. Deng, Thermal optimisation of polymer\nextrusion using in-process monitoring techniques, Applied Thermal Engineering 53 (2) (2012) 405\u2013413.\n[51] C. Abeykoon, M. McAfee, S. Thompson, K. Li, A. L. Kelly, E. C. Brown, Investigation of torque fluctuations in extrusion through monitoring\nof motor variables, Proceedings of 26th PPS Annual Europe/Africa regional meeting, Larnaca, Cyprus (2009) Paper No: 22-0.\n[52] C. Abeykoon, J. Vera-Sorroche, A. L. Kelly, E. C. Brown, P. D. Coates, K. B. Howell, J. Deng, K. Li, N. Karnachi, E. Harkin-Jones,\nM. Price, Investigation of the process energy demand in polymer extrusion: A review and an experimental study, Applied Energy 136 (2014)\n726-737.\n[53] C. Abeykoon, A. L. Kelly, E. C. Brown, P. D. Coates, The effect of materials, process settings and screw geometry on energy consumption\nand melt temperature in single screw extrusion, Applied Energy (2016) 880-894.\n[54] R. S. Mallouk, J. M. Mckelvey, Power requirements of melt extruders, Industrial and engineering chemistry 45 (1953) 987\u2013989.\n[55] K. Wilczynski, A computer model for single-screw plasticating extrusion, Polymer-Plastics Technology and Engineering 35 (3) (1996)\n449-477.\n[56] E. Lai, D. W. Yu, Modeling of the plasticating process in a single screw extruder: A fast track approach, Polymer Engineering and Science\n40 (5) (2000) 1074-1084.\n[57] C. Abeykoon, M. McAfee, K. Li, P. J. Martin, J. Deng, A. L. Kelly, Modelling the effects of operating conditions on motor power consump-\ntion in single screw extrusion, Proceedings of LSMS 2010 and ICSEE 2010 6329 (2) (2010) 9-20.\n[58] T. Wildi, Electrical machines, drives, and power systems, 4th Edition, Prentice Hall, 2000.\n[59] S. J. Chapman, Electrical machinery fundamentals, McGraw-Hill, 2005, pp 261-264, 524-527.\n[60] Motor efficiency depends upon power factorAvailable at: http://www.actel.com/documents/Motor_PowerFactor_WP.pdf.\n[61] L. Halsz, Control methods in polymer processing, Elsevier, 1993.\n[62] C. Abeykoon, M. McAfee, K. Li, P. J. Martin, A. L. Kelly, The inferential monitoring of screw load torque to predict process fluctuations in\npolymer extrusion, J. Mater. Process. Technol. 211 (12) (2011) 1907-1918.\n[63] R. M. Green, A comparison of AC and DC extruder drives, IEEE Industry Applications Society Annual Meeting 2 (1989) 1663\u20131666.\n[64] N. Mulakken, Taufik, M. Anwari, Power quality measurements of a dc motor drive, International Confernce on Power and Energy Systems,\nLangkawi, malaysia (2008) 98\u2013101.\n[65] S.-H. Kim, Electric Motor Control: DC, AC, and BLDC Motors, Elsevier, 2017.\n[66] F. Eickelberg, Continuous vacuum metallizing: it's more than tinsel . . . much more, Modern Plastics 54 (1977) 42-45.\n[67] C. S. P. Kumar, S. P. Sabberwal, A. K. Mukharji, Power factor measurement and correction techniques, Electric Power Systems Research\n32 (2) (1994) 141-43.\n[68] R. A. Sfeir, Measured savings of DC to AC drive retrofit in plastic extrusion, Proceedings of the 13th Industrial Energy Technology Confer-\nence, New Orleans, USA.\n[69] Extruder line updated with ac vector drives, Available at: http://www.yaskawa.com/site/dmdrive.nsf, Last viewed: 10 of August 2012.\n[70] D. C. Lounsbury, A. G. Karafilidis, Optimizing extruder drive selection: A review of the factors involved, SPE ANTEC Technical Papers\n(1995) 76-79.\n[71] A. A. not stated, Induction heating applications - the processes, the equipment, the benefits. kha provide full details, Internet: http://www.efd-\n21\ninduction.com/ (Year not stated, last access date not stated) Pages not stated.\n[72] B. F. T. Gregory E. Tremblay, Dan Sowle, Investigation of injection moulding performance using induction barrel heating. kha provide full\ndetails, Journal not stated (Year not stated) Pages not stated.\n[73] T. Deligio, Screws and barrels the focus of energy-savings efforts, Modern Plastic Worldwide (2008) 44\u201345.\n[74] J. M. McKelvey, Theory of adiabatic extruder operation, Industrial & Engineering Chemistry 46 (4) (1954) 660\u2013664.\n[75] W. T. Rice, Conservation of energy and raw materials by utilization of gear pumps in conjunction with extruder processes, Energy Conference\n& Exhibition, New Orleans, LA, USAASME Publication 80-Pet-2.\n[76] E. L. Steward, Control of melt temperature on single screw extruders, SPE ANTEC Technical Papers 1, paper No: 813.\n[77] K. Bhattacharjee, Energy conservation opportunities in industrial waste heat recovery systems, Energy Engineering 107 (2010) 7-13.\n[78] J. J. Baron, C. E. McChesney, S. M. Sinker, Extrusion blow moldable polyethylene terephthalate, SPE ANTEC Technical Papers (1981)\n713-715.\n[79] Energy saving pet, Modern Plastics 58 (1981) 72.\n[80] Pp that cuts users' energy costs, Modern Plastics 58 (1981) 14.\n[81] R. D. Deanin, Decreasing energy requirements in plastics processing, Polymer-Plastics Technology and Engineering 14 (1980) 37\u201345.\n[82] B. I. D. V. H. P. K. H. d. W. Y. Yang, R. Boom, Recycling of composite materials, Chemical Engineering and Processing 51 (2012) 53\u201368.\n[83] J. J. M Wollschlaeger, T Sauter, The future of industrial communication, IEEE Industrial Electronics magazine (2017) 17-27.\n[84] E. K. S. N. RY Zhong, X Xu, Intelligent manufacturing in the context of industry 4.0: A review, Engineering 3 (2017) 616\u2013630.\n[85] M. Loeffler, A. Tschiesner, The internet of things and the future of manufacturing, Web interview https://www.mckinsey.com/business-\nfunctions/digital-mckinsey/our-insights/the-internet-of-things-and-the-future-of-manufacturing.\n[86] All things alexa, Amazon Webpage https://www.amazon.com/Amazon-Echo-And-Alexa-Devices/b?node=9818047011.\n[87] M. K. M. H. K.-S. K. SMR ISLAM, D KWAK, The internet of things for health care: A comprehensive survey, IEEE Access 3 (2015)\n678-708.\n[88] G. M. F Shrouf, J Ordieres, Smart factories in industry 4.0: A review of the concept and of energy management approached in production\nbased on the internet of things paradigm, 2014 IEEE International Conference on Industrial Engineering and Engineering Management,\n9-12 December, Bandar Sunway, Malaysia.\n[89] G. S. T. Stock, Opportunities of sustainable manufacturing in industry 4.0, Procedia CIRP 40 (2016) 536 - 54.\n[90] A. F. S. Y. L. JH Ang, C Goh, Energy-efficient through-life smart design, manufacturing and operation of ships in an industry 4.0 environ-\nment, Energies 10 (2017) doi:10.3390/en10050610.\n[91] J. P. Ibar, Control of polymer properties by melt vibration technology: A review, Polym Eng. Sci. 38 (1) (1998) 1-20.\n[92] J. Y. Chen, Y. Z. Chen, H. L. Li, S. Y. Lai, J. D. Jow, Physical and chemical effects of ultrasound vibration on polymer melt in extrusion.,\nUltrason Sonochem 17 (2010) 66-71.\n[93] Y. C. Zhang, H. L. Li, Functionalization of high density polyethylene with maleic anhydride in the melt state through ultrasonic initiation.,\nPolym Eng. Sci. 43 (4) (2003) 774-782.\n[94] Y. Z. Chen, H. L. Li, Effect of ultrasound on the viscoelasticity and rheology of polystyrene extruded through a slit die., J. Appl. Polym. Sci.\n100 (4) (2006) 2907-2911.\n[95] I. Isayev, S. Mandelbaum, Effect of ultrasonic waves on foam extrusion., Polym. Eng. Sci. 31 (14) (1991) 1051\u20131056.\n[96] Y. R. Cao, H. L. Li, Influence of ultrasound on the processing and structure of polypropylene during extrusion., Polym. Eng. Sci. 42 (7)\n(2002) 1534-1540.\n[97] X. Q. Liu, K. Li, M. McAfee, J. Deng, 'soft-sensor' for real-time monitoring of melt viscosity in polymer extrusion process., Proc. IEEE\nConference on Decision and Control, GA, USA (2010) 3469-3474.\n[98] Y. G. Son, Determination of shear viscosity and shear rate from pressure drop and flow rate relationship in a rectangular channel., Polymer\n48 (2007) 632-637.\n[99] M. McAfee, G. McNally, Real-time measurement of melt viscosity in single-screw extrusion., T I Meas. Control 28 (5) (2006) 481\u201397.\n[100] B. K. Nguyen, G. McNally, A. Clarke, Automatic extruder for recycled polymers with ultrasound and temperature control system., Proc.\nIMechE Part E: J Process Mechanical Engineering (2015) DOI: 10.1177/0954408914552335.\n[101] C. Abeykoon, K. Li, M. McAfee, P. J. Martin, Q. Niu, A. L. Kelly, J. Deng, A new model based approach for the prediction and optimisation\nof thermal homogeneity in single screw extrusion, Control Eng. Pract. 19 (8) (2011) 862-874.\n[102] C. Abeykoon, A novel soft sensor for real-time monitoring of die melt temperature profile in polymer extrusion, IEEE Trans. Ind. Electron.\n61 (12) (2014) 7113-7123.\n[103] C. Abeykoon, A novel model-based controller for polymer extrusion, IEEE Trans. Fuzzy Syst. 22 (6) (2014) 1413\u20131430.\n[104] Dassault systemes, simulia portfolio., http://www.3ds.com/products-services/simulia/products/.\n[105] Ansys products., http://www.ansys.com/en-GB/Products.\n[106] Comsol, multiphysics software product suite., https://uk.comsol.com/products.\n[107] Msc software, products., http://www.mscsoftware.com/products.\n[108] Esi, software solutions., https://www.esi-group.com/software-solutions.\n[109] Livermore software technology corporation, Is-dyna., http://www.lstc.com/products/ls-dyna.\n[110] O. C. Zienkiewicz, R. L. Taylor, J. Z. Zhu, The Finite Element Method: Its Basis and Fundamentals, 7th Edition, Butterworth-Heinemann,\nOxford, 2013.\n[111] D. J. Benson, Computational methods in lagrangian and eulerian hydrocodes., Computer Methods in Applied Mechanics and Engineering\n99 (2-3) (1992) 235-394.\n[112] R. F. Stellingwerf, C. A. Wingate, Impact modelling with sph., Memorie della Societa Astronomia Italiana 65 (4) (1994) 1117\u20131128.\n[113] G. R. Liu, M. B. Liu, Smoothed Particle Hydrodynamics., World Scientific Publishing Company, Singapore, 2003.\n[114] Y. Mu, G. Zhao, A. Chen, X. Wu, Numerical investigation of the thermally and flow induced crystallization behavior of semi-crystalline\npolymers by using finite elementfinite difference method., Computers & Chemical Engineering 46 (2012) 190\u2013204.\n[115] Y. Mu, G. Zhao, A. Chen, X. Wu, Modeling and simulation of polymer melts flow in the extrusion process of plastic profile with metal\n22\ninsert., The International Journal of Advanced Manufacturing Technology 67 (1) (2013) 629\u2013646.\n[116] Y. Mu, G. Zhao, X. Wu, L. Hang, H. Chu, Continuous modeling and simulation of flow-swell-crystallization behaviors of viscoelastic\npolymer melts in the hollow profile extrusion process., Applied Mathematical Modelling 39 (3-4) (2015) 1352-1368.\n[117] C. Abeykoon, P. J. Martin, K. Li, A. L. Kelly, Dynamic modelling of die melt temperature profile in polymer extrusion: Effects of process\nsettings, screw geometry and material., Applied Mathematical Modelling 38 (4) (2014) 1224\u20131236.\n[118] F. Zairi, Should Havelwith TwoDots, B. Aour, J. Gloaguen, M. Nat-Abdelaziz, J. Lefebvre, Numerical modelling of elasticviscoplastic equal\nchannel angular extrusion process of a polymer., Computational Materials Science 38 (1) (2006) 202-212.\n[119] J. Dryden, Stress concentration in polymer extrusion dies., Acta Mechanica 219 (3) (2011) 269\u2013279.\n[120] J. Liker, The Toyota Way: 14 Management Principles from the World's Greatest Manufacturer., McGraw-Hill, New York, 2003.\n[121] M. George, J. Maxey, D. Rowlands, M. Upton, The Lean Six Sigma Pocket Toolbook: A Quick Reference Guide to Nearly 100 Tools for\nImproving Quality and Speed., McGraw-Hill, New York, 2003.\n[122] M. Wagner, G. Montague, M. Tham, Neural networks for steady state modelling of an extruder., Artificial Intelligence in Engineering 11 (4)\n(1997) 375-382.\n[123] B. McKay, B. Lennox, M. Willis, G. Barton, G. Montague, Extruder modelling: a comparison of two paradigms., Control '96, UKACC\nInternational Conference on (Conf. Publ. No. 427), September 1996, DOI 10.1049/cp:19960643 2 (1996) 734\u2013739.\n[124] Z.-L. Chen, P.-Y. Chao, S.-H. Chiu, Proposal of an empirical viscosity model for quality control in the polymer extrusion process., Polymer\nTesting 22 (5) (2003) 601\u2013607.\n[125] M. McAfee, Enhancing process insight in polymer extrusion by grey box modelling., Transactions of the Institute of Measurement and\nControl 29 (5) (2007) 467-488.\n[126] M. McAfee, S. Thompson, A novel approach to dynamic modelling of polymer extrusion for improved process control., Proceedings of the\nInstitution of Mechanical Engineers, Part I: Journal of Systems and Control Engineering 221 (4) (2007) 617-628.\n[127] X. Liu, K. Li, M. McAfee, B. K. Nguyen, G. M. McNally, Dynamic gray-box modeling for on-line monitoring of polymer extrusion\nviscosity., Polymer Engineering & Science 52 (6) (2012) 1332\u20131341.\n[128] J. Deng, K. Li, E. Harkin-Jones, M. Price, M. Fei, A. Kelly, J. Vera-Sorroche, P. Coates, E. Brown, Low-cost process monitoring for polymer\nextrusion., Transactions of the Institute of Measurement and Control 36 (3) (2014) 382-390.\n[129] E. G. Carrano, D. G. Coelho, A. Gaspar-Cunha, E. F. Wanner, R. H. Takahashi, Feedback-control operators for improved pareto-set descrip-\ntion: Application to a polymer extrusion process., Engineering Applications of Artificial Intelligence 38 (2015) 147\u2013167.\n[130] P. G. Drazin, Nonlinear systems., Cambridge University Press, Cambridge, 1992.\n[131] G. McKinley, W. Raiford, R. Brown, R. C. Armstrong, Nonlinear dynamics of viscoelastic flow in axisymmetric abrupt contractions., Journal\nof Fluid Mechanics 223 (3-4) (1991) 411-456.\n[132] M. D. Graham, Wall slip and the nonlinear dynamics of large amplitude oscillatory shear flows., Journal of Rheology 39 (4) (1995) 697-712.\n[133] D. E. Smith, D. A. Tortorelli, C. L. TuckerIII, Optimal design for polymer extrusion. part i: Sensitivity analysis for nonlinear steady-state\nsystems., Computer Methods in Applied Mechanics and Engineering 167 (3-4) (1998) 283\u2013302.\n[134] D. E. Smith, D. A. Tortorelli, C. L. TuckerIII, Optimal design for polymer extrusion. part ii: Sensitivity analysis for weakly-coupled\nnonlinear steady-state systems., Computer Methods in Applied Mechanics and Engineering 167 (3-4) (1998) 303\u2013323.\n[135] A. J. McMillan, A non-linear friction model for self-excited vibrations., Journal of Sound and Vibration 205 (3) (1997) 323\u2013335.\n[136] J. C. Pirkle(Jr.), R. D. Braatz, Instabilities and multiplicities in non-isothermal blown film extrusion including the effects of crystallization.,\nJournal of Process Control 21 (3) (2011) 405-414.\n23\n"}, "expected_output": {"claims": [{"unit": "%", "value": 45, "evidence": ["enhancements to machinery could be identified, where an economic case could be made on the basis of energy cost\nsaving. Some examples are given below.\nIn the late 1970s, Chung et al. [12] found that for a 63.5 mm diameter extruder mechanical energy efficiency of\n62% was typical, and for larger extruders the energy efficiency was lower. In 1981, Kruder and Nunn [29] claimed\nthat energy efficiency of extruders can range from 45%-75%. It was noted that the energy efficiency depended on the\ntransmission mechanism, screw design, product geometry, nature of polymer feedstock and the production rate, while\nthe major energy losses of an extruder occur as a result of the forced cooling process step, and the losses associated\nwith the drive and transmission unit. At low screw speeds, barrel heaters consume a considerably higher portion of\nenergy than at higher speeds, and significant energy savings could be made by running the processes at the highest\npossible power factor. Additionally, this work presented information on energy demand and losses of each individual\ncomponent of an extruder.\nSubsequently through the 1980s, most research into energy efficiency was focussed on the screw efficiency and\nmass flow rate. A reduction in the overall power requirement for an extruder can be achieved through the use of a\ngeared pump at the end of the extruder to increase the mass flow rate (McKelvey [30]). In 1985, Strauch et al. [31]\ncarried out an energy consumption study on a 63.5 mm diameter single screw extruder, and observed that most of the\nenergy was consumed by the mechanical parts, with less significant levels of consumption in process heating. The\nenergy conversion was then assessed and it was found that heating the water in the cooling system accounted for more\nthan half of the energy supplied.\nDuring the late 1980s and 1990s, the manufacturing sector was making many changes, with a view to improving\nproductivity and quality. Driven by the advances made in Japanese manufacturing, the main focus during that time\nwas on management methods, such as Total Quality Management, LEAN, and Six Sigma. These efforts initially\naddressed cost and time issues, where the biggest economic benefits were to be found. Latterly, interest in energy\nefficiency began to be seen as not only cost reduction opportunity but also as an environmental imperative.\nIn the context of power consumption in the extruder, in 1997 Anderson et al. [32] recognised that for the processing\nof most plastics, from room temperature, the specific energy consumption (SEC) of the extruder motor should be in\nthe\nrange of 0.0822 to 0.1644 kW.hr/kg.\nAt around the same time, a study by Falkner in 1997 [33] showed that motor operations accounted for over 65%\nof the 1994 UK industrial electricity usage. Asserting that more than 10% of this energy could be attributed to\ninefficiency, Falkner argued that this represented a loss of about \u00a30.5 billion to the annual UK economy. These values\naccounted for motor energy utilisation across multiple industrial sectors, but it should be recognised that the electric\nmotors in plastics industry processing machines are a major power consumers.\nA more detailed study by Rosato et al. in 2001 [34] observed that energy losses of between 3 and 20% can arise in\nthe transmissions and control systems. Despite this, a conclusion was made that because plastics have lower specific\nenergy requirements compared with most conventional raw materials, they are still highly competitive.\nFive years later, Womer et al. [35] considered the energy efficiency of extruder cooling. The results demonstrated\nthat water cooling systems consume more energy compared with air cooling, irrespective of the particular plastic being\nprocessed. As a result, a recommendation was made to use air only cooling unless extensive cooling was expressly\nrequired.\nIn 2010 [36], the plastics industry was recognised to be one of the major UK industries with a similar trend\napplying globally. On that basis any improvement in process energy efficiency would lead to a considerable reduction\nin global energy requirement. Also in 2010, Cantor [37] presented measurements of SEC, where the impact of the\nmotor and of each individual heater zone, with respect to the overall specific energy consumption, was separately\nrecorded. It was observed that the heaters account for over 95% of the supplied energy. In a slightly later study\nby Heur and Verheijen [38], the authors studied differences from one plant to another, and recommended the use of\nfrequency controllers to enable more precise process control.\nThe earliest mention of an Industry 4.0 implementation to energy efficiency control was by Jing et al. (2014) [39]\nwhich proposed the use of real-time monitoring. The rationale was to render unnecessary the installation of power\nmeters or the development of data-driven models. A fuzzy logic controller controlled the high melt quality in a single\nscrew extruder, and was shown to be a cheaper alternative to using a gear pump. This also paved the way for achieving\ngreater extruder energy efficiency by optimising the temperature settings.\nA number of other works [35, 40, 41, 42, 43, 44, 45] consider the drive motor efficiency compared with other\ndevices. The conclusion to be drawn is that the drive motor should be the primary design consideration for process\n6\n", "In 1981, Kruder and Nunn [29] claimed that energy efficiency of extruders can range from 45%-75%.", "Chung et al. [12] found that for a 63.5 mm diameter extruder mechanical energy efficiency of 62% was typical, and for larger extruders the energy efficiency was lower.", "Kruder and Nunn [29] claimed that energy efficiency of extruders can range from 45%-75%."]}, {"unit": "%", "value": 75, "evidence": ["enhancements to machinery could be identified, where an economic case could be made on the basis of energy cost\nsaving. Some examples are given below.\nIn the late 1970s, Chung et al. [12] found that for a 63.5 mm diameter extruder mechanical energy efficiency of\n62% was typical, and for larger extruders the energy efficiency was lower. In 1981, Kruder and Nunn [29] claimed\nthat energy efficiency of extruders can range from 45%-75%. It was noted that the energy efficiency depended on the\ntransmission mechanism, screw design, product geometry, nature of polymer feedstock and the production rate, while\nthe major energy losses of an extruder occur as a result of the forced cooling process step, and the losses associated\nwith the drive and transmission unit. At low screw speeds, barrel heaters consume a considerably higher portion of\nenergy than at higher speeds, and significant energy savings could be made by running the processes at the highest\npossible power factor. Additionally, this work presented information on energy demand and losses of each individual\ncomponent of an extruder.\nSubsequently through the 1980s, most research into energy efficiency was focussed on the screw efficiency and\nmass flow rate. A reduction in the overall power requirement for an extruder can be achieved through the use of a\ngeared pump at the end of the extruder to increase the mass flow rate (McKelvey [30]). In 1985, Strauch et al. [31]\ncarried out an energy consumption study on a 63.5 mm diameter single screw extruder, and observed that most of the\nenergy was consumed by the mechanical parts, with less significant levels of consumption in process heating. The\nenergy conversion was then assessed and it was found that heating the water in the cooling system accounted for more\nthan half of the energy supplied.\nDuring the late 1980s and 1990s, the manufacturing sector was making many changes, with a view to improving\nproductivity and quality. Driven by the advances made in Japanese manufacturing, the main focus during that time\nwas on management methods, such as Total Quality Management, LEAN, and Six Sigma. These efforts initially\naddressed cost and time issues, where the biggest economic benefits were to be found. Latterly, interest in energy\nefficiency began to be seen as not only cost reduction opportunity but also as an environmental imperative.\nIn the context of power consumption in the extruder, in 1997 Anderson et al. [32] recognised that for the processing\nof most plastics, from room temperature, the specific energy consumption (SEC) of the extruder motor should be in\nthe\nrange of 0.0822 to 0.1644 kW.hr/kg.\nAt around the same time, a study by Falkner in 1997 [33] showed that motor operations accounted for over 65%\nof the 1994 UK industrial electricity usage. Asserting that more than 10% of this energy could be attributed to\ninefficiency, Falkner argued that this represented a loss of about \u00a30.5 billion to the annual UK economy. These values\naccounted for motor energy utilisation across multiple industrial sectors, but it should be recognised that the electric\nmotors in plastics industry processing machines are a major power consumers.\nA more detailed study by Rosato et al. in 2001 [34] observed that energy losses of between 3 and 20% can arise in\nthe transmissions and control systems. Despite this, a conclusion was made that because plastics have lower specific\nenergy requirements compared with most conventional raw materials, they are still highly competitive.\nFive years later, Womer et al. [35] considered the energy efficiency of extruder cooling. The results demonstrated\nthat water cooling systems consume more energy compared with air cooling, irrespective of the particular plastic being\nprocessed. As a result, a recommendation was made to use air only cooling unless extensive cooling was expressly\nrequired.\nIn 2010 [36], the plastics industry was recognised to be one of the major UK industries with a similar trend\napplying globally. On that basis any improvement in process energy efficiency would lead to a considerable reduction\nin global energy requirement. Also in 2010, Cantor [37] presented measurements of SEC, where the impact of the\nmotor and of each individual heater zone, with respect to the overall specific energy consumption, was separately\nrecorded. It was observed that the heaters account for over 95% of the supplied energy. In a slightly later study\nby Heur and Verheijen [38], the authors studied differences from one plant to another, and recommended the use of\nfrequency controllers to enable more precise process control.\nThe earliest mention of an Industry 4.0 implementation to energy efficiency control was by Jing et al. (2014) [39]\nwhich proposed the use of real-time monitoring. The rationale was to render unnecessary the installation of power\nmeters or the development of data-driven models. A fuzzy logic controller controlled the high melt quality in a single\nscrew extruder, and was shown to be a cheaper alternative to using a gear pump. This also paved the way for achieving\ngreater extruder energy efficiency by optimising the temperature settings.\nA number of other works [35, 40, 41, 42, 43, 44, 45] consider the drive motor efficiency compared with other\ndevices. The conclusion to be drawn is that the drive motor should be the primary design consideration for process\n6\n", "In 1981, Kruder and Nunn [29] claimed that energy efficiency of extruders can range from 45%-75%.", "Chung et al. [12] found that for a 63.5 mm diameter extruder mechanical energy efficiency of 62% was typical, and for larger extruders the energy efficiency was lower.", "Kruder and Nunn [29] claimed that energy efficiency of extruders can range from 45%-75%."]}, {"unit": "%", "value": 62, "evidence": ["enhancements to machinery could be identified, where an economic case could be made on the basis of energy cost\nsaving. Some examples are given below.\nIn the late 1970s, Chung et al. [12] found that for a 63.5 mm diameter extruder mechanical energy efficiency of\n62% was typical, and for larger extruders the energy efficiency was lower. In 1981, Kruder and Nunn [29] claimed\nthat energy efficiency of extruders can range from 45%-75%. It was noted that the energy efficiency depended on the\ntransmission mechanism, screw design, product geometry, nature of polymer feedstock and the production rate, while\nthe major energy losses of an extruder occur as a result of the forced cooling process step, and the losses associated\nwith the drive and transmission unit. At low screw speeds, barrel heaters consume a considerably higher portion of\nenergy than at higher speeds, and significant energy savings could be made by running the processes at the highest\npossible power factor. Additionally, this work presented information on energy demand and losses of each individual\ncomponent of an extruder.\nSubsequently through the 1980s, most research into energy efficiency was focussed on the screw efficiency and\nmass flow rate. A reduction in the overall power requirement for an extruder can be achieved through the use of a\ngeared pump at the end of the extruder to increase the mass flow rate (McKelvey [30]). In 1985, Strauch et al. [31]\ncarried out an energy consumption study on a 63.5 mm diameter single screw extruder, and observed that most of the\nenergy was consumed by the mechanical parts, with less significant levels of consumption in process heating. The\nenergy conversion was then assessed and it was found that heating the water in the cooling system accounted for more\nthan half of the energy supplied.\nDuring the late 1980s and 1990s, the manufacturing sector was making many changes, with a view to improving\nproductivity and quality. Driven by the advances made in Japanese manufacturing, the main focus during that time\nwas on management methods, such as Total Quality Management, LEAN, and Six Sigma. These efforts initially\naddressed cost and time issues, where the biggest economic benefits were to be found. Latterly, interest in energy\nefficiency began to be seen as not only cost reduction opportunity but also as an environmental imperative.\nIn the context of power consumption in the extruder, in 1997 Anderson et al. [32] recognised that for the processing\nof most plastics, from room temperature, the specific energy consumption (SEC) of the extruder motor should be in\nthe\nrange of 0.0822 to 0.1644 kW.hr/kg.\nAt around the same time, a study by Falkner in 1997 [33] showed that motor operations accounted for over 65%\nof the 1994 UK industrial electricity usage. Asserting that more than 10% of this energy could be attributed to\ninefficiency, Falkner argued that this represented a loss of about \u00a30.5 billion to the annual UK economy. These values\naccounted for motor energy utilisation across multiple industrial sectors, but it should be recognised that the electric\nmotors in plastics industry processing machines are a major power consumers.\nA more detailed study by Rosato et al. in 2001 [34] observed that energy losses of between 3 and 20% can arise in\nthe transmissions and control systems. Despite this, a conclusion was made that because plastics have lower specific\nenergy requirements compared with most conventional raw materials, they are still highly competitive.\nFive years later, Womer et al. [35] considered the energy efficiency of extruder cooling. The results demonstrated\nthat water cooling systems consume more energy compared with air cooling, irrespective of the particular plastic being\nprocessed. As a result, a recommendation was made to use air only cooling unless extensive cooling was expressly\nrequired.\nIn 2010 [36], the plastics industry was recognised to be one of the major UK industries with a similar trend\napplying globally. On that basis any improvement in process energy efficiency would lead to a considerable reduction\nin global energy requirement. Also in 2010, Cantor [37] presented measurements of SEC, where the impact of the\nmotor and of each individual heater zone, with respect to the overall specific energy consumption, was separately\nrecorded. It was observed that the heaters account for over 95% of the supplied energy. In a slightly later study\nby Heur and Verheijen [38], the authors studied differences from one plant to another, and recommended the use of\nfrequency controllers to enable more precise process control.\nThe earliest mention of an Industry 4.0 implementation to energy efficiency control was by Jing et al. (2014) [39]\nwhich proposed the use of real-time monitoring. The rationale was to render unnecessary the installation of power\nmeters or the development of data-driven models. A fuzzy logic controller controlled the high melt quality in a single\nscrew extruder, and was shown to be a cheaper alternative to using a gear pump. This also paved the way for achieving\ngreater extruder energy efficiency by optimising the temperature settings.\nA number of other works [35, 40, 41, 42, 43, 44, 45] consider the drive motor efficiency compared with other\ndevices. The conclusion to be drawn is that the drive motor should be the primary design consideration for process\n6\n"]}]}, "metadata": {"product_category": "Machinery & equipment", "request_id": "req_2ced6848ba546c28"}}
{"id": "30b93980-f568-4cb3-aa6f-b4298583b82e", "input": {"query": "What is the motor-specific energy consumption for extruders in kWh per kg of polymer produced?", "source_url": "https://ajmcmillan.co.uk/AcademicPublications/EnergyInExtrusion_AcceptedVersion.pdf", "document_text": "Energy efficiency in extrusion-related polymer processing: a review of state of\nthe art and potential efficiency improvements\nChamil Abeykoon\u00aa,*, Alison McMillan, Bao Kha Nguyen\n\"North West Composites Centre and Aerospace Research Institute, Department of Materials, Faculty of Science and Engineering, University of\nManchester, Oxford Road, Manchester, M13 9PL, UK\nb Faculty of Arts, Science and Technology, Wrexham Glyndwr University, Wrexham, LL11 2AW, UK\n\"School of Engineering and Informatics, University of Sussex, Brighton, BN1 9QT, UK\nAbstract\nEnergy saving and industrial pollution have become increasingly important issues, therefore the identification and\nadoption of more energy efficient machines and industrial processes are now industrial priorities, and worthy topics\nfor further development through academic research. Polymeric materials are a major raw material, finding widespread\napplication to a range of current industrial machine components as well as multiple products and packaging found\nin our daily life. Polymer extrusion serves as a particular example of polymer processing techniques, representative\nof others in as much as there are analogous intermediate stages in the processing. Processing techniques which re-\nquire such intermediate stages include the manufacture of blown film, blow moulding, thermo-forming, and injection\nmoulding. Hence, the study of polymer extrusion is a representative paradigm for a wider range of processing tech-\nniques. Since polymer processing is an energy intensive process and accounts for a huge share (maybe more than 1/3)\nof the materials processing sector, any improvement to the process would contribute significantly to global energy\nsavings. This work presents a review of studies, which focus on, or appertain to, the energy consumption of extrusion\nrelated polymer processing applications. Typical energy demand and losses during processing are considered, and\npossible approaches for improving the process energy efficiency while maintaining the required end product quality\nare considered. Overall, this work provides a detailed discussion about how and where energy is utilized; how, where\nand why energy losses occur; and sets out approaches for optimizing the process energy efficiency.\nKeywords:\nEnergy consumption, Energy losses, Energy savings, Polymer extrusion, Process monitoring, Process control,\nMaterials processing, Energy efficiency, Industry 4.0, Circular economy, Dynamical systems\n1. Introduction\n1.1. Market demand for polymers\nAs the number of applications for polymer materials in high volume manufacturing sectors, such as packaging, con-\ntinues to grow, it is timely to consider manufacturing process optimisation from the energy efficiency point of view.\nAt the present time, the increasing adoption of thermoplastics for use in high performance component applications,\nsuch as automotive and aerospace, has meant that product quality has been the prime focus of process optimisation.\nAs the manufacturing processes associated with polymer processing have become more mature, there has been a cor-\nrespondingly greater utilisation of in-line sensors, and adoption of Industry 4.0 protocols. This has enabled greater\nunderstanding of the material performance under processing temperatures and pressures, thereby providing the neces-\nsary input data needed for high fidelity computational modelling. In the field of computational optimisation, there have\nbeen signficant advances in algorithms development, with the result that a much bigger class of multi-variable and\n*Corresponding author. +441613062540\nEmail address: chamil. abeykoon@manchester.ac.uk (Chamil Abeykoon)\nPreprint submitted to Elsevier\nMay 20, 2021\nmulti-objective problems can be addressed. As a result, the possibility to broaden the scope of process optimisation\ncan now be grasped.\nClearly, for both high volume and high performance applications, energy efficient manufacture is a desirable\ngoal, which not only leads to reduced manufacturing costs but also addresses National and International energy and\nCO2 reduction targets. As a result, scrutiny of the energy required in an energy intensive manufacturing process\nsuch as the extrusion process is driven by both business and environmental imperatives. The payment of energy\nbills for unnecessary usage reduces profit margins and hence increases the end product/service prices for customers.\nMeanwhile, because CO2 emissions are now very clearly understood to be detrimental to the environment, energy\nusage will be increasingly subject to disincentives such as high fuel commodity pricing and taxation.\nThe scale of the plastics industry is internationally huge, and expanding. For example, in 2015 in the UK [1], there\nwere of the order of 6,200 plastics companies, employing nearly 170,000 people, and with a combined annual sales\nturnover of over \u00a323.5 bn, of which one third represented exports. According to the reports of PlasticsEurope [2], by\nthe year 2016 the European plastics industry comprised of more than 60,000 companies, employing more than 1.5\nmillion people, and with total sales exceeding 350 EUR bn. Globally, plastics production has grown from 204 to 335\nmillion tonnes between 2002 and 2016. The statistics presented in Figures 1 and 2 illustrate this growing demand.\nMillion tonnes\n350\n300\n250\n200\n150\n100\nPE\nPP\nPVC\nPS-EPS\nABS-SAN\n50\n01\n2005\n2011\n2012\n2013\n2017\n2020\n2025\nFigure 1: Major thermoplastics: World demand distribution, by polymer between years 2005-2025 [3]\n%/year\n6\n2005-2012\n2012-2017\n5\n4\n3\n2\n1\n0\nPE\nPP\nPVC\nPS-EPS\nABS-SAN\nTOTAL WORLD'\nFigure 2: Major thermoplastics: World consumption growth rate, by polymer (2005-2012 and 2012-2017)[3]\nGiven this level is sustained, the level of growth in demand, and the development and accessibility of new poly-\nmer processing capability, it is clear that improvements in process energy efficiency could have a significant impact\non global energy savings [4, 5]. Furthermore, the European Best Practice Guide [6] claims, \u201cPlastics are the material\nfor the 21st century\u201d, explaining that a 3 Megatonne CO2 emission reduction could be achieved in Europe by a 10%\nreduction in the plastics industry energy consumption. With the current capacity of polymers and plastics manufac-\nturing sector, it is one of the largest energy consumers in industrial manufacturing and also a major source of global\nwaste generation. Meantime, the energy savings/optimization in the manufacturing sector is considered as one of the\nmain pillars of modern circular economy concept and both manufactures and consumers have been forced to re-think\nthe current take-make-waste extractive industrial model for reusing materials form end-of-life components/devices,\nwhere polymers/plastics industry is one of the major focuses of this concept [7].\n1.2. The polymer extrusion process\nA polymer \"extruder\u201d machine processes materials by forcing them through a set of processing stages. The\noperation and basic processing stages are described in Figure 3 below. The screw passes material through a cylindrical\nHopper\nBarrel\nBand type\nheaters\nControl unit\nDie\nGear box\nDrive\nmotor\nCooling fans\nScrew\nSolids conveying\nMelting Melt conveying \u00a6\nFigure 3: Operational schematic of a single screw extruder\nbarrel, around which heaters are wrapped, to provide the necessary heat for material melting. In addition to this\nexternally provided heat, a significant amount of heat is generated internally, inside the barrel, as a result of the\nmechanical work of the screw (i.e. the work done against viscous and frictional forces). The feed material absorbs\nheat as it is conveyed along the screw and is expected to be in the fully molten state at the point that the molten\nmaterial is forced into a die to form into the desired shape.\nCurrently, different types of extruders (e.g. single screw, multi screw, and disc/drum types) are available in in-\ndustry, while screws with different geometrical designs are commercially available. Moreover, a number of process\nmonitoring devices are used, to observe process functionality, and for diagnosing possible processing problems.\nFrankland [8], President of Frankland Plastics Consulting, LLC, explains this in detail in his on-line article about\nestimating extrusion melt temperature. The most significant points are that the mechanical energy feed into the drive\nis converted by the screw action on the material to create heat, and thus melting of the polymer. The energy share\nrequired for material conveying is relatively smaller, as is the energy supplied to the barrel heaters. He also lists energy\nlosses and their sources. More details on the polymer extrusion process and its operation can be found in the literature\n[9, 10, 11].\nManufacturing process stability is a key concern, and variation in the material temperature presents a challenge to\nthe end product quality control. For this reason most commercial polymer producers avoid operating their extruders at\nat the higher screw speeds. This is unfortunate since at higher speeds, and therefore at higher workpiece temperatures,\nthere is more potential for process energy efficiency improvement because the material viscosity is reduced and thus\nthe forming forces required are lower. Moreover, this undesirable cost is repeated, since many thermoplastic polymers\nare extruded more than once before their final products are manufactured [12]. Better concatenation of extrusion\nprocess steps would lead to greater energy reduction, by maintaining or controlling the heat in the workpiece during\nprocessing, thereby avoiding the need to re-heat.\n1.2.1. Basic processing mechanisms\nZones within the polymer processing screw can be broadly designated, based on the functional activity taking place\nwithin that zone, see Figure 3. The points of transition between zones are not generally well defined, as they depend\non the processing conditions and the materials.\na. Solids conveying\nIn this zone, the polymer is preheated before passing into the subsequent zones. While flowing along this zone,\nmaterial starts to absorb heat from the barrel heaters, but the mechanical heat generated by frictional and viscous\n3\nmechanisms is dominant in this zone [13, 14, 10, 15]. Generally, the screw channel depth is maintained constant in\norder to provide a constant material feed to the subsequent zones.\nThe first comprehensive theory for the action of solids conveying was developed by Darnell and Mol [16] in the\n1950s and this quantitative description still remains as the widely accepted model for solids conveying in extrusion.\nb. Melting or Plastication\nExperiments for studying the polymer extrusion melting mechanism were first carried out by Maddock and Street\nin 1959 [17]. The melting mechanism proposed by Maddock for single screw extruders still remains as the most\nwidely accepted melting mechanism in polymer extrusion. The Maddock melting mechanism is only a qualitative\ndescription of melting which occurs in single screw extruders. Maddock used a visual inspection method to investigate\nthe melting process by stopping the screw rotation suddenly during the process and 'freezing' the polymer by cooling\nthe barrel and screw rapidly. Later, Tadmor also extended the understanding of melting mechanism of extrusion\nprocesses [18, 19, 20, 21].\nAs the material reaches the \"end\" of the solids conveying zone, it begins to melt, and as such is considered to\nhave entered into the melting zone. As the material becomes soft, further heat will be added to the process by means\nof viscous dissipation of the material (i.e. work done against the viscoelastic nature of the material). Both solid\nand molten polymers co-exist in this zone. The solid bed would comprise both compacted solid polymer abutting\nthe \"trailing flight\u201d, and the melt pool pushing against the \u201cpushing flight\", as shown in Figure 4. As the material\nFlow direction\nSolid/Melt\ninterface\nPushing\nflight\nCirculating\nmelt pool\nExtruder barrel\nTrailing\nSolid bed\nflight\nScrew\nFigure 4: An illustration of the typical arrangement of the solid bed and melt pool inside a screw channel for a single-flighted conventional screw\nproceeds along the screw, the proportion of melt pool to solid bed increases. The screw channel depth is therefore\ndesigned to become smaller, which influences flow rate and mixing; however, the actual screw length at which melting\noccurs depends on a range of parameters such as screw geometry, operating conditions and physical properties of the\npolymer [20].\nAs was claimed by Severs [22], the plastication or melting process has a direct impact on the quality of the material\nproperties of final product, and thus must be carefully controlled. Tadmor, Klein and Gogos [18, 19, 21] proposed an\nequation for calculating the rate of melting, (Q), in a screw channel, and is given by Eq. (1).\nQ2 =\n-\n[Pm \u00d7 Vbx { km (Tb \u2212 Tmelt) +\u014b\n2 {Cp (Tm-Ts) + 1}\n11/2\n(1)\nWhere Pm is the melt density, Vbx is the transverse component of the barrel velocity, km is the thermal conductivity\nof the molten material, Tmelt is the melt temperature, T is the barrel temperature, \u014b is the melt viscosity, V; is the\nresultant relative velocity, Cp is the polymer specific heat capacity, and \u03bb is the temperature of the solid bed.\nThis equation clearly demonstrates that the melting rate can be increased by increasing the screw rotational speed\n[14]; however, for higher speeds, the polymer passes through more quickly, giving less time for temperature stabilisa-\ntion, and thus more variation in melt viscosity [10, 14]. As a result, to ensure controlled plastication, it is necessary to\ncontrol the melting rate, and this in turn depends on the material being proceeded, process set conditions, and nature\nof the processing unit/machine [23].\nc. Melt conveying\nMelt conveying starts as complete melting is achieved. The screw channel depth is constant along the zone and\nis shallower than in the other two zones. During this stage further heating and mixing of the melt takes place as the\npolymer is smeared by the tip of the screw flight against the barrel wall. Material has to be moved towards the die\nwith enough force to overcome the head pressure generated at the die - this is known as the \u201cdie head pressure\".\nMelt output rate from this zone depends on a combination of two main factors: the rate of the rotation of the screw\nand the screw channel pressure gradient [24]. Proper mixing of material is another requirement for the flow through\nthis zone. The melt conveying zone of some of the new screw designs is fitted with efficient mixer units to ensure\ngood mixing performance (e.g. the barrier flighted screw with a Maddock mixer).\nStudies on melt conveying operation of extrusion were reported very much earlier than in the other two zones.\nOne of the initial studies was carried out in 1920s [25, 26], which proposed the calculation of the melt conveying rate\nby considering the melt flow as a laminar fully developed flow. This is still a widely accepted model.\nIn addition to the above mentioned mechanism/theories, several other works have been reported later on improving\nthe understanding of these three main mechanisms and more details can be found in the literature [14, 9, 10, 11, 27].\n2. Energy required for materials processing\nThe assessment of energy requirements is not straight forward. The overall extrusion process can be broken down\ninto smaller activities, but even then, the power demands at each stage depend in a complex way on a large number of\nprocessing parameters. A useful energy flow model was developed by Severs [22], as presented in Figure 5.\nEquipment cooling\nLosses\nit\nCooling\n\u2191\nPolymer\nsolid\n\u2192 Melting\n\u2192 Forming\nSolidification\nPolymer\nproduct\nMotor power (mechanical energy)\nHeating system (thermal energy)\nElectric power\nFigure 5: Typical energy flow diagram for an extrusion process\nThe energy, Eu, used by an extruder for useful work in material melting and forming, [28], is given by Eq. (2):\n=\nEu Ein Elosses\n(2)\nwhere Ein is the energy input to the extruder and Elosses is the energy expended that does not contribute to the extrusion\nprocess. Thus, the energy efficiency can be given by Eq. (3):\nnextruder =\nEin - Elosses\nEin\n\u00d7 100%\n(3)\nIn these equations, the energy inputs (Ein) should be related to the energy consumed by the electrical components\nsuch as drive motor, barrel/die heaters, barrel/motor cooling fans, water pump/s, instrumentation in the control unit,\netc.. The energy losses are always associated with all the components and also occur due to forced cooling and via\nnatural convection and radiation, which can be accounted under Elosses. In general, the drive motor and the barrel and\ndie heaters are the source of the highest energy losses. In typical polymer extrusion processes, recovery of such lost\nenergy is impractical, as this is largely released as heat energy to water or air. More details concerning the energy\nrequired for polymer processing and the thermodynamic efficiency of an extruder have been discussed by the authors\npreviously [28].\n3. Prior art in extruder energy evaluation, monitoring and modelling\n3.1. Energy Consumption studies\nIn considering the energy consumption in any industrial process, the first step is to review the process capability\nof the existing or available plant machinery, and the power consumption. On that basis, potential modifications or\n5\nenhancements to machinery could be identified, where an economic case could be made on the basis of energy cost\nsaving. Some examples are given below.\nIn the late 1970s, Chung et al. [12] found that for a 63.5 mm diameter extruder mechanical energy efficiency of\n62% was typical, and for larger extruders the energy efficiency was lower. In 1981, Kruder and Nunn [29] claimed\nthat energy efficiency of extruders can range from 45%-75%. It was noted that the energy efficiency depended on the\ntransmission mechanism, screw design, product geometry, nature of polymer feedstock and the production rate, while\nthe major energy losses of an extruder occur as a result of the forced cooling process step, and the losses associated\nwith the drive and transmission unit. At low screw speeds, barrel heaters consume a considerably higher portion of\nenergy than at higher speeds, and significant energy savings could be made by running the processes at the highest\npossible power factor. Additionally, this work presented information on energy demand and losses of each individual\ncomponent of an extruder.\nSubsequently through the 1980s, most research into energy efficiency was focussed on the screw efficiency and\nmass flow rate. A reduction in the overall power requirement for an extruder can be achieved through the use of a\ngeared pump at the end of the extruder to increase the mass flow rate (McKelvey [30]). In 1985, Strauch et al. [31]\ncarried out an energy consumption study on a 63.5 mm diameter single screw extruder, and observed that most of the\nenergy was consumed by the mechanical parts, with less significant levels of consumption in process heating. The\nenergy conversion was then assessed and it was found that heating the water in the cooling system accounted for more\nthan half of the energy supplied.\nDuring the late 1980s and 1990s, the manufacturing sector was making many changes, with a view to improving\nproductivity and quality. Driven by the advances made in Japanese manufacturing, the main focus during that time\nwas on management methods, such as Total Quality Management, LEAN, and Six Sigma. These efforts initially\naddressed cost and time issues, where the biggest economic benefits were to be found. Latterly, interest in energy\nefficiency began to be seen as not only cost reduction opportunity but also as an environmental imperative.\nIn the context of power consumption in the extruder, in 1997 Anderson et al. [32] recognised that for the processing\nof most plastics, from room temperature, the specific energy consumption (SEC) of the extruder motor should be in\nthe\nrange of 0.0822 to 0.1644 kW.hr/kg.\nAt around the same time, a study by Falkner in 1997 [33] showed that motor operations accounted for over 65%\nof the 1994 UK industrial electricity usage. Asserting that more than 10% of this energy could be attributed to\ninefficiency, Falkner argued that this represented a loss of about \u00a30.5 billion to the annual UK economy. These values\naccounted for motor energy utilisation across multiple industrial sectors, but it should be recognised that the electric\nmotors in plastics industry processing machines are a major power consumers.\nA more detailed study by Rosato et al. in 2001 [34] observed that energy losses of between 3 and 20% can arise in\nthe transmissions and control systems. Despite this, a conclusion was made that because plastics have lower specific\nenergy requirements compared with most conventional raw materials, they are still highly competitive.\nFive years later, Womer et al. [35] considered the energy efficiency of extruder cooling. The results demonstrated\nthat water cooling systems consume more energy compared with air cooling, irrespective of the particular plastic being\nprocessed. As a result, a recommendation was made to use air only cooling unless extensive cooling was expressly\nrequired.\nIn 2010 [36], the plastics industry was recognised to be one of the major UK industries with a similar trend\napplying globally. On that basis any improvement in process energy efficiency would lead to a considerable reduction\nin global energy requirement. Also in 2010, Cantor [37] presented measurements of SEC, where the impact of the\nmotor and of each individual heater zone, with respect to the overall specific energy consumption, was separately\nrecorded. It was observed that the heaters account for over 95% of the supplied energy. In a slightly later study\nby Heur and Verheijen [38], the authors studied differences from one plant to another, and recommended the use of\nfrequency controllers to enable more precise process control.\nThe earliest mention of an Industry 4.0 implementation to energy efficiency control was by Jing et al. (2014) [39]\nwhich proposed the use of real-time monitoring. The rationale was to render unnecessary the installation of power\nmeters or the development of data-driven models. A fuzzy logic controller controlled the high melt quality in a single\nscrew extruder, and was shown to be a cheaper alternative to using a gear pump. This also paved the way for achieving\ngreater extruder energy efficiency by optimising the temperature settings.\nA number of other works [35, 40, 41, 42, 43, 44, 45] consider the drive motor efficiency compared with other\ndevices. The conclusion to be drawn is that the drive motor should be the primary design consideration for process\n6\nengineering the energy efficiency of the whole extrusion plant.\n3.2. Influence of process set parameters\nIn 2001, Rauwendaal [10] recognised the significance of process settings, and presented an account of a procedure\nto minimise power consumption. A little later, in 2003, Rasid and Wood [46] investigated the influence of individual\nbarrel zone temperatures and found that the solids conveying zone temperature had the greatest influence on overall\npower consumption.\nStudies carried out between 2004 and 2012, [47, 48, 49, 50], examined various process parameters and their\ninfluence on SEC. In addition to noting the effect of material viscosity, variation in energy consumption was also seen\nfor different designs of screw, and there were greater melt temperature fluctuations at higher screw speeds: another\nexample of the ever-present tension between cost and quality. The simultaneous need to achieve both energy efficient\noperation and finished part quality remains a challenge.\nStudies carried out by Abeykoon et al. [51, 28, 52, 53] between 2009 and 2016, focussed on the relationship\nbetween the process energy demands of the motor and barrel heating and melt thermal stability. The effects of the\nsettings for these processes, the screw geometry and choice of material were explored.\n3.3. Modelling\nFollowing a thorough trawl of the published scientific literature, it has become clear that relatively little work has been\nundertaken to model extruder energy consumption.\nThe earliest work in this area was by Mallouk and Mckelvey [54] in 1953, where a mathematical equation was\ndeveloped, based on assumptions of isothermal, Newtonian flow, in a screw channel with constant section. In 1996,\nWilczynski [55] developed a computer model where the five zones of the extruder plus the die were considered\nseparately. Subsequently, in 2000, Lai and Yu [56] also proposed a mathematical model for the calculation of energy\nconsumption based on screw speed, and including viscosity. In Abyekoon et al.'s [57, 52] studies of a single screw\nextruder, the data collected was analysed using static nonlinear polynomial models. The conclusion of the analysis\nwas that choosing energy efficient process settings would also lead to thermal stability.\nObviously, the availability of advanced modelling methods for predicting energy consumption, based on process\nparameters, would enable process operators to select optimum operating conditions. In particular, models which\nincorporate both energy consumption and melt thermal quality would be preferred but the development of such models\nis quite challenging. Melt thermal quality and energy efficiency present opposite behaviours with respect to the\nprocessing speed: the thermal quality deteriorates while the energy efficiency improves. Since the industrial sector\nhas to meet strict environmental regulations to minimize the carbon footprint, any reduction in the energy demands\nfor polymer processing would support future sustainability.\n3.4. General considerations in energy usage\nAccording to basic electricity principles, the typical power consumption of a DC and an AC device (PDC and PAC) is\ngiven by equations (4) and (5), respectively [58, 59],\nPDC = V XI\nPAC = VXIX cos\n(4)\n(5)\nwith I being the supply current, V voltage, and cos & the \"displacement power factor\u201d. From these, the power demand\nof any device in an extrusion plant can be evaluated; however, the energy losses related to each device might vary\nfrom component to component.\nThe power factor is an important consideration in the assessment of the energy usage of an electrical machine or\nprocess, and is defined as either the \"displacement power factor\" which is the cos o in Eq. (5) or the \"true power\nfactor\" which is given by Eq. (6).\nTrue\npower factor\n=\nTrue (or active) power\n(6)\n7\nApparent power\nImpedance\nApparent power (S)\n(units: VA)\nphase angle\n(units: VAR)\nReactive power (Q)\nActive power (P)\n(units: W)\nFigure 6: Power triangle showing the relationship between active, apparent and reactive powers\nThe true power factor lies in the range 0 \u2013 1, for which the running of the machine or process with true power factor\nequal to one would be the best possible energy efficient operating condition (when the impedance phase angle shown in\nFigure 6 is equal to zero). For a true power factor of less than one, the energy supplied to the load is not used optimally.\nIn such a case, a higher current must be drawn to compensate for the phase shift, o. Where industrial customers operate\nwith power factors below around 0.95, [60], this represents unbalanced additional power demands from the power\nsupplier, and hence additional infrastructure demand leading to additional costs. Furthermore, electrical devices are\nattributed with a 1\u00b2\u00d7R heat loss (R is the electrical resistance), so that increasing the required current while reducing\nthe power factor results in an increase of power loss as heat.\nMeasurements of power factor and total power consumption, for a DC motor driven 63.5 mm in diameter single\nscrew extruder operated at different screw speeds, are shown in Figure 7.\nPower factor\n30\n0.2\n+\n0.8\n0.6\nTotal power (kW)\nSS (rpm)\n\u00a6(a)\n0\n35\n\u00a6(b)\n30\n20\n10\n100\n(c)\n80\n60\n40\n20\n0\n0\n50\n150\n250\n320\nTime (s)\nFigure 7: (a). Power factor, (b). Total extruder power, (c). Screw speed [52]\nFrom this, it can be seen that both the power factor and the total power required are greater for greater processing\nspeeds; however, for higher processing speeds, the temperature uniformity of the process melt output deteriorates\nsignificantly, leading to poor product quality, [28, 52, 53]. Hence, the running of these processes at higher speeds and\n8\nwith the highest possible power factor is problematic, despite being desirable for energy efficiency.\n4. Potential for energy efficiency improvements\nEnergy demands and losses are illustrated in the form of an energy flow diagram, Figure 8. This diagram may be\nEnergy content in\nthe feed material\nDrive motor\nEnergy used\nfor material\nmelting and\nExternal\nheating/cooling\nOther losses\nEnergy for\nother auxiliary\ndevices\nForced\nDrive motor\nlosses\nTransmission cooling\nlosses (gear\nlosses\nbox)\nNatural\ncooling\nlosses\nforming\nFigure 8: A typical energy flow diagram for an extruder [52]\nextended for any auxiliary devices connected with the plant.\n4.1. Drive motor and gear box\nThe key component of any extrusion machine is the screw, which can be driven by a controllable direct current\n(DC) or an alternating current (AC) motor, or indeed by a hydraulic drive [10, 61]. The screw and the motor are\nconnected through a gear box with fixed or adjustable transmission ratio, as shown in Figure 3. For the case of\nan extruder with a DC motor drive, see the schematic, presented in Figure 9. Additionally, the machine can have\nsensing and control devices related to its operation, for example, PID temperature controllers to control set barrel/die\ntemperatures. Extruders with AC motor drives are essentially similar (with no rectifier), and may have additional\ncomponents depending on the type of the motor.\n(@set\nDactual)\nSet screw speed\n(@set)\n+\nPID Motor\nspeed controller\nActual screw\nspeed (actual)\nDC motor\nGear box\nScrew\nArmature voltage (Va)\nchanges to adjust the\nmotor speed\n@actual\nTachometer\ngenerator\nFigure 9: A schematic of an extruder drive mechanism\nFigure 10 shows measured motor power and total power consumptions, for the case of a 63.5 mm diameter single\nscrew extruder with a DC motor, driven at different screw speeds. The contribution of heaters to the total power\ndemand is also indicated. As marked on Figure 10, all the heaters were turned off at around 330 s and this has led to\nsmooth out the total power signal which were fluctuating due to the on-off action of the barrel/die heaters.\nThe drive motor is one of the major energy consuming components of an extruder [31, 34, 40], and, along with the\ngear box, is also responsible for significant energy losses, typically accounting for around 20% of the power supplied\nto an extruder [29]. In particular, DC motors are inefficient when operated at below the rated speed. Commercially\navailable DC motors fall into three main categories: \u201cpermanent magnet\u201d, \u201cseparately excited\u201d and \u201cself-excited\".\nThe first two are more commonly used. A block diagram for a polymer processing extruder with a \u201cseparately excited\ndirect current\" (SEDC) motor is shown in Figure 11.\n9\nPower (kW)\nSS (rpm)\n100\n60\n(a)\n100\n200\n300\n400\n500\n30\n600\n(b)\nTotal power\n20\n0\nMotor power\nAll the heaters turned-off\n-10\n0\n100\n200\n300\nTime (S)\n400\n500\n600\nFigure 10: (a). Screw speed (SS), (b). Motor power and total power signals over the time [52]\nTL\nElectrical dynamics\nGear Box\nVa\n1\nTm\n1\n@m\n@sc\n+\nK\u2081\n+\nLmS + Rm\nJmS+Bm\nN\nV\u2081\nEb\nMechanical dynamics\nKf\nKm\nSpeed controller\nFigure 11: Block diagram of an extruder with a variable field DC motor [62]\n10\n\n\nIn this figure, T is the motor torque, T is the load torque on the screw, Ra and La are the armature resistance and\narmature inductance respectively, K, and K, are the torque constants related to the field and motor, respectively, Ia is\nthe armature current, V+ is the field voltage and Bm and Jm are the damping constant and the steady-state inertia of the\nloaded screw, respectively. For extruders with a permanent magnet motor, the same block diagram is valid without\nthe branch related to the separately excited field (with V\u0192 and K\u0192 block).\nReasons for the popularity of DC motor drives in the polymer processing industry [63, 64, 52] include:\n\u2022 Smooth operation over a wide speed range,\n\u2022 Simplicity in speed control,\n\u2022 Production of a constant/consistent torque from zero to base speed,\nRelatively low power/energy consumption,\n\u2022 Relatively smaller size compared to other drive types with the same capacity,\n\u2022 Compact and simple power circuit, engaged with Silicon-controlled rectifiers,\n\u2022\nEasy installation,\n\u2022 High reliability,\n\u2022 Low intial capital cost, and\n\u2022 Less noisy than AC motors.\nDrawbacks of DC motor drives include the need for maintenance of brushes and commutator as well as the energy\nloss known as \"brush loss\u201d [59]. Green [63] observed that the best power factor that can be achieved by a DC motor\noperated at its top speed is of approximately 0.87, whereas for the best possible efficiency the power factor should be\nclose to 1.\nCurrently, AC motor drives are increasing in popularity since the power factor can be maintained constant across\nthe entire speed range. As a result, companies can avoid paying penalty charges for lagging power factor conditions.\nThe most significant drawback of AC motors is that they require a constant current to produce a constant torque, hence\ndemanding constant cooling regardless of the motor speed [63], and this results in an additional energy cost. Other\nissues include the fact that AC motors are generally larger, by a factor of 1.5-2.1, in volume, as well as being more\ncomplex than DC motors. These issues are becoming less critical nowadays, thanks to advances in electronics such\nas large scale chips and micro-processors. For both types of motors, while under operation at the rated speed, the\nmaximum energy efficiency can be achieved, industrial extruders are typically operated at lower speeds in order to\navoid undesirable fluctuations, particularly of the melt thermal quality.\nBarlow [40] argues that because the displacement power factor of a DC motor drive is proportional to the speed,\nthe power factor reduces as the motor slows down (see Figure 7). Further, it is pointed out that because the diode\nbridge of the input section of a pulse-width-modulated AC vector control drive rectifies the AC into DC, and that the\nenergy is stored in capacitors, the current and voltage waveforms are mutually in phase, and hence the motor operates\nat a power factor, in the range of 0.90 to 0.98. More information on these motor drives can be found in the literature\n[65].\nIt seems that a significant amount of electrical energy may be lost simply as a result of the low power factor\noperation of motor drives [44]. Here, Eickelberg [66] suggests that use of capacitors may be one of the solutions to\nthis problem, to smoothen the power supply. Several examples of the use of capacitors by commercial processes are\nprovided, but it is stated that this is unlikely to be a practical solution for polymer extrusion processes because of\nthe variability that occurs in the load. The installation of a more appropriate form of power factor correction would\nrequire investigation of the relevant issues [67].\nKent [41] observes that in extrusion plant energy usage assessments the energy requirements of motors in equip-\nment such as extruders and injection moulding machines is often over-looked. Other authors [68, 69] report that\nconsiderable energy savings can be achieved by replacing DC motors with AC motors. Lounsbury and Karafilidis\n[70] present factors to be considered in the selection of a drive motor.\n4.2. Barrel and die heaters\nNormally, three different types of heating method can be identified in extrusion. These are known as \"resistance\nheating\u201d, \u201cinduction heating\u201d and \u201cfluid heating\u201d [61].\n11\nResistance heating: This is also called electrical heating, and is the type most frequently used in extruders.\nUsually, electric heaters offer several advantages over fluid and steam heating, such as the possibility of covering\na broader temperature range, cleanliness, easy maintenance, low cost, and better efficiency. As a result of these\nadvantages, fluid and steam heaters have been replaced by electric heaters in modern applications. Currently extruders\ntypically have between two and ten heating zones depending on the size of the extruder.\nIn the most common conventional resistance heaters, the heat from the resistance wire is transferred to ceramic\nsegments that surround the outer surface of the barrel. This heats the barrel up until its inner surface is hot: the heat\nis then transferred to the plastic in the machine so it can be processed. With this heating method, much of the heat\ngenerated is wasted.\nInduction heating: In this case, an AC current is passed through the primary coil surrounding the extruder barrel.\nThis gives rise to an eddy current. Where the material being processed has significant relative permeability, heat may\nalso be generated by magnetic hysteresis. A high energy density can be achieved quickly with induction heating. The\nfrequency of the AC is selected depending on the size and material type and the heat penetration depth.\nWith the advantage of rapid heating and energy efficiency, induction heating has been used in many industrial\napplications [71]. The key advantage of induction heating over resistance heating is that the extruder barrel itself\nbecomes the heating element. This eliminates the conduction problem that exists with conventional heaters. As there\nare no ceramic layers, clamping bands, or water jackets to heat up, the heat is direct and instantaneous. Furthermore,\nthe induction heating generates a very even and precise heat profile, ensuring consistent heating of the polymer melt\nand leading to improved product quality. The application of induction heating to polymer processing has been devel-\noped by the Nordson Xaloy Company, which claims a reduction in heating related energy consumption of up to 50%\ncompared with typical band electric heaters [72, 73].\nFluid heating: Fluid heating uses hot liquid or steam passing through pipes/tubes. This can ensure even temper-\nature distribution but has significant disadvantages, such as demanding high levels of maintenance, the possibility of\nleaking or corrosion, and system complexity.\nAn alternative all three heating methods, is to supply the energy to the material being processed as drive power\nrather than as heater power [44, 74]. In this way shearing of the material leads to both heating and mixing. Of\ncourse, direct heating and cooling can be essential, to maintain the process thermal stability, but as guiding principles\nit would be preferable to minimize the direct heating and to avoid exceeding the melt temperature, in order to avoid\nunnecessary energy costs.\n4.3. Control electronics and monitoring devices\nA number of control and monitoring devices are used in extrusion lines, such as speed and temperature controllers and\nindicators; pressure indicators and gauges; dimension scanners; feed monitoring devices; current indicators; relays;\nswitches; alarms; etc.. All of these use some power for their operation; however, this is insignificant compared to the\ntotal energy demand. This is evident from the experimental data presented in Figure 10-(b), the amount consumed by\ncontrol electronics and monitoring devices are shown by the difference between blue and red lines (i.e., the difference\nbetween to total and motor powers) after turning off all the heaters, or by the blue line roughly between 540-570 s\nafter turning off all heaters and the motor.\n4.4. Auxiliary equipment\nAuxiliary equipment, such as pelletizers, gear pumps and screen changes, might also consume enough power to be\nconsidered in energy evaluations. Rice [75] suggests that improved energy efficiency for the entire operation of the\nplant can be achieved by combining processes.\n4.5. Process cooling\nProcess cooling is required where there is a need to remove excess heat in order to maintain the process thermal\nstability. This can be achieved by fan coolers attached along the barrel, or by cooling the screw core or the barrel wall\ninternally, using a cooling fluid such as water or oil. Air cooling provides slower changes in temperature compared\nwith liquid cooling.\n12\nAlthough cooling helps to ensure stable process operation, improper cooling, particularly cooling of the screw,\ncan lead to the generation of undesirable process fluctuations. Strauch [31] argues that excessive screw core cooling\ncan lead to a reduction in throughput rates, while at the same time affecting the melt temperature. This also impacts\non the pumping stability as a direct consequence of altering the viscosity of melt. Periodic or random temperature\nvariations of the extruder metal surfaces can arise as a result of cooling problems with the screw or barrel, and these\nmay lead to melt flow problems such as melt viscosity fluctuations [76].\nThe research of Womer et al. [35] into the effects of cooling indicated that, where water cooling is used rather\nthan air cooling, the extruder consumes more energy, irrespective of the material being processed. In consequence,\nit is recommended to use air cooling only, in conjunction with a properly designed screw, unless extensive cooling is\nrequired.\n5. Trends in polymer processing energy efficiency improvements\n5.1. Machine development and operational modifications\nAmong the current research and development \"hot topics\" for process energy efficiency, the three most likely to deliver\nsignificant benefits [63] are: (i) the design of direct drive extruders, (ii) improvements in heater and barrel design to\nreduce heat losses, and (iii) thedevelopment of \u201cadvanced vector control alternating current (AC) drives\u201d. In addition\nto these, there are reports which claim that direct drive machines offer further advantages including energy efficient\noperation, narrow footprint, quiet operation, and low maintenance requirements. The use of insulation blankets has\nalso becoming popular in energy saving of machines.\n\"Load management\u201d offers the possibility to save power in the near term. The idea is to achieve the required energy\ncost reduction by maintaining the load factor, rather than focusing on power consumption per se. Since commercial\nelectrical energy supply rates usually depend on the peak demand made by the customer, this can reduce the cost of the\nelectricity used over the duration of the charging period. To operate an effective load management plan, it is necessary\nto have a plant monitoring system to study the real-time plant power usage. In Kent's [41] discussion of energy saving\nin polymer processes, it is pointed out that the purchase of energy efficient capital equipment is profitable in long-run\ndespite initial capital costs.\n5.2. Waste heat energy recovery\nAlthough a significant amount of process heat is removed purposely to maintain the thermal stability in polymer\nextrusion, there has been insufficient attention on the recovery of waste heat for useful work.\nThe challenge [42, 77] is to find a means of re-use for that energy. Future research is required to explore such\nopportunities. For example, the recovered heat might be used for pre-heating the material prior to feeding into the\nhopper, or it could simply be used for space heating.\nFor some materials, resins need to be pre-heated prior to processing to remove moisture. In this case, part of the\nsupplied energy is lost through the evaporation of the moisture as water vapour, part is lost to heating of the surround-\nings, and the rest contributes to heating the resin. If the drying operation takes place remotely from the processing\nmachine, then the heat absorbed by resin will be lost during transit to the processing machine. Therefore, the re-design\nof the drying system, as an in-line step of the processing machine, should help to reduce overall processing energy\ncosts.\n5.3. Material choice, material recycling and disposal\nIt is usually the case that the manufacture and fabrication of plastic products makes lower energy demands than\nequivalent traditional metallic or glassware products. The development of new resins that can be processed at lower\ntemperatures, and hence for reduced energy, is part of the growing tide of interest to cut process energy expenses\neven further [78, 79, 80, 81]. On the other hand, polymeric materials based waste management has become a global\nconcern. A wide range of recycling techniques are available depending on the types of polymers and the manufacturing\ntechniques used [82]. In regard to process energy costs, the energy consumption for plastics is lower than for materials\nsuch as paper, glass, tin, and aluminium. What is more, Rosato et al. [34], the incineration of plastics as part of\nmunicipal waste yields much more energy than other material waste, such as food waste, paper and rubber, and waste\nvolume can be reduced by 90-98%.\n13\n6. Advanced process monitoring and control: adoption of Industry 4.0\nAdvanced process monitoring and control can play a vital role in achieving good product quality as well as in energy\noptimization. Some approaches are discussed in this section together with experimental results.\n6.1. Industry 4.0 and Internet of Things\nWidespread uptake of on-line or in-line monitoring and control in manufacturing processes has been enabled by\ncomputerised communications, earning it the epithet: \u201cThe fourth industrial revolution\u201d, or \u201cIndustry 4.0\u201d for short\n[83, 84]. The key requirements for an Industry 4.0 manufacturing process are: sensors - the means to observe the\nprocess as it currently stands; actuators - the means to modify the process; electronic communications - the means to\npass sensor or control information; and a decision-maker - to determine the course of action to be taken based on the\ninformation received.\nThe precise nature of the decision-maker is a point of some contention in the literature. For some, the decision-\nmaker is a computer-based Artificial Intelligence (AI), which would work completely autonomously, and steadily\nimproving its decision-making capability based on learned patterns of experience. There is still significant value in\nthe Industry 4.0 infrastructure even without an AI capability making the control decisions. For some applications, a\nrule- or model-based computer program would be effective, and many companies now boast of having an Industry\n4.0 implementation of this form. In some applications, having the sensor information fed to a control centre, means\nthat human decision-making can be facilitated and supported. Such control centres are valuable for the control of\nprocesses that are in remote or difficult to access locations, such as the health-monitoring of in-flight aircraft.\nIndustry 4.0 technology is also becoming an increasingly common tool in the home or in social care settings. Here,\nthe more common terminology is Internet of Things (IoT) [85], referring to communications connectivity between\n\"Smart\" devices. These smart devices are pieces of equipment which can send or receive information, in other words,\nthey are equipped with sensors or actuators. Thus, in the home, one might simply ask \u201cAlexa\u201d [86] to turn on the lights,\nbut in time it could easily be imagined that Alexa could modify the home heating to match the schedule inferred from\nfamily member diaries. In a patient care setting [87], the IoT system could be collecting valuable health-related\ninformation and relaying that to a control centre. Care staff or an AI system might detect anomalies and prompt a\ncheck of the patient. It is easy to see that these concepts are similar, whether applied to the home or to industry, and\nthe technology is pervasive and rapidly developing.\nIn the context of the polymer processing industry, it is clear that Industry 4.0 will not only enhance plant operation\nand its maintenance schedule, but also to energy efficiency [88, 89]. The lessons learned over the past half century,\nand reviewed in earlier sections of this paper, are ready to be applied. Wherever there are frequent variations in the\nfeedstock material, processing rate demand, or other factors, with an Industry 4.0 infrastructure in place, it becomes\npossible to monitor performance over time, and to make controlled changes. Systems health monitoring can be used\nto reduce life limiting loads on mechanical parts [90] as part of the maintenance strategy.\nGreater investment in IoT enabled devices will become an increasing imperative, for any polymer processing plant\nthat wishes to reduce its energy footprint, reduce processing costs, and sustain the mechanical plant more effectively.\nAs soon as the Industry 4.0 infrastructure is operational, development of the decision-making capability can be begun;\nwhether that is to be based on AI principles, control systems mathematics or other rule or model based paradigms.\n6.2. Ultrasound\nThe application of ultrasonic waves to reduce viscosity and thus energy consumption in polymer processing has been\ndiscussed in several recent reports [91, 92, 93, 94, 95, 96]. Maintaining a consistent melt viscosity enables improved\nprocess ability, leads to fewer product defects, reduces energy consumption and reduces materials wastage [97, 98, 99].\nChen et al [92] and Zhang and Li [93] report on the use of ultrasound vibration to influence polypropylene (PP)\nmelt, leading to non-Newtonian flow characteristics with reduced viscosity. Other authors have examined the rela-\ntionships between temperature, pressure, work-stuff throughput, the energy consumption and the ultrasonic intensity\n[95, 96]. The integration of ultrasound into a closed loop extruder control system has been developed and introduced\nin detailed by Nguyen et al. [100], who showed that controlling just the temperature, or just the ultrasonic output,\ngave a better time response and reduced energy consumption, than when trying to control both the temperature and\nultrasound together.\n14\n6.3. Closed loop melt temperature control\nMelt temperature can serve as a proxy parameter for melt viscosity and can thus be used to determine the melt quality.\nRecent works by Abeykoon et al [101, 102, 103] incorporate a fuzzy logic approach for the real-time closed loop\ncontrol of melt temperature, and demonstrated excellent performance in achieving desired set temperatures.\nAt present, in the majority of polymer processes, melt pressure and melt temperature are taken as the key param-\neters for process functionally and control. Screw speed, and barrel and die set temperatures are taken as the main\nprocess control parameters, but there is no actual feedback taken from the process melt for making process control\ndecisions, so no corrective actions can be taken to avoid product defects. Furthermore, this affects the production rate\nleading to wasted energy, labour and raw materials. Hence, combined process monitoring and control approaches,\nwhich can observe the melt quality and take control actions would represent a major development of polymer pro-\ncesses.\n7. Applying computational process simulation and dynamical systems to Control\nDevelopments in computational process simulation methods has the potential to revolutionise the design of manu-\nfacturing tooling and processes. Capabilities such as finite element method and computational fluid dynamics have\ndeveloped very significantly, with most commercial packages (e.g. [104, 105, 106, 107, 108, 109]) now offering\nmulti-physics simulation: simulation of static and transient solid mechanics, fluid dynamics, thermodynamics, elec-\ntricomagnetism, and in many cases, much more. Given the improvements in computer hardware, storage capacity\nand the development of parallel processing architecutures, computational analyses with high levels of geometric or\nmaterial modelling complexity can now be readily envisaged. This is a huge opportunity to grasp, and one for which\nmost industrial plants are not fully prepared.\n7.1. Computational mechanics methods\nThe key computational capabilities pertinent to polymer processing include a variety of modelling techniques for vis-\ncoelastic materials irrespectively of whether they should be treated as solids or liquids. Where previously, modelling\nbased on a Lagrangian Finite Element formulation, [110], would have been unable to capture the necessary defor-\nmation, Eulerian formulations, [111], and Smooth Particle Hydrodynamics (SPH), [112] and [113], now have that\npotential. Perhaps it is not enough to model using just one technique or another, but to capture one pertinent aspect\nof the physics in one region of the process, and another aspect in another region using such modelling techniques as\nco-simulation and subdomain modelling.\nNot only it is the type of analysis that is of importance to capture the process simulation requirement, but also the\nway in which the material properties are represented. The distinction between solid and liquid is no longer so clear-\ncut. There are very many material models, developed and applicable to different analysis types, that will capture the\nessential material property physics of any realizable material. Polymers offer particular challenges, but characteristics\nsuch as static stress-strain, strain rate dependence, creep, and temperature dependence are readily modelled. Chemical\nchanges, including exo- or endothermic reactions, and cure shrinkage still present a particular challenge to thermoset\npolymers, but for thermoplastic processing the challenge of representing materials properties at glass transition, and\nextent of crystallization, are, if not straight forward, at least more tractable [114, 115, 116]. For the characterisation\nof polymers undergoing extrusion processing Abeykoon et al, [117], have made significant investigations.\nWith the increase of computing power and the capability for higher model complexity there has been increased\ninterest in the application of computational mechanics methods using sophisticate material models to the simulation\nof the extrusion process. In the past decade authors such as Zairi et al, [118], have been able to model plastic flow,\ntaking account of viscoelastic properties, within a die of finite constrained section, and have been able to make some\nprediction of the microstructure of the end product. While the main focus has been of the effect of the die geometry\non the extruded product, another important consideration is the design of the die tooling, and the loads that must\nwithstand during processing [119].\n15\n7.2. Statistical process control\nThe biggest challenge for the modelling of the polymer extrusion process is to understand the conditions that give rise\nto large variations in product output quality. Where small changes in the processing parameters lead to only small\nchanges in output, even where the change is clearly non-linear, one might take a step-wise approach to linearize the\nparameters and have a measurable and definable set of limits for process control. This is the ideal, textbook, approach\nto manufacturing process excellence, LEAN manufacturing or Six Sigma, [120], [121]. In constrast, where small\nchanges in processing input lead to significant changes in the output quality, a manufacturing engineer would say that\nthe process is not in statistical control. The usual manufacturing engineer's approach for regaining statistical control\nis to monitor parameters to within ever tighter bounds. Clearly this would lead to increased manufacturing costs, and\ncould make the process financially unviable. The approach may ultimately even be completely unsuccessful.\nThe factors that would generally be considered for statistical process control would include the initial conditions:\nhow the process is started up, and the material condition on start-up, factors directly related to the geometry of the\nextruder and the die, environmental conditions and changes, and changes in the raw material batch.\nIn order to control the process, certain actions might be taken. The applied drive torque, a function of the power\nsupplied to the extruder, could be controlled in direct response to measurements relating to the process quality. In-\nevitably there is a delay between making the measurement, and making a change, primarily because the measurement\nis taken at some point down-stream in the process. There is a considerable level of research activity in this area, with\nvarious computational schemes for assessing these measurements and informing the choice of control action to be\ntaken. Clearly this data processing time must be minimised to minimise the feedback delay. Wagner et al, [122], and\nMcKay et al, [123], recognised that the fundamental requirement was to use the measurements to infer the local mate-\nrial viscosity within the process, using artificial intelligence methods such as neural nets. Chen et al, [124], employed\na power law model. Methods based on fuzzy logic were developed by McAfee, [125], McAfee and Thompson, [126],\nand later by Liu et al, [127] and Abeykoon [103]. More recent work includes the soft sensor technique of Deng et al,\n[128] and Abeykoon [102], and the multi-objective optimisation approaches presented by Carrano et al, [129].\n7.3. Non-linear system dynamics\nWhere the dynamics of the manufacturing process make statistical process control challenging, a radically different\napproach is needed. Non-linear systems dynamics [130] has been a topic of considerable research and development,\nmainly by researchers with a strong background in applied mathematics, and with applications including but by no\nmeans limited to manufacturing engineering applications. By developing a non-linear system dynamics representation\nof a manufacturing process, it is possible to explore how process parameters could influence the process dynamics,\nand from this pin-point the controling factors or initial conditions.\nIn the field of polymer extrusion modelling, McKinley, et al, [131], used laser doppler velociometry to visualise\nthe flow towards an abrupt contraction, and found that the flow near the tip of the contraction could show time periodic\nand aperiodic behaviour. Graham [132] modelled the fluid behavious regarding wall slip and was able to replicate the\nlarge amplitude periodic and aperiodic oscillations observed experiementally. Smith et al, [133, 134] demonstrate a\nsteady-state solution through modelling the polymer as a purely viscous non-Newtonian material, and optimising die\ngeometry.\nThe critial feature in the extrusion process is the variation in the material property of the polymer as it passes\nthrough the process. As the material is being worked mechanically, it becomes heated. As it gets hotter, the elastic\nstiffness and viscosity are reduced. The resulting thermal expansion gives rise to a localised increase of pressure.\nAs a result of both the reduced viscosity and increased pressure, the polymer would pass more readily through the\nprocess: and would require less working. Reduced working would result in reduced heating, with the result of inceased\nstiffness, increased viscosity and reduced pressure. It is clear that a cyclic response can be expected. Now consider\nhow the viscosity of the polymer changes with strain rate and the difference of time dependency in glass transition and\ncrystallization, it bceomes clear that there is a level of complexity between production rate, localized temperature and\nviscous response of the material: in the simplistic mass, spring and damper system it is the variability of the damper\nas well as the forcing that is key to understanding the dynamical system.\nA non-linear dynamical systems model of the polymer extrusion process might be modelled as follows. First,\nthe polymer has the properties of both an elastic solid, and a viscous fluid: it can be idealized as a mass, spring\nand damper system, as shown in Figure 12. The non-linear dynamics of such systems have been studied for many\n16\nidealized situations of varying applied force, F(t), and their modelling involves the selection of initial conditions, and\nthen computing the transient behaviour until a long term behaviour becomes apparent.\nk(t)\nx(t)\nc(t)\nm\nF(t)\nFigure 12: Idealised mass, spring and damper system\n-\nFigure 13 shows the time varying position, x(t) and velocity x(t) of the mass, driven by the time varying force\nF(t). The initial conditions for the system might be any point on the x \u2014 x plane, and one might think of particular\nexamples of such initial conditions being the tail end of the solid or dashed curved arrows. As time progresses, the\nsystem would move on from that initial condition, moving towards the head of the arrow. The dashed arrows indicate\nthe evolution away from an unstable limit cycle \u2013 the dashed closed loop. Any initial condition that begins on the\nunstable limit cycle will eventually decay inwards or outwards, and migrate towards a stable limit of some sort. The\nsolid arrows indicate evolution towards either a stable limit cycle the solid closed loop, or towards a steady-state\nsolution \u2015 the solid dot. The stable limit cycle represents a periodic motion. The area enclosed by the unstable limit\ncycle gives an indication of the relative level of damping in the system: it is important to note that increasing the\ndamping would increase the area within the unstable limit cycle, so that a larger proportion of initial conditions would\nultimately lead to a steady-state solution; however in the case of initial conditions sufficiently close to the stable limit\ncycle, periodic motion would still ensue. The notion that damping removes energy from a dynamic system is true, but\nthe energy required for periodic motion is renewed at every cycle by the time varying force, F(t).\nIn the context of polymer processing, periodic motion represents both unnecessary energy utilisation, and reduced\nprocess control. This explanation is more easily understood when described in the context of everyday experience.\nx(t)\nx(t)\nFigure 13: Time evolving position and velocity diagram\nConsider a friction driven vibration such as the vibration of a bowed violin string, brake squeal, or a singing wine glass.\nIn each case, the string, the brake disc and the glass have a stiffness, a density and some form of damping is present.\nUnder particular forcing conditions such as applied by a well-rosined violin bow, a certain braking pressure or being\nstroked with a wet finger, these structures vibrate at a well-defined pitch. The fact that the pitch is well-defined means\nthat the vibration is periodic. Under different forcing conditions: a bow without rosin, a different braking pressure\nor a dry finger, the contact slips with constant velocity. This is a steady-state solution: a force is being applied and\nso there is a displacement, but it is a constant force and a constant displacement. For a detailed account of such an\nanalysis, and the computational algorithms used [135].\n17\nThis description is necessarily rather simplistic, but it provides a basis for reviewing the pertinent features required\nfor modelling the extrusion system. In place of displacement and velocity one should be considering the extruded\nvolume of material as a function of time. In place of force, consider the power output of the machine and what\nresistance it meets from the material being processed. In an ideal process, one would wish for constant extruded\nvolume and constant levels of resistance, so that with an appropriate choice of initial conditions and damping, a steady-\nstate outcome can be achieved. The elastic response and the viscosity of the polymer depend on temperature and strain\nrate, and as such are the time dependent system stiffness, k(t), and time dependent damping, c(t), respectively. Recent\nwork in this vein, but applied to blown film extrusion, is presented in a detailed review paper by Pirkle et al, [136].\n8. Conclusions\nThere have been some very significant developments in computational modelling capability, which means that the\nsimulation of manufacturing processes such as polymer extrusion is now feasible. Multi-physics approaches, combin-\ning material flow and thermal behaviour of the material can, in theory, be modelled. Advances in computer hardware\nmean that models with very high levels of geometric complexity, and therefore high numbers of computational degrees\nof freedom are within reach. Co-simulation modelling, including the modelling of not only the material undergoing\nextrusion, but also of the working state of the dies and the extrusion machine itself, is also within reach, meaning\nthat optimal design of tooling for improved life and reduced machine maintenance demands are additional areas for\nfurther development.\nComplex models of the process and material property variation can also be used to create the simplified models\nrequired for a non-linear systems dynamics modelling approach. On that basis, measurable parameters that would\nhave an effect on the stability of the real production process can be examined, and warning limits found. Even where\nsuch an approach might not be able provide an accurate prediction of limits, it could provide insight that would lead\nto practical solutions or avoidance of particular operating regimes.\n9. List of Abbreviations\nTerm\nDefinition\nAC\nAlternating current\nDC\nDirect current\nIoT\nInternet of Things\nSEC\nSpecific energy consumption\nSEDC Separately excited direct current\n10. List of Symbols\n18\nTerm\nDefinition\nBm\nDamping constant\n\u0421\u0440\nSpecific heat capacity\nEin\nEnergy consumed\nElosses\nEnergy loss\nF(t)\nA time varying force\nI\nLine current\nIa\nArmature current\nIf\nField current\nJm\nSteady-state inertia of the loaded screw\nkm\nThermal conductivity\nKf\nTorque constant related to the field\nK\u2081\nTorque constant related to the motor\nLa\nArmature inductance\nN\nGear ratio\nP\nActive power\nQ\nReactive power\nR\nElectrical resistance\nS\nApparent power\nRa\nArmature resistance\nTb\nExtruder barrel set temperature\nTL\nLoad torque on the screw\nTmelt\nMelt temperature\nTm\nMotor torque\nx(t)\nA time varying position\nx(t)\nA time varying position velocity\nV\nLine voltage\nVa\nArmature voltage\nVbx\nThe component of the barrel velocity in the transverse direction\n19\nV\u0192\nField voltage\nVj\nThe resultant relative velocity\nTerm\nDefinition\nCOS\nThe displacement power factor\nPm\nMelt density\n\u03b7\nMelt viscosity\nWactual\nActual screw speed\nW set\n\u03bb\nC\u2081\n\u03a9\nSet screw speed\nTemperature of the solid bed\nRate of melting\nReferences\n[1] British plastics federation, about the british plastics industry, Available at: http://www.bpf.co.uk/Industry/Default.aspx., Last viewed: 27 of\nFebruary 2019.\n[2] Plasticseurope: Plastics the facts 2017 an analysis of european plasticsproduction, demand and waste data, Available at:\nhttps://www.plasticseurope.org/application/files/5715/1717/4180/Plastics the facts2017 FINAL for website ne page.pdf, Last viewed: 27 of\nFebruary 2019.\n[3] Market outlook:\nAvailable at:\nPolymers world turned on its head (icis chemical business),\nhttp://www.icis.com/resources/news/2013/10/25/9718765/market-outlook-polymers-world-turned-on-its-head/, Last viewed: 31 of\nAugust 2016.\n[4] Available at: http://www.plasticseurope.org/, Last viewed: 10 of August 2012.\n[5] J. Vlachopoulos, D. Strutt, Overview: Polymer processing, Materials Science and Technology 19 (2003) 1161\u20131169.\n[6] European best practice guide: Low energy plastics processing, Available at: http://www.tangram.co.uk/TI-Energy-LowLast viewed: 27 of\nFebruary 2019.\n[7] K. L. O. C. S. L. G. Kaur, K. Uisan, Recent trends in green and sustainable chemistry & waste valorisation: Rethinking plastics in a\ncirculareconomy, Current Opinion in Green and Sustainable Chemistry 9 (2018) 30\u201339.\n[8] J. Frankland, Mission (nearly) impossible: Estimating extrusion melt temperature, Plsatics TechnologyAugust Issue.\n[9] M. J. Stevens, J. A. Covas, Extruder principles and operation, 2nd Edition, Chapman & Hall, 1995.\n[10] C. Rauwendaal, Polymer Extrusion, 4th Edition, Hanser, 2001.\n[11] C. Abeykoon, Polymer Extrusion: A Study on Thermal Monitoring Techniques and Melting Issues, Lap Lambert Academic Publish-\ning: Verlag, 2012.\n[12] C. I. Chung, E. M. Mount, D. E. McClelland, Energy efficiency in plasticating screw extrusion, Proceedings of 176th Symposium in Energy\nConservation in Textile and Polymer Processing (1979) 21-36.\n[13] J. Agassant, Polymer Processing, Hanser-Munich, 1991.\n[14] A. N. Wilkinson, A. J. Ryan, Polymer Processing and Structure Development, Springer-Netherlands, 1998.\n[15] T. Osswald, J. Hernandez-Oritz, Polymer Processing, Hanser-Munich, 2006.\n[16] W. H. Darnell, E. A. J. Mol, Solids conveying in extruders, SPE Journal 12 (1956) 20-29.\n[17] B. H. Maddock, A visual analysis of flow and mixing in extruder screws, SPE Journal 15 (1959) 383-389.\n[18] Z. Tadmor, Fundamental of plasticating extrusion. i. a theoretical model for melting, Polymer Engineering and Science 6 (1966) 185\u2013190.\n[19] Z. Tadmor, I. Klein, The formulation of a mathematical model for plasticating screw extrusion, Rubber Chemistry and Technology 42 (1969)\n780-789.\n[20] Z. Tadmor, I. Klein, Engineering principles of plasticating extrusion, 1st Edition, Society of Plastics Engineers, 1970.\n[21] Z. Tadmor, C. Gogos, Principles of polymer processing, Wiley-New York, 1979.\n[22] E. Severs, Rheology of polymers, Van Nostrand Reinhold Inc, 1962.\n[23] C. Abeykoon, P. Perez, A. L. Kelly, he effect of materials' rheology on process energy consumption in polymer extrusion, Polymer Engi-\nneering and Science 60 (6) (2020) 1244-1265.\n[24] Y. Breaux, J.-Y. Charmeau, M. Moguedet, A simple model of throughput and pressure development for single screw, Journal of Materials\nProcessing Technology 209 (2009) 611-618.\n[25] H. S. Rowell, R. D. Finlayson, Screw viscosity pumps, Engineering 114 (1922) 606\u2013607.\n[26] H. S. Rowell, R. D. Finlayson, Screw viscosity pumps, Engineering 126 (1928) 249-387.\n[27] C. Abeykoon, Modelling and control of melt temperature in polymer extrusion, Ph.D. thesis, Queen's University Belfast, United Kingdom\n(2011).\n[28] C. Abeykoon, J. Vera-Sorroche, A. L. Kelly, E. C. Brown, P. D. Coates, J. Deng, K. Li, N. Karnachi, E. Harkin-Jones, M. Price, Process\nefficiency in polymer extrusion: Correlation between the energy demand and melt thermal stability, Applied Energy 135 (2014) 560-571.\n[29] G. A. Kruder, R. E. Nunn, Optimizing energy utilization in extrusion processing, SPE ANTEC Technical Papers (1981) 648-652.\n20\n\n\n[30] J. M. McKelvey, Energy utilisation in extrusion, Advances in Polymer Technology 3 (1983) 205-211.\n[31] T. H. Strauch, F. Gliese, G. Menges, Energy usage and conservation in extrusion plants, Plastics and Rubber Institute (Polymer Extrusion\nIII Conference), London, England 1 (1985) 2.1-2.10.\n[32] J. Anderson, C. Rauwendaal, M. del P. Noriega, Troubleshooting extrusion problems, SPE ANTEC Technical Papers 1 (1997) 127\u2013134.\n[33] H. Falkner, Energy cost savings in motive power, Water and Environment Journal 11 (1997) 225-227.\n[34] D. V. Rosato, N. R. Schott, D. V. Rosato, M. G. Rosato, Plastics engineering, manufacturing & data handbook: Fundamental and processes,\nVol. 2, Kluwer Academic Publishers, 2001.\n[35] T. W. Womer, W. S. Smith, R. P. Wheeler, Comparison of two different cooling methods for extrusion processes, SPE ANTEC Technical\nPapers (2006) 796-801.\n[36] About the industry, Available at: http://www.bpf.co.uk, Last viewed: 27 of February 2012.\n[37] K. M. Cantor, Analyzing extruder energy consumption, SPE ANTEC Technical Papers 2 (2010) 1300-1306.\n[38] R. V. Heur, M. Verheijen, Power quality & utilization guide: Plastics industry-energy efficiency, Available at: http: //www.leonardo-\nenergy.org/files/root/pdf/Application/Guide/Plastics.pdf, Last viewed: 15 of August 2012.\n[39] J. Deng, K. Li, E. Harkin-Jones, M. Price, N. Karnachi, A. Kelly, J. Vera-Sorroche, P. Coates, E. Brown, M. Fei, Energy monitoring and\nquality control of a single screw extruder, Applied Energy 113 (2014) 1775\u20131785.\n[40] S. Barlow, Reducing electrical energy costs for extrusion processes, SPE ANTEC Technical Papers (2009) 1157-1162.\n[41] R. Kent, Energy management in plastics processing - part 10: Motors and drives, Available at: http://www.tangram.co.uk, Last viewed: 22\nof february 2010.\n[42] B. Drury, The control techniques drives and controls handbook, The Institution of Electrical Engineers, London-UK, 2010, pp 308-312.\n[43] J. V. DeJuneas, How to reduce electrical energy in the extrusion process, TAPPI Pap Synth Conf (1975) 49\u201350.\n[44] J. M. Dealy, Energy conservation in plastics processing: A review, Polymer Engineering and Science 29 (1982) 528-535.\n[45] C. I. Chung, Energy efficinecy in single screw extrusion, SPE ANTEC Technical Papers (1983) 919-920.\n[46] R. Rasid, A. K. Wood, Effect of process variables on melt tempearture profiles in extrusion process using single screw plastics extruder,\nPlast. Rubb. Comp. 32 (5) (2003) 193\u2013198.\n[47] E. C. Brown, A. L. Kelly, P. D. Coates, Melt temperature homogeneity in single screw extrusion: effect of material type and screw geometry,\nSPE ANTEC Technical Papers (2004) 183\u2013187.\n[48] A. L. Kelly, E. C. Brown, P. D. Coates, The effect of screw geometry on melt temperature profile in single screw extrusion, Polymer\nEngineering and Science 46 (12) (2006) 1706-1714.\n[49] A. L. Kelly, J. V. Sorroche, E. C. Brown, P. D. Coates, Improving thermal efficiency of single screw extrusion, SPE ANTEC Technical\nPapers 2 (2012) 1080\u20131087.\n[50] J. V. Sorroche, A. L. Kelly, E. C. Brown, P. D. Coates, N. Karnachi, E. Harkin-Jones, K. Li, J. Deng, Thermal optimisation of polymer\nextrusion using in-process monitoring techniques, Applied Thermal Engineering 53 (2) (2012) 405\u2013413.\n[51] C. Abeykoon, M. McAfee, S. Thompson, K. Li, A. L. Kelly, E. C. Brown, Investigation of torque fluctuations in extrusion through monitoring\nof motor variables, Proceedings of 26th PPS Annual Europe/Africa regional meeting, Larnaca, Cyprus (2009) Paper No: 22-0.\n[52] C. Abeykoon, J. Vera-Sorroche, A. L. Kelly, E. C. Brown, P. D. Coates, K. B. Howell, J. Deng, K. Li, N. Karnachi, E. Harkin-Jones,\nM. Price, Investigation of the process energy demand in polymer extrusion: A review and an experimental study, Applied Energy 136 (2014)\n726-737.\n[53] C. Abeykoon, A. L. Kelly, E. C. Brown, P. D. Coates, The effect of materials, process settings and screw geometry on energy consumption\nand melt temperature in single screw extrusion, Applied Energy (2016) 880-894.\n[54] R. S. Mallouk, J. M. Mckelvey, Power requirements of melt extruders, Industrial and engineering chemistry 45 (1953) 987\u2013989.\n[55] K. Wilczynski, A computer model for single-screw plasticating extrusion, Polymer-Plastics Technology and Engineering 35 (3) (1996)\n449-477.\n[56] E. Lai, D. W. Yu, Modeling of the plasticating process in a single screw extruder: A fast track approach, Polymer Engineering and Science\n40 (5) (2000) 1074-1084.\n[57] C. Abeykoon, M. McAfee, K. Li, P. J. Martin, J. Deng, A. L. Kelly, Modelling the effects of operating conditions on motor power consump-\ntion in single screw extrusion, Proceedings of LSMS 2010 and ICSEE 2010 6329 (2) (2010) 9-20.\n[58] T. Wildi, Electrical machines, drives, and power systems, 4th Edition, Prentice Hall, 2000.\n[59] S. J. Chapman, Electrical machinery fundamentals, McGraw-Hill, 2005, pp 261-264, 524-527.\n[60] Motor efficiency depends upon power factorAvailable at: http://www.actel.com/documents/Motor_PowerFactor_WP.pdf.\n[61] L. Halsz, Control methods in polymer processing, Elsevier, 1993.\n[62] C. Abeykoon, M. McAfee, K. Li, P. J. Martin, A. L. Kelly, The inferential monitoring of screw load torque to predict process fluctuations in\npolymer extrusion, J. Mater. Process. Technol. 211 (12) (2011) 1907-1918.\n[63] R. M. Green, A comparison of AC and DC extruder drives, IEEE Industry Applications Society Annual Meeting 2 (1989) 1663\u20131666.\n[64] N. Mulakken, Taufik, M. Anwari, Power quality measurements of a dc motor drive, International Confernce on Power and Energy Systems,\nLangkawi, malaysia (2008) 98\u2013101.\n[65] S.-H. Kim, Electric Motor Control: DC, AC, and BLDC Motors, Elsevier, 2017.\n[66] F. Eickelberg, Continuous vacuum metallizing: it's more than tinsel . . . much more, Modern Plastics 54 (1977) 42-45.\n[67] C. S. P. Kumar, S. P. Sabberwal, A. K. Mukharji, Power factor measurement and correction techniques, Electric Power Systems Research\n32 (2) (1994) 141-43.\n[68] R. A. Sfeir, Measured savings of DC to AC drive retrofit in plastic extrusion, Proceedings of the 13th Industrial Energy Technology Confer-\nence, New Orleans, USA.\n[69] Extruder line updated with ac vector drives, Available at: http://www.yaskawa.com/site/dmdrive.nsf, Last viewed: 10 of August 2012.\n[70] D. C. Lounsbury, A. G. Karafilidis, Optimizing extruder drive selection: A review of the factors involved, SPE ANTEC Technical Papers\n(1995) 76-79.\n[71] A. A. not stated, Induction heating applications - the processes, the equipment, the benefits. kha provide full details, Internet: http://www.efd-\n21\ninduction.com/ (Year not stated, last access date not stated) Pages not stated.\n[72] B. F. T. Gregory E. Tremblay, Dan Sowle, Investigation of injection moulding performance using induction barrel heating. kha provide full\ndetails, Journal not stated (Year not stated) Pages not stated.\n[73] T. Deligio, Screws and barrels the focus of energy-savings efforts, Modern Plastic Worldwide (2008) 44\u201345.\n[74] J. M. McKelvey, Theory of adiabatic extruder operation, Industrial & Engineering Chemistry 46 (4) (1954) 660\u2013664.\n[75] W. T. Rice, Conservation of energy and raw materials by utilization of gear pumps in conjunction with extruder processes, Energy Conference\n& Exhibition, New Orleans, LA, USAASME Publication 80-Pet-2.\n[76] E. L. Steward, Control of melt temperature on single screw extruders, SPE ANTEC Technical Papers 1, paper No: 813.\n[77] K. Bhattacharjee, Energy conservation opportunities in industrial waste heat recovery systems, Energy Engineering 107 (2010) 7-13.\n[78] J. J. Baron, C. E. McChesney, S. M. Sinker, Extrusion blow moldable polyethylene terephthalate, SPE ANTEC Technical Papers (1981)\n713-715.\n[79] Energy saving pet, Modern Plastics 58 (1981) 72.\n[80] Pp that cuts users' energy costs, Modern Plastics 58 (1981) 14.\n[81] R. D. Deanin, Decreasing energy requirements in plastics processing, Polymer-Plastics Technology and Engineering 14 (1980) 37\u201345.\n[82] B. I. D. V. H. P. K. H. d. W. Y. Yang, R. Boom, Recycling of composite materials, Chemical Engineering and Processing 51 (2012) 53\u201368.\n[83] J. J. M Wollschlaeger, T Sauter, The future of industrial communication, IEEE Industrial Electronics magazine (2017) 17-27.\n[84] E. K. S. N. RY Zhong, X Xu, Intelligent manufacturing in the context of industry 4.0: A review, Engineering 3 (2017) 616\u2013630.\n[85] M. Loeffler, A. Tschiesner, The internet of things and the future of manufacturing, Web interview https://www.mckinsey.com/business-\nfunctions/digital-mckinsey/our-insights/the-internet-of-things-and-the-future-of-manufacturing.\n[86] All things alexa, Amazon Webpage https://www.amazon.com/Amazon-Echo-And-Alexa-Devices/b?node=9818047011.\n[87] M. K. M. H. K.-S. K. SMR ISLAM, D KWAK, The internet of things for health care: A comprehensive survey, IEEE Access 3 (2015)\n678-708.\n[88] G. M. F Shrouf, J Ordieres, Smart factories in industry 4.0: A review of the concept and of energy management approached in production\nbased on the internet of things paradigm, 2014 IEEE International Conference on Industrial Engineering and Engineering Management,\n9-12 December, Bandar Sunway, Malaysia.\n[89] G. S. T. Stock, Opportunities of sustainable manufacturing in industry 4.0, Procedia CIRP 40 (2016) 536 - 54.\n[90] A. F. S. Y. L. JH Ang, C Goh, Energy-efficient through-life smart design, manufacturing and operation of ships in an industry 4.0 environ-\nment, Energies 10 (2017) doi:10.3390/en10050610.\n[91] J. P. Ibar, Control of polymer properties by melt vibration technology: A review, Polym Eng. Sci. 38 (1) (1998) 1-20.\n[92] J. Y. Chen, Y. Z. Chen, H. L. Li, S. Y. Lai, J. D. Jow, Physical and chemical effects of ultrasound vibration on polymer melt in extrusion.,\nUltrason Sonochem 17 (2010) 66-71.\n[93] Y. C. Zhang, H. L. Li, Functionalization of high density polyethylene with maleic anhydride in the melt state through ultrasonic initiation.,\nPolym Eng. Sci. 43 (4) (2003) 774-782.\n[94] Y. Z. Chen, H. L. Li, Effect of ultrasound on the viscoelasticity and rheology of polystyrene extruded through a slit die., J. Appl. Polym. Sci.\n100 (4) (2006) 2907-2911.\n[95] I. Isayev, S. Mandelbaum, Effect of ultrasonic waves on foam extrusion., Polym. Eng. Sci. 31 (14) (1991) 1051\u20131056.\n[96] Y. R. Cao, H. L. Li, Influence of ultrasound on the processing and structure of polypropylene during extrusion., Polym. Eng. Sci. 42 (7)\n(2002) 1534-1540.\n[97] X. Q. Liu, K. Li, M. McAfee, J. Deng, 'soft-sensor' for real-time monitoring of melt viscosity in polymer extrusion process., Proc. IEEE\nConference on Decision and Control, GA, USA (2010) 3469-3474.\n[98] Y. G. Son, Determination of shear viscosity and shear rate from pressure drop and flow rate relationship in a rectangular channel., Polymer\n48 (2007) 632-637.\n[99] M. McAfee, G. McNally, Real-time measurement of melt viscosity in single-screw extrusion., T I Meas. Control 28 (5) (2006) 481\u201397.\n[100] B. K. Nguyen, G. McNally, A. Clarke, Automatic extruder for recycled polymers with ultrasound and temperature control system., Proc.\nIMechE Part E: J Process Mechanical Engineering (2015) DOI: 10.1177/0954408914552335.\n[101] C. Abeykoon, K. Li, M. McAfee, P. J. Martin, Q. Niu, A. L. Kelly, J. Deng, A new model based approach for the prediction and optimisation\nof thermal homogeneity in single screw extrusion, Control Eng. Pract. 19 (8) (2011) 862-874.\n[102] C. Abeykoon, A novel soft sensor for real-time monitoring of die melt temperature profile in polymer extrusion, IEEE Trans. Ind. Electron.\n61 (12) (2014) 7113-7123.\n[103] C. Abeykoon, A novel model-based controller for polymer extrusion, IEEE Trans. Fuzzy Syst. 22 (6) (2014) 1413\u20131430.\n[104] Dassault systemes, simulia portfolio., http://www.3ds.com/products-services/simulia/products/.\n[105] Ansys products., http://www.ansys.com/en-GB/Products.\n[106] Comsol, multiphysics software product suite., https://uk.comsol.com/products.\n[107] Msc software, products., http://www.mscsoftware.com/products.\n[108] Esi, software solutions., https://www.esi-group.com/software-solutions.\n[109] Livermore software technology corporation, Is-dyna., http://www.lstc.com/products/ls-dyna.\n[110] O. C. Zienkiewicz, R. L. Taylor, J. Z. Zhu, The Finite Element Method: Its Basis and Fundamentals, 7th Edition, Butterworth-Heinemann,\nOxford, 2013.\n[111] D. J. Benson, Computational methods in lagrangian and eulerian hydrocodes., Computer Methods in Applied Mechanics and Engineering\n99 (2-3) (1992) 235-394.\n[112] R. F. Stellingwerf, C. A. Wingate, Impact modelling with sph., Memorie della Societa Astronomia Italiana 65 (4) (1994) 1117\u20131128.\n[113] G. R. Liu, M. B. Liu, Smoothed Particle Hydrodynamics., World Scientific Publishing Company, Singapore, 2003.\n[114] Y. Mu, G. Zhao, A. Chen, X. Wu, Numerical investigation of the thermally and flow induced crystallization behavior of semi-crystalline\npolymers by using finite elementfinite difference method., Computers & Chemical Engineering 46 (2012) 190\u2013204.\n[115] Y. Mu, G. Zhao, A. Chen, X. Wu, Modeling and simulation of polymer melts flow in the extrusion process of plastic profile with metal\n22\ninsert., The International Journal of Advanced Manufacturing Technology 67 (1) (2013) 629\u2013646.\n[116] Y. Mu, G. Zhao, X. Wu, L. Hang, H. Chu, Continuous modeling and simulation of flow-swell-crystallization behaviors of viscoelastic\npolymer melts in the hollow profile extrusion process., Applied Mathematical Modelling 39 (3-4) (2015) 1352-1368.\n[117] C. Abeykoon, P. J. Martin, K. Li, A. L. Kelly, Dynamic modelling of die melt temperature profile in polymer extrusion: Effects of process\nsettings, screw geometry and material., Applied Mathematical Modelling 38 (4) (2014) 1224\u20131236.\n[118] F. Zairi, Should Havelwith TwoDots, B. Aour, J. Gloaguen, M. Nat-Abdelaziz, J. Lefebvre, Numerical modelling of elasticviscoplastic equal\nchannel angular extrusion process of a polymer., Computational Materials Science 38 (1) (2006) 202-212.\n[119] J. Dryden, Stress concentration in polymer extrusion dies., Acta Mechanica 219 (3) (2011) 269\u2013279.\n[120] J. Liker, The Toyota Way: 14 Management Principles from the World's Greatest Manufacturer., McGraw-Hill, New York, 2003.\n[121] M. George, J. Maxey, D. Rowlands, M. Upton, The Lean Six Sigma Pocket Toolbook: A Quick Reference Guide to Nearly 100 Tools for\nImproving Quality and Speed., McGraw-Hill, New York, 2003.\n[122] M. Wagner, G. Montague, M. Tham, Neural networks for steady state modelling of an extruder., Artificial Intelligence in Engineering 11 (4)\n(1997) 375-382.\n[123] B. McKay, B. Lennox, M. Willis, G. Barton, G. Montague, Extruder modelling: a comparison of two paradigms., Control '96, UKACC\nInternational Conference on (Conf. Publ. No. 427), September 1996, DOI 10.1049/cp:19960643 2 (1996) 734\u2013739.\n[124] Z.-L. Chen, P.-Y. Chao, S.-H. Chiu, Proposal of an empirical viscosity model for quality control in the polymer extrusion process., Polymer\nTesting 22 (5) (2003) 601\u2013607.\n[125] M. McAfee, Enhancing process insight in polymer extrusion by grey box modelling., Transactions of the Institute of Measurement and\nControl 29 (5) (2007) 467-488.\n[126] M. McAfee, S. Thompson, A novel approach to dynamic modelling of polymer extrusion for improved process control., Proceedings of the\nInstitution of Mechanical Engineers, Part I: Journal of Systems and Control Engineering 221 (4) (2007) 617-628.\n[127] X. Liu, K. Li, M. McAfee, B. K. Nguyen, G. M. McNally, Dynamic gray-box modeling for on-line monitoring of polymer extrusion\nviscosity., Polymer Engineering & Science 52 (6) (2012) 1332\u20131341.\n[128] J. Deng, K. Li, E. Harkin-Jones, M. Price, M. Fei, A. Kelly, J. Vera-Sorroche, P. Coates, E. Brown, Low-cost process monitoring for polymer\nextrusion., Transactions of the Institute of Measurement and Control 36 (3) (2014) 382-390.\n[129] E. G. Carrano, D. G. Coelho, A. Gaspar-Cunha, E. F. Wanner, R. H. Takahashi, Feedback-control operators for improved pareto-set descrip-\ntion: Application to a polymer extrusion process., Engineering Applications of Artificial Intelligence 38 (2015) 147\u2013167.\n[130] P. G. Drazin, Nonlinear systems., Cambridge University Press, Cambridge, 1992.\n[131] G. McKinley, W. Raiford, R. Brown, R. C. Armstrong, Nonlinear dynamics of viscoelastic flow in axisymmetric abrupt contractions., Journal\nof Fluid Mechanics 223 (3-4) (1991) 411-456.\n[132] M. D. Graham, Wall slip and the nonlinear dynamics of large amplitude oscillatory shear flows., Journal of Rheology 39 (4) (1995) 697-712.\n[133] D. E. Smith, D. A. Tortorelli, C. L. TuckerIII, Optimal design for polymer extrusion. part i: Sensitivity analysis for nonlinear steady-state\nsystems., Computer Methods in Applied Mechanics and Engineering 167 (3-4) (1998) 283\u2013302.\n[134] D. E. Smith, D. A. Tortorelli, C. L. TuckerIII, Optimal design for polymer extrusion. part ii: Sensitivity analysis for weakly-coupled\nnonlinear steady-state systems., Computer Methods in Applied Mechanics and Engineering 167 (3-4) (1998) 303\u2013323.\n[135] A. J. McMillan, A non-linear friction model for self-excited vibrations., Journal of Sound and Vibration 205 (3) (1997) 323\u2013335.\n[136] J. C. Pirkle(Jr.), R. D. Braatz, Instabilities and multiplicities in non-isothermal blown film extrusion including the effects of crystallization.,\nJournal of Process Control 21 (3) (2011) 405-414.\n23\n"}, "expected_output": {"claims": [{"unit": "kWh/kg", "value": 0.0822, "evidence": ["the specific energy consumption (SEC) of the extruder motor should be in the range of 0.0822 to 0.1644 kW.hr/kg.", "enhancements to machinery could be identified, where an economic case could be made on the basis of energy cost\nsaving. Some examples are given below.\nIn the late 1970s, Chung et al. [12] found that for a 63.5 mm diameter extruder mechanical energy efficiency of\n62% was typical, and for larger extruders the energy efficiency was lower. In 1981, Kruder and Nunn [29] claimed\nthat energy efficiency of extruders can range from 45%-75%. It was noted that the energy efficiency depended on the\ntransmission mechanism, screw design, product geometry, nature of polymer feedstock and the production rate, while\nthe major energy losses of an extruder occur as a result of the forced cooling process step, and the losses associated\nwith the drive and transmission unit. At low screw speeds, barrel heaters consume a considerably higher portion of\nenergy than at higher speeds, and significant energy savings could be made by running the processes at the highest\npossible power factor. Additionally, this work presented information on energy demand and losses of each individual\ncomponent of an extruder.\nSubsequently through the 1980s, most research into energy efficiency was focussed on the screw efficiency and\nmass flow rate. A reduction in the overall power requirement for an extruder can be achieved through the use of a\ngeared pump at the end of the extruder to increase the mass flow rate (McKelvey [30]). In 1985, Strauch et al. [31]\ncarried out an energy consumption study on a 63.5 mm diameter single screw extruder, and observed that most of the\nenergy was consumed by the mechanical parts, with less significant levels of consumption in process heating. The\nenergy conversion was then assessed and it was found that heating the water in the cooling system accounted for more\nthan half of the energy supplied.\nDuring the late 1980s and 1990s, the manufacturing sector was making many changes, with a view to improving\nproductivity and quality. Driven by the advances made in Japanese manufacturing, the main focus during that time\nwas on management methods, such as Total Quality Management, LEAN, and Six Sigma. These efforts initially\naddressed cost and time issues, where the biggest economic benefits were to be found. Latterly, interest in energy\nefficiency began to be seen as not only cost reduction opportunity but also as an environmental imperative.\nIn the context of power consumption in the extruder, in 1997 Anderson et al. [32] recognised that for the processing\nof most plastics, from room temperature, the specific energy consumption (SEC) of the extruder motor should be in\nthe\nrange of 0.0822 to 0.1644 kW.hr/kg.\nAt around the same time, a study by Falkner in 1997 [33] showed that motor operations accounted for over 65%\nof the 1994 UK industrial electricity usage. Asserting that more than 10% of this energy could be attributed to\ninefficiency, Falkner argued that this represented a loss of about \u00a30.5 billion to the annual UK economy. These values\naccounted for motor energy utilisation across multiple industrial sectors, but it should be recognised that the electric\nmotors in plastics industry processing machines are a major power consumers.\nA more detailed study by Rosato et al. in 2001 [34] observed that energy losses of between 3 and 20% can arise in\nthe transmissions and control systems. Despite this, a conclusion was made that because plastics have lower specific\nenergy requirements compared with most conventional raw materials, they are still highly competitive.\nFive years later, Womer et al. [35] considered the energy efficiency of extruder cooling. The results demonstrated\nthat water cooling systems consume more energy compared with air cooling, irrespective of the particular plastic being\nprocessed. As a result, a recommendation was made to use air only cooling unless extensive cooling was expressly\nrequired.\nIn 2010 [36], the plastics industry was recognised to be one of the major UK industries with a similar trend\napplying globally. On that basis any improvement in process energy efficiency would lead to a considerable reduction\nin global energy requirement. Also in 2010, Cantor [37] presented measurements of SEC, where the impact of the\nmotor and of each individual heater zone, with respect to the overall specific energy consumption, was separately\nrecorded. It was observed that the heaters account for over 95% of the supplied energy. In a slightly later study\nby Heur and Verheijen [38], the authors studied differences from one plant to another, and recommended the use of\nfrequency controllers to enable more precise process control.\nThe earliest mention of an Industry 4.0 implementation to energy efficiency control was by Jing et al. (2014) [39]\nwhich proposed the use of real-time monitoring. The rationale was to render unnecessary the installation of power\nmeters or the development of data-driven models. A fuzzy logic controller controlled the high melt quality in a single\nscrew extruder, and was shown to be a cheaper alternative to using a gear pump. This also paved the way for achieving\ngreater extruder energy efficiency by optimising the temperature settings.\nA number of other works [35, 40, 41, 42, 43, 44, 45] consider the drive motor efficiency compared with other\ndevices. The conclusion to be drawn is that the drive motor should be the primary design consideration for process\n6\n", "in 1997 Anderson et al. [32] recognised that for the processing of most plastics, from room temperature, the specific energy consumption (SEC) of the extruder motor should be in the range of 0.0822 to 0.1644 kW.hr/kg."]}, {"unit": "kWh/kg", "value": 0.1644, "evidence": ["the specific energy consumption (SEC) of the extruder motor should be in the range of 0.0822 to 0.1644 kW.hr/kg.", "enhancements to machinery could be identified, where an economic case could be made on the basis of energy cost\nsaving. Some examples are given below.\nIn the late 1970s, Chung et al. [12] found that for a 63.5 mm diameter extruder mechanical energy efficiency of\n62% was typical, and for larger extruders the energy efficiency was lower. In 1981, Kruder and Nunn [29] claimed\nthat energy efficiency of extruders can range from 45%-75%. It was noted that the energy efficiency depended on the\ntransmission mechanism, screw design, product geometry, nature of polymer feedstock and the production rate, while\nthe major energy losses of an extruder occur as a result of the forced cooling process step, and the losses associated\nwith the drive and transmission unit. At low screw speeds, barrel heaters consume a considerably higher portion of\nenergy than at higher speeds, and significant energy savings could be made by running the processes at the highest\npossible power factor. Additionally, this work presented information on energy demand and losses of each individual\ncomponent of an extruder.\nSubsequently through the 1980s, most research into energy efficiency was focussed on the screw efficiency and\nmass flow rate. A reduction in the overall power requirement for an extruder can be achieved through the use of a\ngeared pump at the end of the extruder to increase the mass flow rate (McKelvey [30]). In 1985, Strauch et al. [31]\ncarried out an energy consumption study on a 63.5 mm diameter single screw extruder, and observed that most of the\nenergy was consumed by the mechanical parts, with less significant levels of consumption in process heating. The\nenergy conversion was then assessed and it was found that heating the water in the cooling system accounted for more\nthan half of the energy supplied.\nDuring the late 1980s and 1990s, the manufacturing sector was making many changes, with a view to improving\nproductivity and quality. Driven by the advances made in Japanese manufacturing, the main focus during that time\nwas on management methods, such as Total Quality Management, LEAN, and Six Sigma. These efforts initially\naddressed cost and time issues, where the biggest economic benefits were to be found. Latterly, interest in energy\nefficiency began to be seen as not only cost reduction opportunity but also as an environmental imperative.\nIn the context of power consumption in the extruder, in 1997 Anderson et al. [32] recognised that for the processing\nof most plastics, from room temperature, the specific energy consumption (SEC) of the extruder motor should be in\nthe\nrange of 0.0822 to 0.1644 kW.hr/kg.\nAt around the same time, a study by Falkner in 1997 [33] showed that motor operations accounted for over 65%\nof the 1994 UK industrial electricity usage. Asserting that more than 10% of this energy could be attributed to\ninefficiency, Falkner argued that this represented a loss of about \u00a30.5 billion to the annual UK economy. These values\naccounted for motor energy utilisation across multiple industrial sectors, but it should be recognised that the electric\nmotors in plastics industry processing machines are a major power consumers.\nA more detailed study by Rosato et al. in 2001 [34] observed that energy losses of between 3 and 20% can arise in\nthe transmissions and control systems. Despite this, a conclusion was made that because plastics have lower specific\nenergy requirements compared with most conventional raw materials, they are still highly competitive.\nFive years later, Womer et al. [35] considered the energy efficiency of extruder cooling. The results demonstrated\nthat water cooling systems consume more energy compared with air cooling, irrespective of the particular plastic being\nprocessed. As a result, a recommendation was made to use air only cooling unless extensive cooling was expressly\nrequired.\nIn 2010 [36], the plastics industry was recognised to be one of the major UK industries with a similar trend\napplying globally. On that basis any improvement in process energy efficiency would lead to a considerable reduction\nin global energy requirement. Also in 2010, Cantor [37] presented measurements of SEC, where the impact of the\nmotor and of each individual heater zone, with respect to the overall specific energy consumption, was separately\nrecorded. It was observed that the heaters account for over 95% of the supplied energy. In a slightly later study\nby Heur and Verheijen [38], the authors studied differences from one plant to another, and recommended the use of\nfrequency controllers to enable more precise process control.\nThe earliest mention of an Industry 4.0 implementation to energy efficiency control was by Jing et al. (2014) [39]\nwhich proposed the use of real-time monitoring. The rationale was to render unnecessary the installation of power\nmeters or the development of data-driven models. A fuzzy logic controller controlled the high melt quality in a single\nscrew extruder, and was shown to be a cheaper alternative to using a gear pump. This also paved the way for achieving\ngreater extruder energy efficiency by optimising the temperature settings.\nA number of other works [35, 40, 41, 42, 43, 44, 45] consider the drive motor efficiency compared with other\ndevices. The conclusion to be drawn is that the drive motor should be the primary design consideration for process\n6\n", "in 1997 Anderson et al. [32] recognised that for the processing of most plastics, from room temperature, the specific energy consumption (SEC) of the extruder motor should be in the range of 0.0822 to 0.1644 kW.hr/kg."]}]}, "metadata": {"product_category": "Machinery & equipment", "request_id": "req_2ced6848ba546c28"}}
{"id": "1065e3bc44a627972aa9b1bf", "input": {"query": "What is the motor-specific energy consumption for extruders in kWh per kg? What is the overall extruder efficiency and total energy consumption per kg of polymer processed?", "source_url": "https://ajmcmillan.co.uk/AcademicPublications/EnergyInExtrusion_AcceptedVersion.pdf", "document_text": "Energy efficiency in extrusion-related polymer processing: a review of state of\nthe art and potential efficiency improvements\nChamil Abeykoon\u00aa,*, Alison McMillan, Bao Kha Nguyen\n\"North West Composites Centre and Aerospace Research Institute, Department of Materials, Faculty of Science and Engineering, University of\nManchester, Oxford Road, Manchester, M13 9PL, UK\nb Faculty of Arts, Science and Technology, Wrexham Glyndwr University, Wrexham, LL11 2AW, UK\n\"School of Engineering and Informatics, University of Sussex, Brighton, BN1 9QT, UK\nAbstract\nEnergy saving and industrial pollution have become increasingly important issues, therefore the identification and\nadoption of more energy efficient machines and industrial processes are now industrial priorities, and worthy topics\nfor further development through academic research. Polymeric materials are a major raw material, finding widespread\napplication to a range of current industrial machine components as well as multiple products and packaging found\nin our daily life. Polymer extrusion serves as a particular example of polymer processing techniques, representative\nof others in as much as there are analogous intermediate stages in the processing. Processing techniques which re-\nquire such intermediate stages include the manufacture of blown film, blow moulding, thermo-forming, and injection\nmoulding. Hence, the study of polymer extrusion is a representative paradigm for a wider range of processing tech-\nniques. Since polymer processing is an energy intensive process and accounts for a huge share (maybe more than 1/3)\nof the materials processing sector, any improvement to the process would contribute significantly to global energy\nsavings. This work presents a review of studies, which focus on, or appertain to, the energy consumption of extrusion\nrelated polymer processing applications. Typical energy demand and losses during processing are considered, and\npossible approaches for improving the process energy efficiency while maintaining the required end product quality\nare considered. Overall, this work provides a detailed discussion about how and where energy is utilized; how, where\nand why energy losses occur; and sets out approaches for optimizing the process energy efficiency.\nKeywords:\nEnergy consumption, Energy losses, Energy savings, Polymer extrusion, Process monitoring, Process control,\nMaterials processing, Energy efficiency, Industry 4.0, Circular economy, Dynamical systems\n1. Introduction\n1.1. Market demand for polymers\nAs the number of applications for polymer materials in high volume manufacturing sectors, such as packaging, con-\ntinues to grow, it is timely to consider manufacturing process optimisation from the energy efficiency point of view.\nAt the present time, the increasing adoption of thermoplastics for use in high performance component applications,\nsuch as automotive and aerospace, has meant that product quality has been the prime focus of process optimisation.\nAs the manufacturing processes associated with polymer processing have become more mature, there has been a cor-\nrespondingly greater utilisation of in-line sensors, and adoption of Industry 4.0 protocols. This has enabled greater\nunderstanding of the material performance under processing temperatures and pressures, thereby providing the neces-\nsary input data needed for high fidelity computational modelling. In the field of computational optimisation, there have\nbeen signficant advances in algorithms development, with the result that a much bigger class of multi-variable and\n*Corresponding author. +441613062540\nEmail address: chamil. abeykoon@manchester.ac.uk (Chamil Abeykoon)\nPreprint submitted to Elsevier\nMay 20, 2021\nmulti-objective problems can be addressed. As a result, the possibility to broaden the scope of process optimisation\ncan now be grasped.\nClearly, for both high volume and high performance applications, energy efficient manufacture is a desirable\ngoal, which not only leads to reduced manufacturing costs but also addresses National and International energy and\nCO2 reduction targets. As a result, scrutiny of the energy required in an energy intensive manufacturing process\nsuch as the extrusion process is driven by both business and environmental imperatives. The payment of energy\nbills for unnecessary usage reduces profit margins and hence increases the end product/service prices for customers.\nMeanwhile, because CO2 emissions are now very clearly understood to be detrimental to the environment, energy\nusage will be increasingly subject to disincentives such as high fuel commodity pricing and taxation.\nThe scale of the plastics industry is internationally huge, and expanding. For example, in 2015 in the UK [1], there\nwere of the order of 6,200 plastics companies, employing nearly 170,000 people, and with a combined annual sales\nturnover of over \u00a323.5 bn, of which one third represented exports. According to the reports of PlasticsEurope [2], by\nthe year 2016 the European plastics industry comprised of more than 60,000 companies, employing more than 1.5\nmillion people, and with total sales exceeding 350 EUR bn. Globally, plastics production has grown from 204 to 335\nmillion tonnes between 2002 and 2016. The statistics presented in Figures 1 and 2 illustrate this growing demand.\nMillion tonnes\n350\n300\n250\n200\n150\n100\nPE\nPP\nPVC\nPS-EPS\nABS-SAN\n50\n01\n2005\n2011\n2012\n2013\n2017\n2020\n2025\nFigure 1: Major thermoplastics: World demand distribution, by polymer between years 2005-2025 [3]\n%/year\n6\n2005-2012\n2012-2017\n5\n4\n3\n2\n1\n0\nPE\nPP\nPVC\nPS-EPS\nABS-SAN\nTOTAL WORLD'\nFigure 2: Major thermoplastics: World consumption growth rate, by polymer (2005-2012 and 2012-2017)[3]\nGiven this level is sustained, the level of growth in demand, and the development and accessibility of new poly-\nmer processing capability, it is clear that improvements in process energy efficiency could have a significant impact\non global energy savings [4, 5]. Furthermore, the European Best Practice Guide [6] claims, \u201cPlastics are the material\nfor the 21st century\u201d, explaining that a 3 Megatonne CO2 emission reduction could be achieved in Europe by a 10%\nreduction in the plastics industry energy consumption. With the current capacity of polymers and plastics manufac-\nturing sector, it is one of the largest energy consumers in industrial manufacturing and also a major source of global\nwaste generation. Meantime, the energy savings/optimization in the manufacturing sector is considered as one of the\nmain pillars of modern circular economy concept and both manufactures and consumers have been forced to re-think\nthe current take-make-waste extractive industrial model for reusing materials form end-of-life components/devices,\nwhere polymers/plastics industry is one of the major focuses of this concept [7].\n1.2. The polymer extrusion process\nA polymer \"extruder\u201d machine processes materials by forcing them through a set of processing stages. The\noperation and basic processing stages are described in Figure 3 below. The screw passes material through a cylindrical\nHopper\nBarrel\nBand type\nheaters\nControl unit\nDie\nGear box\nDrive\nmotor\nCooling fans\nScrew\nSolids conveying\nMelting Melt conveying \u00a6\nFigure 3: Operational schematic of a single screw extruder\nbarrel, around which heaters are wrapped, to provide the necessary heat for material melting. In addition to this\nexternally provided heat, a significant amount of heat is generated internally, inside the barrel, as a result of the\nmechanical work of the screw (i.e. the work done against viscous and frictional forces). The feed material absorbs\nheat as it is conveyed along the screw and is expected to be in the fully molten state at the point that the molten\nmaterial is forced into a die to form into the desired shape.\nCurrently, different types of extruders (e.g. single screw, multi screw, and disc/drum types) are available in in-\ndustry, while screws with different geometrical designs are commercially available. Moreover, a number of process\nmonitoring devices are used, to observe process functionality, and for diagnosing possible processing problems.\nFrankland [8], President of Frankland Plastics Consulting, LLC, explains this in detail in his on-line article about\nestimating extrusion melt temperature. The most significant points are that the mechanical energy feed into the drive\nis converted by the screw action on the material to create heat, and thus melting of the polymer. The energy share\nrequired for material conveying is relatively smaller, as is the energy supplied to the barrel heaters. He also lists energy\nlosses and their sources. More details on the polymer extrusion process and its operation can be found in the literature\n[9, 10, 11].\nManufacturing process stability is a key concern, and variation in the material temperature presents a challenge to\nthe end product quality control. For this reason most commercial polymer producers avoid operating their extruders at\nat the higher screw speeds. This is unfortunate since at higher speeds, and therefore at higher workpiece temperatures,\nthere is more potential for process energy efficiency improvement because the material viscosity is reduced and thus\nthe forming forces required are lower. Moreover, this undesirable cost is repeated, since many thermoplastic polymers\nare extruded more than once before their final products are manufactured [12]. Better concatenation of extrusion\nprocess steps would lead to greater energy reduction, by maintaining or controlling the heat in the workpiece during\nprocessing, thereby avoiding the need to re-heat.\n1.2.1. Basic processing mechanisms\nZones within the polymer processing screw can be broadly designated, based on the functional activity taking place\nwithin that zone, see Figure 3. The points of transition between zones are not generally well defined, as they depend\non the processing conditions and the materials.\na. Solids conveying\nIn this zone, the polymer is preheated before passing into the subsequent zones. While flowing along this zone,\nmaterial starts to absorb heat from the barrel heaters, but the mechanical heat generated by frictional and viscous\n3\nmechanisms is dominant in this zone [13, 14, 10, 15]. Generally, the screw channel depth is maintained constant in\norder to provide a constant material feed to the subsequent zones.\nThe first comprehensive theory for the action of solids conveying was developed by Darnell and Mol [16] in the\n1950s and this quantitative description still remains as the widely accepted model for solids conveying in extrusion.\nb. Melting or Plastication\nExperiments for studying the polymer extrusion melting mechanism were first carried out by Maddock and Street\nin 1959 [17]. The melting mechanism proposed by Maddock for single screw extruders still remains as the most\nwidely accepted melting mechanism in polymer extrusion. The Maddock melting mechanism is only a qualitative\ndescription of melting which occurs in single screw extruders. Maddock used a visual inspection method to investigate\nthe melting process by stopping the screw rotation suddenly during the process and 'freezing' the polymer by cooling\nthe barrel and screw rapidly. Later, Tadmor also extended the understanding of melting mechanism of extrusion\nprocesses [18, 19, 20, 21].\nAs the material reaches the \"end\" of the solids conveying zone, it begins to melt, and as such is considered to\nhave entered into the melting zone. As the material becomes soft, further heat will be added to the process by means\nof viscous dissipation of the material (i.e. work done against the viscoelastic nature of the material). Both solid\nand molten polymers co-exist in this zone. The solid bed would comprise both compacted solid polymer abutting\nthe \"trailing flight\u201d, and the melt pool pushing against the \u201cpushing flight\", as shown in Figure 4. As the material\nFlow direction\nSolid/Melt\ninterface\nPushing\nflight\nCirculating\nmelt pool\nExtruder barrel\nTrailing\nSolid bed\nflight\nScrew\nFigure 4: An illustration of the typical arrangement of the solid bed and melt pool inside a screw channel for a single-flighted conventional screw\nproceeds along the screw, the proportion of melt pool to solid bed increases. The screw channel depth is therefore\ndesigned to become smaller, which influences flow rate and mixing; however, the actual screw length at which melting\noccurs depends on a range of parameters such as screw geometry, operating conditions and physical properties of the\npolymer [20].\nAs was claimed by Severs [22], the plastication or melting process has a direct impact on the quality of the material\nproperties of final product, and thus must be carefully controlled. Tadmor, Klein and Gogos [18, 19, 21] proposed an\nequation for calculating the rate of melting, (Q), in a screw channel, and is given by Eq. (1).\nQ2 =\n-\n[Pm \u00d7 Vbx { km (Tb \u2212 Tmelt) +\u014b\n2 {Cp (Tm-Ts) + 1}\n11/2\n(1)\nWhere Pm is the melt density, Vbx is the transverse component of the barrel velocity, km is the thermal conductivity\nof the molten material, Tmelt is the melt temperature, T is the barrel temperature, \u014b is the melt viscosity, V; is the\nresultant relative velocity, Cp is the polymer specific heat capacity, and \u03bb is the temperature of the solid bed.\nThis equation clearly demonstrates that the melting rate can be increased by increasing the screw rotational speed\n[14]; however, for higher speeds, the polymer passes through more quickly, giving less time for temperature stabilisa-\ntion, and thus more variation in melt viscosity [10, 14]. As a result, to ensure controlled plastication, it is necessary to\ncontrol the melting rate, and this in turn depends on the material being proceeded, process set conditions, and nature\nof the processing unit/machine [23].\nc. Melt conveying\nMelt conveying starts as complete melting is achieved. The screw channel depth is constant along the zone and\nis shallower than in the other two zones. During this stage further heating and mixing of the melt takes place as the\npolymer is smeared by the tip of the screw flight against the barrel wall. Material has to be moved towards the die\nwith enough force to overcome the head pressure generated at the die - this is known as the \u201cdie head pressure\".\nMelt output rate from this zone depends on a combination of two main factors: the rate of the rotation of the screw\nand the screw channel pressure gradient [24]. Proper mixing of material is another requirement for the flow through\nthis zone. The melt conveying zone of some of the new screw designs is fitted with efficient mixer units to ensure\ngood mixing performance (e.g. the barrier flighted screw with a Maddock mixer).\nStudies on melt conveying operation of extrusion were reported very much earlier than in the other two zones.\nOne of the initial studies was carried out in 1920s [25, 26], which proposed the calculation of the melt conveying rate\nby considering the melt flow as a laminar fully developed flow. This is still a widely accepted model.\nIn addition to the above mentioned mechanism/theories, several other works have been reported later on improving\nthe understanding of these three main mechanisms and more details can be found in the literature [14, 9, 10, 11, 27].\n2. Energy required for materials processing\nThe assessment of energy requirements is not straight forward. The overall extrusion process can be broken down\ninto smaller activities, but even then, the power demands at each stage depend in a complex way on a large number of\nprocessing parameters. A useful energy flow model was developed by Severs [22], as presented in Figure 5.\nEquipment cooling\nLosses\nit\nCooling\n\u2191\nPolymer\nsolid\n\u2192 Melting\n\u2192 Forming\nSolidification\nPolymer\nproduct\nMotor power (mechanical energy)\nHeating system (thermal energy)\nElectric power\nFigure 5: Typical energy flow diagram for an extrusion process\nThe energy, Eu, used by an extruder for useful work in material melting and forming, [28], is given by Eq. (2):\n=\nEu Ein Elosses\n(2)\nwhere Ein is the energy input to the extruder and Elosses is the energy expended that does not contribute to the extrusion\nprocess. Thus, the energy efficiency can be given by Eq. (3):\nnextruder =\nEin - Elosses\nEin\n\u00d7 100%\n(3)\nIn these equations, the energy inputs (Ein) should be related to the energy consumed by the electrical components\nsuch as drive motor, barrel/die heaters, barrel/motor cooling fans, water pump/s, instrumentation in the control unit,\netc.. The energy losses are always associated with all the components and also occur due to forced cooling and via\nnatural convection and radiation, which can be accounted under Elosses. In general, the drive motor and the barrel and\ndie heaters are the source of the highest energy losses. In typical polymer extrusion processes, recovery of such lost\nenergy is impractical, as this is largely released as heat energy to water or air. More details concerning the energy\nrequired for polymer processing and the thermodynamic efficiency of an extruder have been discussed by the authors\npreviously [28].\n3. Prior art in extruder energy evaluation, monitoring and modelling\n3.1. Energy Consumption studies\nIn considering the energy consumption in any industrial process, the first step is to review the process capability\nof the existing or available plant machinery, and the power consumption. On that basis, potential modifications or\n5\nenhancements to machinery could be identified, where an economic case could be made on the basis of energy cost\nsaving. Some examples are given below.\nIn the late 1970s, Chung et al. [12] found that for a 63.5 mm diameter extruder mechanical energy efficiency of\n62% was typical, and for larger extruders the energy efficiency was lower. In 1981, Kruder and Nunn [29] claimed\nthat energy efficiency of extruders can range from 45%-75%. It was noted that the energy efficiency depended on the\ntransmission mechanism, screw design, product geometry, nature of polymer feedstock and the production rate, while\nthe major energy losses of an extruder occur as a result of the forced cooling process step, and the losses associated\nwith the drive and transmission unit. At low screw speeds, barrel heaters consume a considerably higher portion of\nenergy than at higher speeds, and significant energy savings could be made by running the processes at the highest\npossible power factor. Additionally, this work presented information on energy demand and losses of each individual\ncomponent of an extruder.\nSubsequently through the 1980s, most research into energy efficiency was focussed on the screw efficiency and\nmass flow rate. A reduction in the overall power requirement for an extruder can be achieved through the use of a\ngeared pump at the end of the extruder to increase the mass flow rate (McKelvey [30]). In 1985, Strauch et al. [31]\ncarried out an energy consumption study on a 63.5 mm diameter single screw extruder, and observed that most of the\nenergy was consumed by the mechanical parts, with less significant levels of consumption in process heating. The\nenergy conversion was then assessed and it was found that heating the water in the cooling system accounted for more\nthan half of the energy supplied.\nDuring the late 1980s and 1990s, the manufacturing sector was making many changes, with a view to improving\nproductivity and quality. Driven by the advances made in Japanese manufacturing, the main focus during that time\nwas on management methods, such as Total Quality Management, LEAN, and Six Sigma. These efforts initially\naddressed cost and time issues, where the biggest economic benefits were to be found. Latterly, interest in energy\nefficiency began to be seen as not only cost reduction opportunity but also as an environmental imperative.\nIn the context of power consumption in the extruder, in 1997 Anderson et al. [32] recognised that for the processing\nof most plastics, from room temperature, the specific energy consumption (SEC) of the extruder motor should be in\nthe\nrange of 0.0822 to 0.1644 kW.hr/kg.\nAt around the same time, a study by Falkner in 1997 [33] showed that motor operations accounted for over 65%\nof the 1994 UK industrial electricity usage. Asserting that more than 10% of this energy could be attributed to\ninefficiency, Falkner argued that this represented a loss of about \u00a30.5 billion to the annual UK economy. These values\naccounted for motor energy utilisation across multiple industrial sectors, but it should be recognised that the electric\nmotors in plastics industry processing machines are a major power consumers.\nA more detailed study by Rosato et al. in 2001 [34] observed that energy losses of between 3 and 20% can arise in\nthe transmissions and control systems. Despite this, a conclusion was made that because plastics have lower specific\nenergy requirements compared with most conventional raw materials, they are still highly competitive.\nFive years later, Womer et al. [35] considered the energy efficiency of extruder cooling. The results demonstrated\nthat water cooling systems consume more energy compared with air cooling, irrespective of the particular plastic being\nprocessed. As a result, a recommendation was made to use air only cooling unless extensive cooling was expressly\nrequired.\nIn 2010 [36], the plastics industry was recognised to be one of the major UK industries with a similar trend\napplying globally. On that basis any improvement in process energy efficiency would lead to a considerable reduction\nin global energy requirement. Also in 2010, Cantor [37] presented measurements of SEC, where the impact of the\nmotor and of each individual heater zone, with respect to the overall specific energy consumption, was separately\nrecorded. It was observed that the heaters account for over 95% of the supplied energy. In a slightly later study\nby Heur and Verheijen [38], the authors studied differences from one plant to another, and recommended the use of\nfrequency controllers to enable more precise process control.\nThe earliest mention of an Industry 4.0 implementation to energy efficiency control was by Jing et al. (2014) [39]\nwhich proposed the use of real-time monitoring. The rationale was to render unnecessary the installation of power\nmeters or the development of data-driven models. A fuzzy logic controller controlled the high melt quality in a single\nscrew extruder, and was shown to be a cheaper alternative to using a gear pump. This also paved the way for achieving\ngreater extruder energy efficiency by optimising the temperature settings.\nA number of other works [35, 40, 41, 42, 43, 44, 45] consider the drive motor efficiency compared with other\ndevices. The conclusion to be drawn is that the drive motor should be the primary design consideration for process\n6\nengineering the energy efficiency of the whole extrusion plant.\n3.2. Influence of process set parameters\nIn 2001, Rauwendaal [10] recognised the significance of process settings, and presented an account of a procedure\nto minimise power consumption. A little later, in 2003, Rasid and Wood [46] investigated the influence of individual\nbarrel zone temperatures and found that the solids conveying zone temperature had the greatest influence on overall\npower consumption.\nStudies carried out between 2004 and 2012, [47, 48, 49, 50], examined various process parameters and their\ninfluence on SEC. In addition to noting the effect of material viscosity, variation in energy consumption was also seen\nfor different designs of screw, and there were greater melt temperature fluctuations at higher screw speeds: another\nexample of the ever-present tension between cost and quality. The simultaneous need to achieve both energy efficient\noperation and finished part quality remains a challenge.\nStudies carried out by Abeykoon et al. [51, 28, 52, 53] between 2009 and 2016, focussed on the relationship\nbetween the process energy demands of the motor and barrel heating and melt thermal stability. The effects of the\nsettings for these processes, the screw geometry and choice of material were explored.\n3.3. Modelling\nFollowing a thorough trawl of the published scientific literature, it has become clear that relatively little work has been\nundertaken to model extruder energy consumption.\nThe earliest work in this area was by Mallouk and Mckelvey [54] in 1953, where a mathematical equation was\ndeveloped, based on assumptions of isothermal, Newtonian flow, in a screw channel with constant section. In 1996,\nWilczynski [55] developed a computer model where the five zones of the extruder plus the die were considered\nseparately. Subsequently, in 2000, Lai and Yu [56] also proposed a mathematical model for the calculation of energy\nconsumption based on screw speed, and including viscosity. In Abyekoon et al.'s [57, 52] studies of a single screw\nextruder, the data collected was analysed using static nonlinear polynomial models. The conclusion of the analysis\nwas that choosing energy efficient process settings would also lead to thermal stability.\nObviously, the availability of advanced modelling methods for predicting energy consumption, based on process\nparameters, would enable process operators to select optimum operating conditions. In particular, models which\nincorporate both energy consumption and melt thermal quality would be preferred but the development of such models\nis quite challenging. Melt thermal quality and energy efficiency present opposite behaviours with respect to the\nprocessing speed: the thermal quality deteriorates while the energy efficiency improves. Since the industrial sector\nhas to meet strict environmental regulations to minimize the carbon footprint, any reduction in the energy demands\nfor polymer processing would support future sustainability.\n3.4. General considerations in energy usage\nAccording to basic electricity principles, the typical power consumption of a DC and an AC device (PDC and PAC) is\ngiven by equations (4) and (5), respectively [58, 59],\nPDC = V XI\nPAC = VXIX cos\n(4)\n(5)\nwith I being the supply current, V voltage, and cos & the \"displacement power factor\u201d. From these, the power demand\nof any device in an extrusion plant can be evaluated; however, the energy losses related to each device might vary\nfrom component to component.\nThe power factor is an important consideration in the assessment of the energy usage of an electrical machine or\nprocess, and is defined as either the \"displacement power factor\" which is the cos o in Eq. (5) or the \"true power\nfactor\" which is given by Eq. (6).\nTrue\npower factor\n=\nTrue (or active) power\n(6)\n7\nApparent power\nImpedance\nApparent power (S)\n(units: VA)\nphase angle\n(units: VAR)\nReactive power (Q)\nActive power (P)\n(units: W)\nFigure 6: Power triangle showing the relationship between active, apparent and reactive powers\nThe true power factor lies in the range 0 \u2013 1, for which the running of the machine or process with true power factor\nequal to one would be the best possible energy efficient operating condition (when the impedance phase angle shown in\nFigure 6 is equal to zero). For a true power factor of less than one, the energy supplied to the load is not used optimally.\nIn such a case, a higher current must be drawn to compensate for the phase shift, o. Where industrial customers operate\nwith power factors below around 0.95, [60], this represents unbalanced additional power demands from the power\nsupplier, and hence additional infrastructure demand leading to additional costs. Furthermore, electrical devices are\nattributed with a 1\u00b2\u00d7R heat loss (R is the electrical resistance), so that increasing the required current while reducing\nthe power factor results in an increase of power loss as heat.\nMeasurements of power factor and total power consumption, for a DC motor driven 63.5 mm in diameter single\nscrew extruder operated at different screw speeds, are shown in Figure 7.\nPower factor\n30\n0.2\n+\n0.8\n0.6\nTotal power (kW)\nSS (rpm)\n\u00a6(a)\n0\n35\n\u00a6(b)\n30\n20\n10\n100\n(c)\n80\n60\n40\n20\n0\n0\n50\n150\n250\n320\nTime (s)\nFigure 7: (a). Power factor, (b). Total extruder power, (c). Screw speed [52]\nFrom this, it can be seen that both the power factor and the total power required are greater for greater processing\nspeeds; however, for higher processing speeds, the temperature uniformity of the process melt output deteriorates\nsignificantly, leading to poor product quality, [28, 52, 53]. Hence, the running of these processes at higher speeds and\n8\nwith the highest possible power factor is problematic, despite being desirable for energy efficiency.\n4. Potential for energy efficiency improvements\nEnergy demands and losses are illustrated in the form of an energy flow diagram, Figure 8. This diagram may be\nEnergy content in\nthe feed material\nDrive motor\nEnergy used\nfor material\nmelting and\nExternal\nheating/cooling\nOther losses\nEnergy for\nother auxiliary\ndevices\nForced\nDrive motor\nlosses\nTransmission cooling\nlosses (gear\nlosses\nbox)\nNatural\ncooling\nlosses\nforming\nFigure 8: A typical energy flow diagram for an extruder [52]\nextended for any auxiliary devices connected with the plant.\n4.1. Drive motor and gear box\nThe key component of any extrusion machine is the screw, which can be driven by a controllable direct current\n(DC) or an alternating current (AC) motor, or indeed by a hydraulic drive [10, 61]. The screw and the motor are\nconnected through a gear box with fixed or adjustable transmission ratio, as shown in Figure 3. For the case of\nan extruder with a DC motor drive, see the schematic, presented in Figure 9. Additionally, the machine can have\nsensing and control devices related to its operation, for example, PID temperature controllers to control set barrel/die\ntemperatures. Extruders with AC motor drives are essentially similar (with no rectifier), and may have additional\ncomponents depending on the type of the motor.\n(@set\nDactual)\nSet screw speed\n(@set)\n+\nPID Motor\nspeed controller\nActual screw\nspeed (actual)\nDC motor\nGear box\nScrew\nArmature voltage (Va)\nchanges to adjust the\nmotor speed\n@actual\nTachometer\ngenerator\nFigure 9: A schematic of an extruder drive mechanism\nFigure 10 shows measured motor power and total power consumptions, for the case of a 63.5 mm diameter single\nscrew extruder with a DC motor, driven at different screw speeds. The contribution of heaters to the total power\ndemand is also indicated. As marked on Figure 10, all the heaters were turned off at around 330 s and this has led to\nsmooth out the total power signal which were fluctuating due to the on-off action of the barrel/die heaters.\nThe drive motor is one of the major energy consuming components of an extruder [31, 34, 40], and, along with the\ngear box, is also responsible for significant energy losses, typically accounting for around 20% of the power supplied\nto an extruder [29]. In particular, DC motors are inefficient when operated at below the rated speed. Commercially\navailable DC motors fall into three main categories: \u201cpermanent magnet\u201d, \u201cseparately excited\u201d and \u201cself-excited\".\nThe first two are more commonly used. A block diagram for a polymer processing extruder with a \u201cseparately excited\ndirect current\" (SEDC) motor is shown in Figure 11.\n9\nPower (kW)\nSS (rpm)\n100\n60\n(a)\n100\n200\n300\n400\n500\n30\n600\n(b)\nTotal power\n20\n0\nMotor power\nAll the heaters turned-off\n-10\n0\n100\n200\n300\nTime (S)\n400\n500\n600\nFigure 10: (a). Screw speed (SS), (b). Motor power and total power signals over the time [52]\nTL\nElectrical dynamics\nGear Box\nVa\n1\nTm\n1\n@m\n@sc\n+\nK\u2081\n+\nLmS + Rm\nJmS+Bm\nN\nV\u2081\nEb\nMechanical dynamics\nKf\nKm\nSpeed controller\nFigure 11: Block diagram of an extruder with a variable field DC motor [62]\n10\n\n\nIn this figure, T is the motor torque, T is the load torque on the screw, Ra and La are the armature resistance and\narmature inductance respectively, K, and K, are the torque constants related to the field and motor, respectively, Ia is\nthe armature current, V+ is the field voltage and Bm and Jm are the damping constant and the steady-state inertia of the\nloaded screw, respectively. For extruders with a permanent magnet motor, the same block diagram is valid without\nthe branch related to the separately excited field (with V\u0192 and K\u0192 block).\nReasons for the popularity of DC motor drives in the polymer processing industry [63, 64, 52] include:\n\u2022 Smooth operation over a wide speed range,\n\u2022 Simplicity in speed control,\n\u2022 Production of a constant/consistent torque from zero to base speed,\nRelatively low power/energy consumption,\n\u2022 Relatively smaller size compared to other drive types with the same capacity,\n\u2022 Compact and simple power circuit, engaged with Silicon-controlled rectifiers,\n\u2022\nEasy installation,\n\u2022 High reliability,\n\u2022 Low intial capital cost, and\n\u2022 Less noisy than AC motors.\nDrawbacks of DC motor drives include the need for maintenance of brushes and commutator as well as the energy\nloss known as \"brush loss\u201d [59]. Green [63] observed that the best power factor that can be achieved by a DC motor\noperated at its top speed is of approximately 0.87, whereas for the best possible efficiency the power factor should be\nclose to 1.\nCurrently, AC motor drives are increasing in popularity since the power factor can be maintained constant across\nthe entire speed range. As a result, companies can avoid paying penalty charges for lagging power factor conditions.\nThe most significant drawback of AC motors is that they require a constant current to produce a constant torque, hence\ndemanding constant cooling regardless of the motor speed [63], and this results in an additional energy cost. Other\nissues include the fact that AC motors are generally larger, by a factor of 1.5-2.1, in volume, as well as being more\ncomplex than DC motors. These issues are becoming less critical nowadays, thanks to advances in electronics such\nas large scale chips and micro-processors. For both types of motors, while under operation at the rated speed, the\nmaximum energy efficiency can be achieved, industrial extruders are typically operated at lower speeds in order to\navoid undesirable fluctuations, particularly of the melt thermal quality.\nBarlow [40] argues that because the displacement power factor of a DC motor drive is proportional to the speed,\nthe power factor reduces as the motor slows down (see Figure 7). Further, it is pointed out that because the diode\nbridge of the input section of a pulse-width-modulated AC vector control drive rectifies the AC into DC, and that the\nenergy is stored in capacitors, the current and voltage waveforms are mutually in phase, and hence the motor operates\nat a power factor, in the range of 0.90 to 0.98. More information on these motor drives can be found in the literature\n[65].\nIt seems that a significant amount of electrical energy may be lost simply as a result of the low power factor\noperation of motor drives [44]. Here, Eickelberg [66] suggests that use of capacitors may be one of the solutions to\nthis problem, to smoothen the power supply. Several examples of the use of capacitors by commercial processes are\nprovided, but it is stated that this is unlikely to be a practical solution for polymer extrusion processes because of\nthe variability that occurs in the load. The installation of a more appropriate form of power factor correction would\nrequire investigation of the relevant issues [67].\nKent [41] observes that in extrusion plant energy usage assessments the energy requirements of motors in equip-\nment such as extruders and injection moulding machines is often over-looked. Other authors [68, 69] report that\nconsiderable energy savings can be achieved by replacing DC motors with AC motors. Lounsbury and Karafilidis\n[70] present factors to be considered in the selection of a drive motor.\n4.2. Barrel and die heaters\nNormally, three different types of heating method can be identified in extrusion. These are known as \"resistance\nheating\u201d, \u201cinduction heating\u201d and \u201cfluid heating\u201d [61].\n11\nResistance heating: This is also called electrical heating, and is the type most frequently used in extruders.\nUsually, electric heaters offer several advantages over fluid and steam heating, such as the possibility of covering\na broader temperature range, cleanliness, easy maintenance, low cost, and better efficiency. As a result of these\nadvantages, fluid and steam heaters have been replaced by electric heaters in modern applications. Currently extruders\ntypically have between two and ten heating zones depending on the size of the extruder.\nIn the most common conventional resistance heaters, the heat from the resistance wire is transferred to ceramic\nsegments that surround the outer surface of the barrel. This heats the barrel up until its inner surface is hot: the heat\nis then transferred to the plastic in the machine so it can be processed. With this heating method, much of the heat\ngenerated is wasted.\nInduction heating: In this case, an AC current is passed through the primary coil surrounding the extruder barrel.\nThis gives rise to an eddy current. Where the material being processed has significant relative permeability, heat may\nalso be generated by magnetic hysteresis. A high energy density can be achieved quickly with induction heating. The\nfrequency of the AC is selected depending on the size and material type and the heat penetration depth.\nWith the advantage of rapid heating and energy efficiency, induction heating has been used in many industrial\napplications [71]. The key advantage of induction heating over resistance heating is that the extruder barrel itself\nbecomes the heating element. This eliminates the conduction problem that exists with conventional heaters. As there\nare no ceramic layers, clamping bands, or water jackets to heat up, the heat is direct and instantaneous. Furthermore,\nthe induction heating generates a very even and precise heat profile, ensuring consistent heating of the polymer melt\nand leading to improved product quality. The application of induction heating to polymer processing has been devel-\noped by the Nordson Xaloy Company, which claims a reduction in heating related energy consumption of up to 50%\ncompared with typical band electric heaters [72, 73].\nFluid heating: Fluid heating uses hot liquid or steam passing through pipes/tubes. This can ensure even temper-\nature distribution but has significant disadvantages, such as demanding high levels of maintenance, the possibility of\nleaking or corrosion, and system complexity.\nAn alternative all three heating methods, is to supply the energy to the material being processed as drive power\nrather than as heater power [44, 74]. In this way shearing of the material leads to both heating and mixing. Of\ncourse, direct heating and cooling can be essential, to maintain the process thermal stability, but as guiding principles\nit would be preferable to minimize the direct heating and to avoid exceeding the melt temperature, in order to avoid\nunnecessary energy costs.\n4.3. Control electronics and monitoring devices\nA number of control and monitoring devices are used in extrusion lines, such as speed and temperature controllers and\nindicators; pressure indicators and gauges; dimension scanners; feed monitoring devices; current indicators; relays;\nswitches; alarms; etc.. All of these use some power for their operation; however, this is insignificant compared to the\ntotal energy demand. This is evident from the experimental data presented in Figure 10-(b), the amount consumed by\ncontrol electronics and monitoring devices are shown by the difference between blue and red lines (i.e., the difference\nbetween to total and motor powers) after turning off all the heaters, or by the blue line roughly between 540-570 s\nafter turning off all heaters and the motor.\n4.4. Auxiliary equipment\nAuxiliary equipment, such as pelletizers, gear pumps and screen changes, might also consume enough power to be\nconsidered in energy evaluations. Rice [75] suggests that improved energy efficiency for the entire operation of the\nplant can be achieved by combining processes.\n4.5. Process cooling\nProcess cooling is required where there is a need to remove excess heat in order to maintain the process thermal\nstability. This can be achieved by fan coolers attached along the barrel, or by cooling the screw core or the barrel wall\ninternally, using a cooling fluid such as water or oil. Air cooling provides slower changes in temperature compared\nwith liquid cooling.\n12\nAlthough cooling helps to ensure stable process operation, improper cooling, particularly cooling of the screw,\ncan lead to the generation of undesirable process fluctuations. Strauch [31] argues that excessive screw core cooling\ncan lead to a reduction in throughput rates, while at the same time affecting the melt temperature. This also impacts\non the pumping stability as a direct consequence of altering the viscosity of melt. Periodic or random temperature\nvariations of the extruder metal surfaces can arise as a result of cooling problems with the screw or barrel, and these\nmay lead to melt flow problems such as melt viscosity fluctuations [76].\nThe research of Womer et al. [35] into the effects of cooling indicated that, where water cooling is used rather\nthan air cooling, the extruder consumes more energy, irrespective of the material being processed. In consequence,\nit is recommended to use air cooling only, in conjunction with a properly designed screw, unless extensive cooling is\nrequired.\n5. Trends in polymer processing energy efficiency improvements\n5.1. Machine development and operational modifications\nAmong the current research and development \"hot topics\" for process energy efficiency, the three most likely to deliver\nsignificant benefits [63] are: (i) the design of direct drive extruders, (ii) improvements in heater and barrel design to\nreduce heat losses, and (iii) thedevelopment of \u201cadvanced vector control alternating current (AC) drives\u201d. In addition\nto these, there are reports which claim that direct drive machines offer further advantages including energy efficient\noperation, narrow footprint, quiet operation, and low maintenance requirements. The use of insulation blankets has\nalso becoming popular in energy saving of machines.\n\"Load management\u201d offers the possibility to save power in the near term. The idea is to achieve the required energy\ncost reduction by maintaining the load factor, rather than focusing on power consumption per se. Since commercial\nelectrical energy supply rates usually depend on the peak demand made by the customer, this can reduce the cost of the\nelectricity used over the duration of the charging period. To operate an effective load management plan, it is necessary\nto have a plant monitoring system to study the real-time plant power usage. In Kent's [41] discussion of energy saving\nin polymer processes, it is pointed out that the purchase of energy efficient capital equipment is profitable in long-run\ndespite initial capital costs.\n5.2. Waste heat energy recovery\nAlthough a significant amount of process heat is removed purposely to maintain the thermal stability in polymer\nextrusion, there has been insufficient attention on the recovery of waste heat for useful work.\nThe challenge [42, 77] is to find a means of re-use for that energy. Future research is required to explore such\nopportunities. For example, the recovered heat might be used for pre-heating the material prior to feeding into the\nhopper, or it could simply be used for space heating.\nFor some materials, resins need to be pre-heated prior to processing to remove moisture. In this case, part of the\nsupplied energy is lost through the evaporation of the moisture as water vapour, part is lost to heating of the surround-\nings, and the rest contributes to heating the resin. If the drying operation takes place remotely from the processing\nmachine, then the heat absorbed by resin will be lost during transit to the processing machine. Therefore, the re-design\nof the drying system, as an in-line step of the processing machine, should help to reduce overall processing energy\ncosts.\n5.3. Material choice, material recycling and disposal\nIt is usually the case that the manufacture and fabrication of plastic products makes lower energy demands than\nequivalent traditional metallic or glassware products. The development of new resins that can be processed at lower\ntemperatures, and hence for reduced energy, is part of the growing tide of interest to cut process energy expenses\neven further [78, 79, 80, 81]. On the other hand, polymeric materials based waste management has become a global\nconcern. A wide range of recycling techniques are available depending on the types of polymers and the manufacturing\ntechniques used [82]. In regard to process energy costs, the energy consumption for plastics is lower than for materials\nsuch as paper, glass, tin, and aluminium. What is more, Rosato et al. [34], the incineration of plastics as part of\nmunicipal waste yields much more energy than other material waste, such as food waste, paper and rubber, and waste\nvolume can be reduced by 90-98%.\n13\n6. Advanced process monitoring and control: adoption of Industry 4.0\nAdvanced process monitoring and control can play a vital role in achieving good product quality as well as in energy\noptimization. Some approaches are discussed in this section together with experimental results.\n6.1. Industry 4.0 and Internet of Things\nWidespread uptake of on-line or in-line monitoring and control in manufacturing processes has been enabled by\ncomputerised communications, earning it the epithet: \u201cThe fourth industrial revolution\u201d, or \u201cIndustry 4.0\u201d for short\n[83, 84]. The key requirements for an Industry 4.0 manufacturing process are: sensors - the means to observe the\nprocess as it currently stands; actuators - the means to modify the process; electronic communications - the means to\npass sensor or control information; and a decision-maker - to determine the course of action to be taken based on the\ninformation received.\nThe precise nature of the decision-maker is a point of some contention in the literature. For some, the decision-\nmaker is a computer-based Artificial Intelligence (AI), which would work completely autonomously, and steadily\nimproving its decision-making capability based on learned patterns of experience. There is still significant value in\nthe Industry 4.0 infrastructure even without an AI capability making the control decisions. For some applications, a\nrule- or model-based computer program would be effective, and many companies now boast of having an Industry\n4.0 implementation of this form. In some applications, having the sensor information fed to a control centre, means\nthat human decision-making can be facilitated and supported. Such control centres are valuable for the control of\nprocesses that are in remote or difficult to access locations, such as the health-monitoring of in-flight aircraft.\nIndustry 4.0 technology is also becoming an increasingly common tool in the home or in social care settings. Here,\nthe more common terminology is Internet of Things (IoT) [85], referring to communications connectivity between\n\"Smart\" devices. These smart devices are pieces of equipment which can send or receive information, in other words,\nthey are equipped with sensors or actuators. Thus, in the home, one might simply ask \u201cAlexa\u201d [86] to turn on the lights,\nbut in time it could easily be imagined that Alexa could modify the home heating to match the schedule inferred from\nfamily member diaries. In a patient care setting [87], the IoT system could be collecting valuable health-related\ninformation and relaying that to a control centre. Care staff or an AI system might detect anomalies and prompt a\ncheck of the patient. It is easy to see that these concepts are similar, whether applied to the home or to industry, and\nthe technology is pervasive and rapidly developing.\nIn the context of the polymer processing industry, it is clear that Industry 4.0 will not only enhance plant operation\nand its maintenance schedule, but also to energy efficiency [88, 89]. The lessons learned over the past half century,\nand reviewed in earlier sections of this paper, are ready to be applied. Wherever there are frequent variations in the\nfeedstock material, processing rate demand, or other factors, with an Industry 4.0 infrastructure in place, it becomes\npossible to monitor performance over time, and to make controlled changes. Systems health monitoring can be used\nto reduce life limiting loads on mechanical parts [90] as part of the maintenance strategy.\nGreater investment in IoT enabled devices will become an increasing imperative, for any polymer processing plant\nthat wishes to reduce its energy footprint, reduce processing costs, and sustain the mechanical plant more effectively.\nAs soon as the Industry 4.0 infrastructure is operational, development of the decision-making capability can be begun;\nwhether that is to be based on AI principles, control systems mathematics or other rule or model based paradigms.\n6.2. Ultrasound\nThe application of ultrasonic waves to reduce viscosity and thus energy consumption in polymer processing has been\ndiscussed in several recent reports [91, 92, 93, 94, 95, 96]. Maintaining a consistent melt viscosity enables improved\nprocess ability, leads to fewer product defects, reduces energy consumption and reduces materials wastage [97, 98, 99].\nChen et al [92] and Zhang and Li [93] report on the use of ultrasound vibration to influence polypropylene (PP)\nmelt, leading to non-Newtonian flow characteristics with reduced viscosity. Other authors have examined the rela-\ntionships between temperature, pressure, work-stuff throughput, the energy consumption and the ultrasonic intensity\n[95, 96]. The integration of ultrasound into a closed loop extruder control system has been developed and introduced\nin detailed by Nguyen et al. [100], who showed that controlling just the temperature, or just the ultrasonic output,\ngave a better time response and reduced energy consumption, than when trying to control both the temperature and\nultrasound together.\n14\n6.3. Closed loop melt temperature control\nMelt temperature can serve as a proxy parameter for melt viscosity and can thus be used to determine the melt quality.\nRecent works by Abeykoon et al [101, 102, 103] incorporate a fuzzy logic approach for the real-time closed loop\ncontrol of melt temperature, and demonstrated excellent performance in achieving desired set temperatures.\nAt present, in the majority of polymer processes, melt pressure and melt temperature are taken as the key param-\neters for process functionally and control. Screw speed, and barrel and die set temperatures are taken as the main\nprocess control parameters, but there is no actual feedback taken from the process melt for making process control\ndecisions, so no corrective actions can be taken to avoid product defects. Furthermore, this affects the production rate\nleading to wasted energy, labour and raw materials. Hence, combined process monitoring and control approaches,\nwhich can observe the melt quality and take control actions would represent a major development of polymer pro-\ncesses.\n7. Applying computational process simulation and dynamical systems to Control\nDevelopments in computational process simulation methods has the potential to revolutionise the design of manu-\nfacturing tooling and processes. Capabilities such as finite element method and computational fluid dynamics have\ndeveloped very significantly, with most commercial packages (e.g. [104, 105, 106, 107, 108, 109]) now offering\nmulti-physics simulation: simulation of static and transient solid mechanics, fluid dynamics, thermodynamics, elec-\ntricomagnetism, and in many cases, much more. Given the improvements in computer hardware, storage capacity\nand the development of parallel processing architecutures, computational analyses with high levels of geometric or\nmaterial modelling complexity can now be readily envisaged. This is a huge opportunity to grasp, and one for which\nmost industrial plants are not fully prepared.\n7.1. Computational mechanics methods\nThe key computational capabilities pertinent to polymer processing include a variety of modelling techniques for vis-\ncoelastic materials irrespectively of whether they should be treated as solids or liquids. Where previously, modelling\nbased on a Lagrangian Finite Element formulation, [110], would have been unable to capture the necessary defor-\nmation, Eulerian formulations, [111], and Smooth Particle Hydrodynamics (SPH), [112] and [113], now have that\npotential. Perhaps it is not enough to model using just one technique or another, but to capture one pertinent aspect\nof the physics in one region of the process, and another aspect in another region using such modelling techniques as\nco-simulation and subdomain modelling.\nNot only it is the type of analysis that is of importance to capture the process simulation requirement, but also the\nway in which the material properties are represented. The distinction between solid and liquid is no longer so clear-\ncut. There are very many material models, developed and applicable to different analysis types, that will capture the\nessential material property physics of any realizable material. Polymers offer particular challenges, but characteristics\nsuch as static stress-strain, strain rate dependence, creep, and temperature dependence are readily modelled. Chemical\nchanges, including exo- or endothermic reactions, and cure shrinkage still present a particular challenge to thermoset\npolymers, but for thermoplastic processing the challenge of representing materials properties at glass transition, and\nextent of crystallization, are, if not straight forward, at least more tractable [114, 115, 116]. For the characterisation\nof polymers undergoing extrusion processing Abeykoon et al, [117], have made significant investigations.\nWith the increase of computing power and the capability for higher model complexity there has been increased\ninterest in the application of computational mechanics methods using sophisticate material models to the simulation\nof the extrusion process. In the past decade authors such as Zairi et al, [118], have been able to model plastic flow,\ntaking account of viscoelastic properties, within a die of finite constrained section, and have been able to make some\nprediction of the microstructure of the end product. While the main focus has been of the effect of the die geometry\non the extruded product, another important consideration is the design of the die tooling, and the loads that must\nwithstand during processing [119].\n15\n7.2. Statistical process control\nThe biggest challenge for the modelling of the polymer extrusion process is to understand the conditions that give rise\nto large variations in product output quality. Where small changes in the processing parameters lead to only small\nchanges in output, even where the change is clearly non-linear, one might take a step-wise approach to linearize the\nparameters and have a measurable and definable set of limits for process control. This is the ideal, textbook, approach\nto manufacturing process excellence, LEAN manufacturing or Six Sigma, [120], [121]. In constrast, where small\nchanges in processing input lead to significant changes in the output quality, a manufacturing engineer would say that\nthe process is not in statistical control. The usual manufacturing engineer's approach for regaining statistical control\nis to monitor parameters to within ever tighter bounds. Clearly this would lead to increased manufacturing costs, and\ncould make the process financially unviable. The approach may ultimately even be completely unsuccessful.\nThe factors that would generally be considered for statistical process control would include the initial conditions:\nhow the process is started up, and the material condition on start-up, factors directly related to the geometry of the\nextruder and the die, environmental conditions and changes, and changes in the raw material batch.\nIn order to control the process, certain actions might be taken. The applied drive torque, a function of the power\nsupplied to the extruder, could be controlled in direct response to measurements relating to the process quality. In-\nevitably there is a delay between making the measurement, and making a change, primarily because the measurement\nis taken at some point down-stream in the process. There is a considerable level of research activity in this area, with\nvarious computational schemes for assessing these measurements and informing the choice of control action to be\ntaken. Clearly this data processing time must be minimised to minimise the feedback delay. Wagner et al, [122], and\nMcKay et al, [123], recognised that the fundamental requirement was to use the measurements to infer the local mate-\nrial viscosity within the process, using artificial intelligence methods such as neural nets. Chen et al, [124], employed\na power law model. Methods based on fuzzy logic were developed by McAfee, [125], McAfee and Thompson, [126],\nand later by Liu et al, [127] and Abeykoon [103]. More recent work includes the soft sensor technique of Deng et al,\n[128] and Abeykoon [102], and the multi-objective optimisation approaches presented by Carrano et al, [129].\n7.3. Non-linear system dynamics\nWhere the dynamics of the manufacturing process make statistical process control challenging, a radically different\napproach is needed. Non-linear systems dynamics [130] has been a topic of considerable research and development,\nmainly by researchers with a strong background in applied mathematics, and with applications including but by no\nmeans limited to manufacturing engineering applications. By developing a non-linear system dynamics representation\nof a manufacturing process, it is possible to explore how process parameters could influence the process dynamics,\nand from this pin-point the controling factors or initial conditions.\nIn the field of polymer extrusion modelling, McKinley, et al, [131], used laser doppler velociometry to visualise\nthe flow towards an abrupt contraction, and found that the flow near the tip of the contraction could show time periodic\nand aperiodic behaviour. Graham [132] modelled the fluid behavious regarding wall slip and was able to replicate the\nlarge amplitude periodic and aperiodic oscillations observed experiementally. Smith et al, [133, 134] demonstrate a\nsteady-state solution through modelling the polymer as a purely viscous non-Newtonian material, and optimising die\ngeometry.\nThe critial feature in the extrusion process is the variation in the material property of the polymer as it passes\nthrough the process. As the material is being worked mechanically, it becomes heated. As it gets hotter, the elastic\nstiffness and viscosity are reduced. The resulting thermal expansion gives rise to a localised increase of pressure.\nAs a result of both the reduced viscosity and increased pressure, the polymer would pass more readily through the\nprocess: and would require less working. Reduced working would result in reduced heating, with the result of inceased\nstiffness, increased viscosity and reduced pressure. It is clear that a cyclic response can be expected. Now consider\nhow the viscosity of the polymer changes with strain rate and the difference of time dependency in glass transition and\ncrystallization, it bceomes clear that there is a level of complexity between production rate, localized temperature and\nviscous response of the material: in the simplistic mass, spring and damper system it is the variability of the damper\nas well as the forcing that is key to understanding the dynamical system.\nA non-linear dynamical systems model of the polymer extrusion process might be modelled as follows. First,\nthe polymer has the properties of both an elastic solid, and a viscous fluid: it can be idealized as a mass, spring\nand damper system, as shown in Figure 12. The non-linear dynamics of such systems have been studied for many\n16\nidealized situations of varying applied force, F(t), and their modelling involves the selection of initial conditions, and\nthen computing the transient behaviour until a long term behaviour becomes apparent.\nk(t)\nx(t)\nc(t)\nm\nF(t)\nFigure 12: Idealised mass, spring and damper system\n-\nFigure 13 shows the time varying position, x(t) and velocity x(t) of the mass, driven by the time varying force\nF(t). The initial conditions for the system might be any point on the x \u2014 x plane, and one might think of particular\nexamples of such initial conditions being the tail end of the solid or dashed curved arrows. As time progresses, the\nsystem would move on from that initial condition, moving towards the head of the arrow. The dashed arrows indicate\nthe evolution away from an unstable limit cycle \u2013 the dashed closed loop. Any initial condition that begins on the\nunstable limit cycle will eventually decay inwards or outwards, and migrate towards a stable limit of some sort. The\nsolid arrows indicate evolution towards either a stable limit cycle the solid closed loop, or towards a steady-state\nsolution \u2015 the solid dot. The stable limit cycle represents a periodic motion. The area enclosed by the unstable limit\ncycle gives an indication of the relative level of damping in the system: it is important to note that increasing the\ndamping would increase the area within the unstable limit cycle, so that a larger proportion of initial conditions would\nultimately lead to a steady-state solution; however in the case of initial conditions sufficiently close to the stable limit\ncycle, periodic motion would still ensue. The notion that damping removes energy from a dynamic system is true, but\nthe energy required for periodic motion is renewed at every cycle by the time varying force, F(t).\nIn the context of polymer processing, periodic motion represents both unnecessary energy utilisation, and reduced\nprocess control. This explanation is more easily understood when described in the context of everyday experience.\nx(t)\nx(t)\nFigure 13: Time evolving position and velocity diagram\nConsider a friction driven vibration such as the vibration of a bowed violin string, brake squeal, or a singing wine glass.\nIn each case, the string, the brake disc and the glass have a stiffness, a density and some form of damping is present.\nUnder particular forcing conditions such as applied by a well-rosined violin bow, a certain braking pressure or being\nstroked with a wet finger, these structures vibrate at a well-defined pitch. The fact that the pitch is well-defined means\nthat the vibration is periodic. Under different forcing conditions: a bow without rosin, a different braking pressure\nor a dry finger, the contact slips with constant velocity. This is a steady-state solution: a force is being applied and\nso there is a displacement, but it is a constant force and a constant displacement. For a detailed account of such an\nanalysis, and the computational algorithms used [135].\n17\nThis description is necessarily rather simplistic, but it provides a basis for reviewing the pertinent features required\nfor modelling the extrusion system. In place of displacement and velocity one should be considering the extruded\nvolume of material as a function of time. In place of force, consider the power output of the machine and what\nresistance it meets from the material being processed. In an ideal process, one would wish for constant extruded\nvolume and constant levels of resistance, so that with an appropriate choice of initial conditions and damping, a steady-\nstate outcome can be achieved. The elastic response and the viscosity of the polymer depend on temperature and strain\nrate, and as such are the time dependent system stiffness, k(t), and time dependent damping, c(t), respectively. Recent\nwork in this vein, but applied to blown film extrusion, is presented in a detailed review paper by Pirkle et al, [136].\n8. Conclusions\nThere have been some very significant developments in computational modelling capability, which means that the\nsimulation of manufacturing processes such as polymer extrusion is now feasible. Multi-physics approaches, combin-\ning material flow and thermal behaviour of the material can, in theory, be modelled. Advances in computer hardware\nmean that models with very high levels of geometric complexity, and therefore high numbers of computational degrees\nof freedom are within reach. Co-simulation modelling, including the modelling of not only the material undergoing\nextrusion, but also of the working state of the dies and the extrusion machine itself, is also within reach, meaning\nthat optimal design of tooling for improved life and reduced machine maintenance demands are additional areas for\nfurther development.\nComplex models of the process and material property variation can also be used to create the simplified models\nrequired for a non-linear systems dynamics modelling approach. On that basis, measurable parameters that would\nhave an effect on the stability of the real production process can be examined, and warning limits found. Even where\nsuch an approach might not be able provide an accurate prediction of limits, it could provide insight that would lead\nto practical solutions or avoidance of particular operating regimes.\n9. List of Abbreviations\nTerm\nDefinition\nAC\nAlternating current\nDC\nDirect current\nIoT\nInternet of Things\nSEC\nSpecific energy consumption\nSEDC Separately excited direct current\n10. List of Symbols\n18\nTerm\nDefinition\nBm\nDamping constant\n\u0421\u0440\nSpecific heat capacity\nEin\nEnergy consumed\nElosses\nEnergy loss\nF(t)\nA time varying force\nI\nLine current\nIa\nArmature current\nIf\nField current\nJm\nSteady-state inertia of the loaded screw\nkm\nThermal conductivity\nKf\nTorque constant related to the field\nK\u2081\nTorque constant related to the motor\nLa\nArmature inductance\nN\nGear ratio\nP\nActive power\nQ\nReactive power\nR\nElectrical resistance\nS\nApparent power\nRa\nArmature resistance\nTb\nExtruder barrel set temperature\nTL\nLoad torque on the screw\nTmelt\nMelt temperature\nTm\nMotor torque\nx(t)\nA time varying position\nx(t)\nA time varying position velocity\nV\nLine voltage\nVa\nArmature voltage\nVbx\nThe component of the barrel velocity in the transverse direction\n19\nV\u0192\nField voltage\nVj\nThe resultant relative velocity\nTerm\nDefinition\nCOS\nThe displacement power factor\nPm\nMelt density\n\u03b7\nMelt viscosity\nWactual\nActual screw speed\nW set\n\u03bb\nC\u2081\n\u03a9\nSet screw speed\nTemperature of the solid bed\nRate of melting\nReferences\n[1] British plastics federation, about the british plastics industry, Available at: http://www.bpf.co.uk/Industry/Default.aspx., Last viewed: 27 of\nFebruary 2019.\n[2] Plasticseurope: Plastics the facts 2017 an analysis of european plasticsproduction, demand and waste data, Available at:\nhttps://www.plasticseurope.org/application/files/5715/1717/4180/Plastics the facts2017 FINAL for website ne page.pdf, Last viewed: 27 of\nFebruary 2019.\n[3] Market outlook:\nAvailable at:\nPolymers world turned on its head (icis chemical business),\nhttp://www.icis.com/resources/news/2013/10/25/9718765/market-outlook-polymers-world-turned-on-its-head/, Last viewed: 31 of\nAugust 2016.\n[4] Available at: http://www.plasticseurope.org/, Last viewed: 10 of August 2012.\n[5] J. Vlachopoulos, D. Strutt, Overview: Polymer processing, Materials Science and Technology 19 (2003) 1161\u20131169.\n[6] European best practice guide: Low energy plastics processing, Available at: http://www.tangram.co.uk/TI-Energy-LowLast viewed: 27 of\nFebruary 2019.\n[7] K. L. O. C. S. L. G. Kaur, K. Uisan, Recent trends in green and sustainable chemistry & waste valorisation: Rethinking plastics in a\ncirculareconomy, Current Opinion in Green and Sustainable Chemistry 9 (2018) 30\u201339.\n[8] J. Frankland, Mission (nearly) impossible: Estimating extrusion melt temperature, Plsatics TechnologyAugust Issue.\n[9] M. J. Stevens, J. A. Covas, Extruder principles and operation, 2nd Edition, Chapman & Hall, 1995.\n[10] C. Rauwendaal, Polymer Extrusion, 4th Edition, Hanser, 2001.\n[11] C. Abeykoon, Polymer Extrusion: A Study on Thermal Monitoring Techniques and Melting Issues, Lap Lambert Academic Publish-\ning: Verlag, 2012.\n[12] C. I. Chung, E. M. Mount, D. E. McClelland, Energy efficiency in plasticating screw extrusion, Proceedings of 176th Symposium in Energy\nConservation in Textile and Polymer Processing (1979) 21-36.\n[13] J. Agassant, Polymer Processing, Hanser-Munich, 1991.\n[14] A. N. Wilkinson, A. J. Ryan, Polymer Processing and Structure Development, Springer-Netherlands, 1998.\n[15] T. Osswald, J. Hernandez-Oritz, Polymer Processing, Hanser-Munich, 2006.\n[16] W. H. Darnell, E. A. J. Mol, Solids conveying in extruders, SPE Journal 12 (1956) 20-29.\n[17] B. H. Maddock, A visual analysis of flow and mixing in extruder screws, SPE Journal 15 (1959) 383-389.\n[18] Z. Tadmor, Fundamental of plasticating extrusion. i. a theoretical model for melting, Polymer Engineering and Science 6 (1966) 185\u2013190.\n[19] Z. Tadmor, I. Klein, The formulation of a mathematical model for plasticating screw extrusion, Rubber Chemistry and Technology 42 (1969)\n780-789.\n[20] Z. Tadmor, I. Klein, Engineering principles of plasticating extrusion, 1st Edition, Society of Plastics Engineers, 1970.\n[21] Z. Tadmor, C. Gogos, Principles of polymer processing, Wiley-New York, 1979.\n[22] E. Severs, Rheology of polymers, Van Nostrand Reinhold Inc, 1962.\n[23] C. Abeykoon, P. Perez, A. L. Kelly, he effect of materials' rheology on process energy consumption in polymer extrusion, Polymer Engi-\nneering and Science 60 (6) (2020) 1244-1265.\n[24] Y. Breaux, J.-Y. Charmeau, M. Moguedet, A simple model of throughput and pressure development for single screw, Journal of Materials\nProcessing Technology 209 (2009) 611-618.\n[25] H. S. Rowell, R. D. Finlayson, Screw viscosity pumps, Engineering 114 (1922) 606\u2013607.\n[26] H. S. Rowell, R. D. Finlayson, Screw viscosity pumps, Engineering 126 (1928) 249-387.\n[27] C. Abeykoon, Modelling and control of melt temperature in polymer extrusion, Ph.D. thesis, Queen's University Belfast, United Kingdom\n(2011).\n[28] C. Abeykoon, J. Vera-Sorroche, A. L. Kelly, E. C. Brown, P. D. Coates, J. Deng, K. Li, N. Karnachi, E. Harkin-Jones, M. Price, Process\nefficiency in polymer extrusion: Correlation between the energy demand and melt thermal stability, Applied Energy 135 (2014) 560-571.\n[29] G. A. Kruder, R. E. Nunn, Optimizing energy utilization in extrusion processing, SPE ANTEC Technical Papers (1981) 648-652.\n20\n\n\n[30] J. M. McKelvey, Energy utilisation in extrusion, Advances in Polymer Technology 3 (1983) 205-211.\n[31] T. H. Strauch, F. Gliese, G. Menges, Energy usage and conservation in extrusion plants, Plastics and Rubber Institute (Polymer Extrusion\nIII Conference), London, England 1 (1985) 2.1-2.10.\n[32] J. Anderson, C. Rauwendaal, M. del P. Noriega, Troubleshooting extrusion problems, SPE ANTEC Technical Papers 1 (1997) 127\u2013134.\n[33] H. Falkner, Energy cost savings in motive power, Water and Environment Journal 11 (1997) 225-227.\n[34] D. V. Rosato, N. R. Schott, D. V. Rosato, M. G. Rosato, Plastics engineering, manufacturing & data handbook: Fundamental and processes,\nVol. 2, Kluwer Academic Publishers, 2001.\n[35] T. W. Womer, W. S. Smith, R. P. Wheeler, Comparison of two different cooling methods for extrusion processes, SPE ANTEC Technical\nPapers (2006) 796-801.\n[36] About the industry, Available at: http://www.bpf.co.uk, Last viewed: 27 of February 2012.\n[37] K. M. Cantor, Analyzing extruder energy consumption, SPE ANTEC Technical Papers 2 (2010) 1300-1306.\n[38] R. V. Heur, M. Verheijen, Power quality & utilization guide: Plastics industry-energy efficiency, Available at: http: //www.leonardo-\nenergy.org/files/root/pdf/Application/Guide/Plastics.pdf, Last viewed: 15 of August 2012.\n[39] J. Deng, K. Li, E. Harkin-Jones, M. Price, N. Karnachi, A. Kelly, J. Vera-Sorroche, P. Coates, E. Brown, M. Fei, Energy monitoring and\nquality control of a single screw extruder, Applied Energy 113 (2014) 1775\u20131785.\n[40] S. Barlow, Reducing electrical energy costs for extrusion processes, SPE ANTEC Technical Papers (2009) 1157-1162.\n[41] R. Kent, Energy management in plastics processing - part 10: Motors and drives, Available at: http://www.tangram.co.uk, Last viewed: 22\nof february 2010.\n[42] B. Drury, The control techniques drives and controls handbook, The Institution of Electrical Engineers, London-UK, 2010, pp 308-312.\n[43] J. V. DeJuneas, How to reduce electrical energy in the extrusion process, TAPPI Pap Synth Conf (1975) 49\u201350.\n[44] J. M. Dealy, Energy conservation in plastics processing: A review, Polymer Engineering and Science 29 (1982) 528-535.\n[45] C. I. Chung, Energy efficinecy in single screw extrusion, SPE ANTEC Technical Papers (1983) 919-920.\n[46] R. Rasid, A. K. Wood, Effect of process variables on melt tempearture profiles in extrusion process using single screw plastics extruder,\nPlast. Rubb. Comp. 32 (5) (2003) 193\u2013198.\n[47] E. C. Brown, A. L. Kelly, P. D. Coates, Melt temperature homogeneity in single screw extrusion: effect of material type and screw geometry,\nSPE ANTEC Technical Papers (2004) 183\u2013187.\n[48] A. L. Kelly, E. C. Brown, P. D. Coates, The effect of screw geometry on melt temperature profile in single screw extrusion, Polymer\nEngineering and Science 46 (12) (2006) 1706-1714.\n[49] A. L. Kelly, J. V. Sorroche, E. C. Brown, P. D. Coates, Improving thermal efficiency of single screw extrusion, SPE ANTEC Technical\nPapers 2 (2012) 1080\u20131087.\n[50] J. V. Sorroche, A. L. Kelly, E. C. Brown, P. D. Coates, N. Karnachi, E. Harkin-Jones, K. Li, J. Deng, Thermal optimisation of polymer\nextrusion using in-process monitoring techniques, Applied Thermal Engineering 53 (2) (2012) 405\u2013413.\n[51] C. Abeykoon, M. McAfee, S. Thompson, K. Li, A. L. Kelly, E. C. Brown, Investigation of torque fluctuations in extrusion through monitoring\nof motor variables, Proceedings of 26th PPS Annual Europe/Africa regional meeting, Larnaca, Cyprus (2009) Paper No: 22-0.\n[52] C. Abeykoon, J. Vera-Sorroche, A. L. Kelly, E. C. Brown, P. D. Coates, K. B. Howell, J. Deng, K. Li, N. Karnachi, E. Harkin-Jones,\nM. Price, Investigation of the process energy demand in polymer extrusion: A review and an experimental study, Applied Energy 136 (2014)\n726-737.\n[53] C. Abeykoon, A. L. Kelly, E. C. Brown, P. D. Coates, The effect of materials, process settings and screw geometry on energy consumption\nand melt temperature in single screw extrusion, Applied Energy (2016) 880-894.\n[54] R. S. Mallouk, J. M. Mckelvey, Power requirements of melt extruders, Industrial and engineering chemistry 45 (1953) 987\u2013989.\n[55] K. Wilczynski, A computer model for single-screw plasticating extrusion, Polymer-Plastics Technology and Engineering 35 (3) (1996)\n449-477.\n[56] E. Lai, D. W. Yu, Modeling of the plasticating process in a single screw extruder: A fast track approach, Polymer Engineering and Science\n40 (5) (2000) 1074-1084.\n[57] C. Abeykoon, M. McAfee, K. Li, P. J. Martin, J. Deng, A. L. Kelly, Modelling the effects of operating conditions on motor power consump-\ntion in single screw extrusion, Proceedings of LSMS 2010 and ICSEE 2010 6329 (2) (2010) 9-20.\n[58] T. Wildi, Electrical machines, drives, and power systems, 4th Edition, Prentice Hall, 2000.\n[59] S. J. Chapman, Electrical machinery fundamentals, McGraw-Hill, 2005, pp 261-264, 524-527.\n[60] Motor efficiency depends upon power factorAvailable at: http://www.actel.com/documents/Motor_PowerFactor_WP.pdf.\n[61] L. Halsz, Control methods in polymer processing, Elsevier, 1993.\n[62] C. Abeykoon, M. McAfee, K. Li, P. J. Martin, A. L. Kelly, The inferential monitoring of screw load torque to predict process fluctuations in\npolymer extrusion, J. Mater. Process. Technol. 211 (12) (2011) 1907-1918.\n[63] R. M. Green, A comparison of AC and DC extruder drives, IEEE Industry Applications Society Annual Meeting 2 (1989) 1663\u20131666.\n[64] N. Mulakken, Taufik, M. Anwari, Power quality measurements of a dc motor drive, International Confernce on Power and Energy Systems,\nLangkawi, malaysia (2008) 98\u2013101.\n[65] S.-H. Kim, Electric Motor Control: DC, AC, and BLDC Motors, Elsevier, 2017.\n[66] F. Eickelberg, Continuous vacuum metallizing: it's more than tinsel . . . much more, Modern Plastics 54 (1977) 42-45.\n[67] C. S. P. Kumar, S. P. Sabberwal, A. K. Mukharji, Power factor measurement and correction techniques, Electric Power Systems Research\n32 (2) (1994) 141-43.\n[68] R. A. Sfeir, Measured savings of DC to AC drive retrofit in plastic extrusion, Proceedings of the 13th Industrial Energy Technology Confer-\nence, New Orleans, USA.\n[69] Extruder line updated with ac vector drives, Available at: http://www.yaskawa.com/site/dmdrive.nsf, Last viewed: 10 of August 2012.\n[70] D. C. Lounsbury, A. G. Karafilidis, Optimizing extruder drive selection: A review of the factors involved, SPE ANTEC Technical Papers\n(1995) 76-79.\n[71] A. A. not stated, Induction heating applications - the processes, the equipment, the benefits. kha provide full details, Internet: http://www.efd-\n21\ninduction.com/ (Year not stated, last access date not stated) Pages not stated.\n[72] B. F. T. Gregory E. Tremblay, Dan Sowle, Investigation of injection moulding performance using induction barrel heating. kha provide full\ndetails, Journal not stated (Year not stated) Pages not stated.\n[73] T. Deligio, Screws and barrels the focus of energy-savings efforts, Modern Plastic Worldwide (2008) 44\u201345.\n[74] J. M. McKelvey, Theory of adiabatic extruder operation, Industrial & Engineering Chemistry 46 (4) (1954) 660\u2013664.\n[75] W. T. Rice, Conservation of energy and raw materials by utilization of gear pumps in conjunction with extruder processes, Energy Conference\n& Exhibition, New Orleans, LA, USAASME Publication 80-Pet-2.\n[76] E. L. Steward, Control of melt temperature on single screw extruders, SPE ANTEC Technical Papers 1, paper No: 813.\n[77] K. Bhattacharjee, Energy conservation opportunities in industrial waste heat recovery systems, Energy Engineering 107 (2010) 7-13.\n[78] J. J. Baron, C. E. McChesney, S. M. Sinker, Extrusion blow moldable polyethylene terephthalate, SPE ANTEC Technical Papers (1981)\n713-715.\n[79] Energy saving pet, Modern Plastics 58 (1981) 72.\n[80] Pp that cuts users' energy costs, Modern Plastics 58 (1981) 14.\n[81] R. D. Deanin, Decreasing energy requirements in plastics processing, Polymer-Plastics Technology and Engineering 14 (1980) 37\u201345.\n[82] B. I. D. V. H. P. K. H. d. W. Y. Yang, R. Boom, Recycling of composite materials, Chemical Engineering and Processing 51 (2012) 53\u201368.\n[83] J. J. M Wollschlaeger, T Sauter, The future of industrial communication, IEEE Industrial Electronics magazine (2017) 17-27.\n[84] E. K. S. N. RY Zhong, X Xu, Intelligent manufacturing in the context of industry 4.0: A review, Engineering 3 (2017) 616\u2013630.\n[85] M. Loeffler, A. Tschiesner, The internet of things and the future of manufacturing, Web interview https://www.mckinsey.com/business-\nfunctions/digital-mckinsey/our-insights/the-internet-of-things-and-the-future-of-manufacturing.\n[86] All things alexa, Amazon Webpage https://www.amazon.com/Amazon-Echo-And-Alexa-Devices/b?node=9818047011.\n[87] M. K. M. H. K.-S. K. SMR ISLAM, D KWAK, The internet of things for health care: A comprehensive survey, IEEE Access 3 (2015)\n678-708.\n[88] G. M. F Shrouf, J Ordieres, Smart factories in industry 4.0: A review of the concept and of energy management approached in production\nbased on the internet of things paradigm, 2014 IEEE International Conference on Industrial Engineering and Engineering Management,\n9-12 December, Bandar Sunway, Malaysia.\n[89] G. S. T. Stock, Opportunities of sustainable manufacturing in industry 4.0, Procedia CIRP 40 (2016) 536 - 54.\n[90] A. F. S. Y. L. JH Ang, C Goh, Energy-efficient through-life smart design, manufacturing and operation of ships in an industry 4.0 environ-\nment, Energies 10 (2017) doi:10.3390/en10050610.\n[91] J. P. Ibar, Control of polymer properties by melt vibration technology: A review, Polym Eng. Sci. 38 (1) (1998) 1-20.\n[92] J. Y. Chen, Y. Z. Chen, H. L. Li, S. Y. Lai, J. D. Jow, Physical and chemical effects of ultrasound vibration on polymer melt in extrusion.,\nUltrason Sonochem 17 (2010) 66-71.\n[93] Y. C. Zhang, H. L. Li, Functionalization of high density polyethylene with maleic anhydride in the melt state through ultrasonic initiation.,\nPolym Eng. Sci. 43 (4) (2003) 774-782.\n[94] Y. Z. Chen, H. L. Li, Effect of ultrasound on the viscoelasticity and rheology of polystyrene extruded through a slit die., J. Appl. Polym. Sci.\n100 (4) (2006) 2907-2911.\n[95] I. Isayev, S. Mandelbaum, Effect of ultrasonic waves on foam extrusion., Polym. Eng. Sci. 31 (14) (1991) 1051\u20131056.\n[96] Y. R. Cao, H. L. Li, Influence of ultrasound on the processing and structure of polypropylene during extrusion., Polym. Eng. Sci. 42 (7)\n(2002) 1534-1540.\n[97] X. Q. Liu, K. Li, M. McAfee, J. Deng, 'soft-sensor' for real-time monitoring of melt viscosity in polymer extrusion process., Proc. IEEE\nConference on Decision and Control, GA, USA (2010) 3469-3474.\n[98] Y. G. Son, Determination of shear viscosity and shear rate from pressure drop and flow rate relationship in a rectangular channel., Polymer\n48 (2007) 632-637.\n[99] M. McAfee, G. McNally, Real-time measurement of melt viscosity in single-screw extrusion., T I Meas. Control 28 (5) (2006) 481\u201397.\n[100] B. K. Nguyen, G. McNally, A. Clarke, Automatic extruder for recycled polymers with ultrasound and temperature control system., Proc.\nIMechE Part E: J Process Mechanical Engineering (2015) DOI: 10.1177/0954408914552335.\n[101] C. Abeykoon, K. Li, M. McAfee, P. J. Martin, Q. Niu, A. L. Kelly, J. Deng, A new model based approach for the prediction and optimisation\nof thermal homogeneity in single screw extrusion, Control Eng. Pract. 19 (8) (2011) 862-874.\n[102] C. Abeykoon, A novel soft sensor for real-time monitoring of die melt temperature profile in polymer extrusion, IEEE Trans. Ind. Electron.\n61 (12) (2014) 7113-7123.\n[103] C. Abeykoon, A novel model-based controller for polymer extrusion, IEEE Trans. Fuzzy Syst. 22 (6) (2014) 1413\u20131430.\n[104] Dassault systemes, simulia portfolio., http://www.3ds.com/products-services/simulia/products/.\n[105] Ansys products., http://www.ansys.com/en-GB/Products.\n[106] Comsol, multiphysics software product suite., https://uk.comsol.com/products.\n[107] Msc software, products., http://www.mscsoftware.com/products.\n[108] Esi, software solutions., https://www.esi-group.com/software-solutions.\n[109] Livermore software technology corporation, Is-dyna., http://www.lstc.com/products/ls-dyna.\n[110] O. C. Zienkiewicz, R. L. Taylor, J. Z. Zhu, The Finite Element Method: Its Basis and Fundamentals, 7th Edition, Butterworth-Heinemann,\nOxford, 2013.\n[111] D. J. Benson, Computational methods in lagrangian and eulerian hydrocodes., Computer Methods in Applied Mechanics and Engineering\n99 (2-3) (1992) 235-394.\n[112] R. F. Stellingwerf, C. A. Wingate, Impact modelling with sph., Memorie della Societa Astronomia Italiana 65 (4) (1994) 1117\u20131128.\n[113] G. R. Liu, M. B. Liu, Smoothed Particle Hydrodynamics., World Scientific Publishing Company, Singapore, 2003.\n[114] Y. Mu, G. Zhao, A. Chen, X. Wu, Numerical investigation of the thermally and flow induced crystallization behavior of semi-crystalline\npolymers by using finite elementfinite difference method., Computers & Chemical Engineering 46 (2012) 190\u2013204.\n[115] Y. Mu, G. Zhao, A. Chen, X. Wu, Modeling and simulation of polymer melts flow in the extrusion process of plastic profile with metal\n22\ninsert., The International Journal of Advanced Manufacturing Technology 67 (1) (2013) 629\u2013646.\n[116] Y. Mu, G. Zhao, X. Wu, L. Hang, H. Chu, Continuous modeling and simulation of flow-swell-crystallization behaviors of viscoelastic\npolymer melts in the hollow profile extrusion process., Applied Mathematical Modelling 39 (3-4) (2015) 1352-1368.\n[117] C. Abeykoon, P. J. Martin, K. Li, A. L. Kelly, Dynamic modelling of die melt temperature profile in polymer extrusion: Effects of process\nsettings, screw geometry and material., Applied Mathematical Modelling 38 (4) (2014) 1224\u20131236.\n[118] F. Zairi, Should Havelwith TwoDots, B. Aour, J. Gloaguen, M. Nat-Abdelaziz, J. Lefebvre, Numerical modelling of elasticviscoplastic equal\nchannel angular extrusion process of a polymer., Computational Materials Science 38 (1) (2006) 202-212.\n[119] J. Dryden, Stress concentration in polymer extrusion dies., Acta Mechanica 219 (3) (2011) 269\u2013279.\n[120] J. Liker, The Toyota Way: 14 Management Principles from the World's Greatest Manufacturer., McGraw-Hill, New York, 2003.\n[121] M. George, J. Maxey, D. Rowlands, M. Upton, The Lean Six Sigma Pocket Toolbook: A Quick Reference Guide to Nearly 100 Tools for\nImproving Quality and Speed., McGraw-Hill, New York, 2003.\n[122] M. Wagner, G. Montague, M. Tham, Neural networks for steady state modelling of an extruder., Artificial Intelligence in Engineering 11 (4)\n(1997) 375-382.\n[123] B. McKay, B. Lennox, M. Willis, G. Barton, G. Montague, Extruder modelling: a comparison of two paradigms., Control '96, UKACC\nInternational Conference on (Conf. Publ. No. 427), September 1996, DOI 10.1049/cp:19960643 2 (1996) 734\u2013739.\n[124] Z.-L. Chen, P.-Y. Chao, S.-H. Chiu, Proposal of an empirical viscosity model for quality control in the polymer extrusion process., Polymer\nTesting 22 (5) (2003) 601\u2013607.\n[125] M. McAfee, Enhancing process insight in polymer extrusion by grey box modelling., Transactions of the Institute of Measurement and\nControl 29 (5) (2007) 467-488.\n[126] M. McAfee, S. Thompson, A novel approach to dynamic modelling of polymer extrusion for improved process control., Proceedings of the\nInstitution of Mechanical Engineers, Part I: Journal of Systems and Control Engineering 221 (4) (2007) 617-628.\n[127] X. Liu, K. Li, M. McAfee, B. K. Nguyen, G. M. McNally, Dynamic gray-box modeling for on-line monitoring of polymer extrusion\nviscosity., Polymer Engineering & Science 52 (6) (2012) 1332\u20131341.\n[128] J. Deng, K. Li, E. Harkin-Jones, M. Price, M. Fei, A. Kelly, J. Vera-Sorroche, P. Coates, E. Brown, Low-cost process monitoring for polymer\nextrusion., Transactions of the Institute of Measurement and Control 36 (3) (2014) 382-390.\n[129] E. G. Carrano, D. G. Coelho, A. Gaspar-Cunha, E. F. Wanner, R. H. Takahashi, Feedback-control operators for improved pareto-set descrip-\ntion: Application to a polymer extrusion process., Engineering Applications of Artificial Intelligence 38 (2015) 147\u2013167.\n[130] P. G. Drazin, Nonlinear systems., Cambridge University Press, Cambridge, 1992.\n[131] G. McKinley, W. Raiford, R. Brown, R. C. Armstrong, Nonlinear dynamics of viscoelastic flow in axisymmetric abrupt contractions., Journal\nof Fluid Mechanics 223 (3-4) (1991) 411-456.\n[132] M. D. Graham, Wall slip and the nonlinear dynamics of large amplitude oscillatory shear flows., Journal of Rheology 39 (4) (1995) 697-712.\n[133] D. E. Smith, D. A. Tortorelli, C. L. TuckerIII, Optimal design for polymer extrusion. part i: Sensitivity analysis for nonlinear steady-state\nsystems., Computer Methods in Applied Mechanics and Engineering 167 (3-4) (1998) 283\u2013302.\n[134] D. E. Smith, D. A. Tortorelli, C. L. TuckerIII, Optimal design for polymer extrusion. part ii: Sensitivity analysis for weakly-coupled\nnonlinear steady-state systems., Computer Methods in Applied Mechanics and Engineering 167 (3-4) (1998) 303\u2013323.\n[135] A. J. McMillan, A non-linear friction model for self-excited vibrations., Journal of Sound and Vibration 205 (3) (1997) 323\u2013335.\n[136] J. C. Pirkle(Jr.), R. D. Braatz, Instabilities and multiplicities in non-isothermal blown film extrusion including the effects of crystallization.,\nJournal of Process Control 21 (3) (2011) 405-414.\n23\n"}, "expected_output": {"claims": [{"unit": "%", "value": 60, "evidence": ["enhancements to machinery could be identified, where an economic case could be made on the basis of energy cost\nsaving. Some examples are given below.\nIn the late 1970s, Chung et al. [12] found that for a 63.5 mm diameter extruder mechanical energy efficiency of\n62% was typical, and for larger extruders the energy efficiency was lower. In 1981, Kruder and Nunn [29] claimed\nthat energy efficiency of extruders can range from 45%-75%. It was noted that the energy efficiency depended on the\ntransmission mechanism, screw design, product geometry, nature of polymer feedstock and the production rate, while\nthe major energy losses of an extruder occur as a result of the forced cooling process step, and the losses associated\nwith the drive and transmission unit. At low screw speeds, barrel heaters consume a considerably higher portion of\nenergy than at higher speeds, and significant energy savings could be made by running the processes at the highest\npossible power factor. Additionally, this work presented information on energy demand and losses of each individual\ncomponent of an extruder.\nSubsequently through the 1980s, most research into energy efficiency was focussed on the screw efficiency and\nmass flow rate. A reduction in the overall power requirement for an extruder can be achieved through the use of a\ngeared pump at the end of the extruder to increase the mass flow rate (McKelvey [30]). In 1985, Strauch et al. [31]\ncarried out an energy consumption study on a 63.5 mm diameter single screw extruder, and observed that most of the\nenergy was consumed by the mechanical parts, with less significant levels of consumption in process heating. The\nenergy conversion was then assessed and it was found that heating the water in the cooling system accounted for more\nthan half of the energy supplied.\nDuring the late 1980s and 1990s, the manufacturing sector was making many changes, with a view to improving\nproductivity and quality. Driven by the advances made in Japanese manufacturing, the main focus during that time\nwas on management methods, such as Total Quality Management, LEAN, and Six Sigma. These efforts initially\naddressed cost and time issues, where the biggest economic benefits were to be found. Latterly, interest in energy\nefficiency began to be seen as not only cost reduction opportunity but also as an environmental imperative.\nIn the context of power consumption in the extruder, in 1997 Anderson et al. [32] recognised that for the processing\nof most plastics, from room temperature, the specific energy consumption (SEC) of the extruder motor should be in\nthe\nrange of 0.0822 to 0.1644 kW.hr/kg.\nAt around the same time, a study by Falkner in 1997 [33] showed that motor operations accounted for over 65%\nof the 1994 UK industrial electricity usage. Asserting that more than 10% of this energy could be attributed to\ninefficiency, Falkner argued that this represented a loss of about \u00a30.5 billion to the annual UK economy. These values\naccounted for motor energy utilisation across multiple industrial sectors, but it should be recognised that the electric\nmotors in plastics industry processing machines are a major power consumers.\nA more detailed study by Rosato et al. in 2001 [34] observed that energy losses of between 3 and 20% can arise in\nthe transmissions and control systems. Despite this, a conclusion was made that because plastics have lower specific\nenergy requirements compared with most conventional raw materials, they are still highly competitive.\nFive years later, Womer et al. [35] considered the energy efficiency of extruder cooling. The results demonstrated\nthat water cooling systems consume more energy compared with air cooling, irrespective of the particular plastic being\nprocessed. As a result, a recommendation was made to use air only cooling unless extensive cooling was expressly\nrequired.\nIn 2010 [36], the plastics industry was recognised to be one of the major UK industries with a similar trend\napplying globally. On that basis any improvement in process energy efficiency would lead to a considerable reduction\nin global energy requirement. Also in 2010, Cantor [37] presented measurements of SEC, where the impact of the\nmotor and of each individual heater zone, with respect to the overall specific energy consumption, was separately\nrecorded. It was observed that the heaters account for over 95% of the supplied energy. In a slightly later study\nby Heur and Verheijen [38], the authors studied differences from one plant to another, and recommended the use of\nfrequency controllers to enable more precise process control.\nThe earliest mention of an Industry 4.0 implementation to energy efficiency control was by Jing et al. (2014) [39]\nwhich proposed the use of real-time monitoring. The rationale was to render unnecessary the installation of power\nmeters or the development of data-driven models. A fuzzy logic controller controlled the high melt quality in a single\nscrew extruder, and was shown to be a cheaper alternative to using a gear pump. This also paved the way for achieving\ngreater extruder energy efficiency by optimising the temperature settings.\nA number of other works [35, 40, 41, 42, 43, 44, 45] consider the drive motor efficiency compared with other\ndevices. The conclusion to be drawn is that the drive motor should be the primary design consideration for process\n6\n", "In 1981, Kruder and Nunn [29] claimed that energy efficiency of extruders can range from 45%-75%.", "Chung et al. [12] found that for a 63.5 mm diameter extruder mechanical energy efficiency of 62% was typical, and for larger extruders the energy efficiency was lower.", "Kruder and Nunn [29] claimed that energy efficiency of extruders can range from 45%-75%."]}, {"unit": "%", "value": 62, "evidence": ["enhancements to machinery could be identified, where an economic case could be made on the basis of energy cost\nsaving. Some examples are given below.\nIn the late 1970s, Chung et al. [12] found that for a 63.5 mm diameter extruder mechanical energy efficiency of\n62% was typical, and for larger extruders the energy efficiency was lower. In 1981, Kruder and Nunn [29] claimed\nthat energy efficiency of extruders can range from 45%-75%. It was noted that the energy efficiency depended on the\ntransmission mechanism, screw design, product geometry, nature of polymer feedstock and the production rate, while\nthe major energy losses of an extruder occur as a result of the forced cooling process step, and the losses associated\nwith the drive and transmission unit. At low screw speeds, barrel heaters consume a considerably higher portion of\nenergy than at higher speeds, and significant energy savings could be made by running the processes at the highest\npossible power factor. Additionally, this work presented information on energy demand and losses of each individual\ncomponent of an extruder.\nSubsequently through the 1980s, most research into energy efficiency was focussed on the screw efficiency and\nmass flow rate. A reduction in the overall power requirement for an extruder can be achieved through the use of a\ngeared pump at the end of the extruder to increase the mass flow rate (McKelvey [30]). In 1985, Strauch et al. [31]\ncarried out an energy consumption study on a 63.5 mm diameter single screw extruder, and observed that most of the\nenergy was consumed by the mechanical parts, with less significant levels of consumption in process heating. The\nenergy conversion was then assessed and it was found that heating the water in the cooling system accounted for more\nthan half of the energy supplied.\nDuring the late 1980s and 1990s, the manufacturing sector was making many changes, with a view to improving\nproductivity and quality. Driven by the advances made in Japanese manufacturing, the main focus during that time\nwas on management methods, such as Total Quality Management, LEAN, and Six Sigma. These efforts initially\naddressed cost and time issues, where the biggest economic benefits were to be found. Latterly, interest in energy\nefficiency began to be seen as not only cost reduction opportunity but also as an environmental imperative.\nIn the context of power consumption in the extruder, in 1997 Anderson et al. [32] recognised that for the processing\nof most plastics, from room temperature, the specific energy consumption (SEC) of the extruder motor should be in\nthe\nrange of 0.0822 to 0.1644 kW.hr/kg.\nAt around the same time, a study by Falkner in 1997 [33] showed that motor operations accounted for over 65%\nof the 1994 UK industrial electricity usage. Asserting that more than 10% of this energy could be attributed to\ninefficiency, Falkner argued that this represented a loss of about \u00a30.5 billion to the annual UK economy. These values\naccounted for motor energy utilisation across multiple industrial sectors, but it should be recognised that the electric\nmotors in plastics industry processing machines are a major power consumers.\nA more detailed study by Rosato et al. in 2001 [34] observed that energy losses of between 3 and 20% can arise in\nthe transmissions and control systems. Despite this, a conclusion was made that because plastics have lower specific\nenergy requirements compared with most conventional raw materials, they are still highly competitive.\nFive years later, Womer et al. [35] considered the energy efficiency of extruder cooling. The results demonstrated\nthat water cooling systems consume more energy compared with air cooling, irrespective of the particular plastic being\nprocessed. As a result, a recommendation was made to use air only cooling unless extensive cooling was expressly\nrequired.\nIn 2010 [36], the plastics industry was recognised to be one of the major UK industries with a similar trend\napplying globally. On that basis any improvement in process energy efficiency would lead to a considerable reduction\nin global energy requirement. Also in 2010, Cantor [37] presented measurements of SEC, where the impact of the\nmotor and of each individual heater zone, with respect to the overall specific energy consumption, was separately\nrecorded. It was observed that the heaters account for over 95% of the supplied energy. In a slightly later study\nby Heur and Verheijen [38], the authors studied differences from one plant to another, and recommended the use of\nfrequency controllers to enable more precise process control.\nThe earliest mention of an Industry 4.0 implementation to energy efficiency control was by Jing et al. (2014) [39]\nwhich proposed the use of real-time monitoring. The rationale was to render unnecessary the installation of power\nmeters or the development of data-driven models. A fuzzy logic controller controlled the high melt quality in a single\nscrew extruder, and was shown to be a cheaper alternative to using a gear pump. This also paved the way for achieving\ngreater extruder energy efficiency by optimising the temperature settings.\nA number of other works [35, 40, 41, 42, 43, 44, 45] consider the drive motor efficiency compared with other\ndevices. The conclusion to be drawn is that the drive motor should be the primary design consideration for process\n6\n"]}, {"unit": "kW.hr/kg", "value": 0.1233, "evidence": ["the specific energy consumption (SEC) of the extruder motor should be in the range of 0.0822 to 0.1644 kW.hr/kg.", "enhancements to machinery could be identified, where an economic case could be made on the basis of energy cost\nsaving. Some examples are given below.\nIn the late 1970s, Chung et al. [12] found that for a 63.5 mm diameter extruder mechanical energy efficiency of\n62% was typical, and for larger extruders the energy efficiency was lower. In 1981, Kruder and Nunn [29] claimed\nthat energy efficiency of extruders can range from 45%-75%. It was noted that the energy efficiency depended on the\ntransmission mechanism, screw design, product geometry, nature of polymer feedstock and the production rate, while\nthe major energy losses of an extruder occur as a result of the forced cooling process step, and the losses associated\nwith the drive and transmission unit. At low screw speeds, barrel heaters consume a considerably higher portion of\nenergy than at higher speeds, and significant energy savings could be made by running the processes at the highest\npossible power factor. Additionally, this work presented information on energy demand and losses of each individual\ncomponent of an extruder.\nSubsequently through the 1980s, most research into energy efficiency was focussed on the screw efficiency and\nmass flow rate. A reduction in the overall power requirement for an extruder can be achieved through the use of a\ngeared pump at the end of the extruder to increase the mass flow rate (McKelvey [30]). In 1985, Strauch et al. [31]\ncarried out an energy consumption study on a 63.5 mm diameter single screw extruder, and observed that most of the\nenergy was consumed by the mechanical parts, with less significant levels of consumption in process heating. The\nenergy conversion was then assessed and it was found that heating the water in the cooling system accounted for more\nthan half of the energy supplied.\nDuring the late 1980s and 1990s, the manufacturing sector was making many changes, with a view to improving\nproductivity and quality. Driven by the advances made in Japanese manufacturing, the main focus during that time\nwas on management methods, such as Total Quality Management, LEAN, and Six Sigma. These efforts initially\naddressed cost and time issues, where the biggest economic benefits were to be found. Latterly, interest in energy\nefficiency began to be seen as not only cost reduction opportunity but also as an environmental imperative.\nIn the context of power consumption in the extruder, in 1997 Anderson et al. [32] recognised that for the processing\nof most plastics, from room temperature, the specific energy consumption (SEC) of the extruder motor should be in\nthe\nrange of 0.0822 to 0.1644 kW.hr/kg.\nAt around the same time, a study by Falkner in 1997 [33] showed that motor operations accounted for over 65%\nof the 1994 UK industrial electricity usage. Asserting that more than 10% of this energy could be attributed to\ninefficiency, Falkner argued that this represented a loss of about \u00a30.5 billion to the annual UK economy. These values\naccounted for motor energy utilisation across multiple industrial sectors, but it should be recognised that the electric\nmotors in plastics industry processing machines are a major power consumers.\nA more detailed study by Rosato et al. in 2001 [34] observed that energy losses of between 3 and 20% can arise in\nthe transmissions and control systems. Despite this, a conclusion was made that because plastics have lower specific\nenergy requirements compared with most conventional raw materials, they are still highly competitive.\nFive years later, Womer et al. [35] considered the energy efficiency of extruder cooling. The results demonstrated\nthat water cooling systems consume more energy compared with air cooling, irrespective of the particular plastic being\nprocessed. As a result, a recommendation was made to use air only cooling unless extensive cooling was expressly\nrequired.\nIn 2010 [36], the plastics industry was recognised to be one of the major UK industries with a similar trend\napplying globally. On that basis any improvement in process energy efficiency would lead to a considerable reduction\nin global energy requirement. Also in 2010, Cantor [37] presented measurements of SEC, where the impact of the\nmotor and of each individual heater zone, with respect to the overall specific energy consumption, was separately\nrecorded. It was observed that the heaters account for over 95% of the supplied energy. In a slightly later study\nby Heur and Verheijen [38], the authors studied differences from one plant to another, and recommended the use of\nfrequency controllers to enable more precise process control.\nThe earliest mention of an Industry 4.0 implementation to energy efficiency control was by Jing et al. (2014) [39]\nwhich proposed the use of real-time monitoring. The rationale was to render unnecessary the installation of power\nmeters or the development of data-driven models. A fuzzy logic controller controlled the high melt quality in a single\nscrew extruder, and was shown to be a cheaper alternative to using a gear pump. This also paved the way for achieving\ngreater extruder energy efficiency by optimising the temperature settings.\nA number of other works [35, 40, 41, 42, 43, 44, 45] consider the drive motor efficiency compared with other\ndevices. The conclusion to be drawn is that the drive motor should be the primary design consideration for process\n6\n", "in 1997 Anderson et al. [32] recognised that for the processing of most plastics, from room temperature, the specific energy consumption (SEC) of the extruder motor should be in the range of 0.0822 to 0.1644 kW.hr/kg."]}]}, "metadata": {"product_category": "Machinery & equipment", "request_id": "req_2ced6848ba546c28"}}
{"id": "5410f7ef9964083a8205757e", "input": {"query": "What is the total electricity consumption in kWh per kg for cotton yarn spinning operations including all subprocesses like carding, drawing, and ring spinning? I need the gate-to-gate spinning energy only.", "source_url": "https://soeagra.com/iaast/iaastmarch2016/2.pdf", "document_text": "International Archive of Applied Sciences and Technology\nInt. Arch. App. Sci. Technol; Vol 7 [1] March 2016: 06-12\n\u00a9 2016 Society of Education, India\n[ISO9001: 2008 Certified Organization]\nwww.soeagra.com/iaast.html\nCODEN: IAASCA\nJAAST\nONLINE ISSN 2277-1565\nPRINT ISSN 0976 - 4828\nORIGINAL ARTICLE\nEnergy consumption and Carbon footprint of Cotton Yarn\nProduction in textile industry\nMadhuri Nigam\u00b9, Prasad Mandade\u00b2, Bhawana Chanana Sabina Sethi4\n\u00b9Department of Fabric and Apparel Science, Lady Irwin College, University of Delhi, India\n2 Dept. of chemical Engg.. Indian Institute of Technology, Powai. .Mumbai, Maharashtra 400076, India\n3 School of Fashion Design and Technology, Amity University, Mumbai campus, India\n4 Department of Fabric and Apparel Science, Lady Irwin College, University of Delhi, India\nEmail: madhurinigam@gmail.com\nABSTRACT\nTextile industry with its diverse and complex processes, poses multiple challenges when it comes to standardising and\nbenchmarking its various processes. This study has analysed energy inputs and related emissions of cotton yarn\nproduction, from cradle to factory gate boundary, including phases from farm, transport to ginning, ginning, transport\nto spinning plant and spinning and packaging. The study is conducted, according to International standards\nOrganisation (ISO) 14040 series of standards for Life Cycle Assessment. Obtaining relevant data for various phases was\namong the challenges addressed in this work, since efforts to compile life cycle inventory data for India are very recent.\nAlso, the relevant data are scattered across diverse sources, or simply not available in the open literature. Data has been\ncollected from various resources such as journal articles, ministry of agriculture data, Indiastat database, personnel\ncommunication, reports etc. All the embodied energy inputs for the cotton production, such as fertilizer, pesticides,\nelectricity, human and animal inputs, seeds and diesel are considered. Embodied energy values from this data shows that\nfarming has large variability in the inputs due to geographical variations and farming type and type of the farmers.\nDiesel and fertilizer input shares the maximum inputs, along with the electricity. Data pertaining to spinning is collected\nfrom personnel communication with a spinning mill in Uttarakhand district, India. Data for packaging inputs such as\nL.D.P.E., H.D.P.E., and cardboard box have been obtained from Ecoinvent database. Electricity is a major input in\nspinning process, with water being the second major input within plant. Current energy and emissions analysis of yarn\nproduction is expected to improve the supply chain by focusing on the phases which have the higher impact which in turn\nwill to enhance decision making in the textile production processes. The analysis revealed that sustainability of farming\nphase can be improved by using modern better management practices such as drip irrigation and use of organic farming\nto reduce the overall impacts.\nKey words: Carbon Footprint, Life Cycle Assessment, CO2eq. - Carbon dioxide equivalent, GHG emissions, Climate change\nReceived 02/01/2016\nCitation of this article\nRevised 12/01/2016\nAccepted 19/01/2016\nRouf ur Rafiq and D M Kumawat. Impact of Cement industry Pollution on Physio-morphological attributes of Apricot\ntree (Prunus armeniaca) around industrial belt Khrew, Kashmir. Int. Arch. App. Sci. Technol; Vol 7 [1] March 2016\n2015: 06-12. DOI.10.15515/iaast.0976-4828.7.1.612\nINTRODUCTION\nThe textile industry is one of the largest industrial sectors in the world. Its supply chain is diverse and\ncomplex, including design, raw material harvesting, spinning, yarn production, dyeing, weaving, cutting,\nstitching and final garment construction. Clothing and textiles contribute to approximately 10% of the\ntotal carbon emissions. Textile industry consumes\n9-10% of total energy available in India and accounts for 20% of total production cost. Thermal and\nelectrical energy demands are met using coal, firewood and electricity. Thermal energy requirement is\nderived from firewood, lignite, coal and fuel oil. Combustion of these fuels contributes direct emission of\nCO2 [1]. India is a major producer of cotton, and ranks 2nd in export of Cotton. It has a 59% share in\nconsumption of raw material in the Indian textile industry. The need for sustainability in textile industry\nis being increasingly emphasised due to associated environmental impacts on soil and water [3].\nCotton cultivation is generally known for its unsustainable nature, due to the overuse of fertilisers,\npesticides and water [4].Cotton production is both a contributor to and a \u2018victim' of climate change.\nAgricultural production, processing, trade and consumption contribute up to 40% of the world's\nIAAST Vol 7[1] March 2016\n6 | Page\n\u00a92016 Society of Education, India\nNigam et al\nemissions when forest clearance is included in the calculation. Cotton production contributes between\n0.3% and 1% of total global GHG emissions [5]. Present cotton growing practices, are not sustainable:\nthey damage soil, water and associated eco-systems, as well as contribute to extremely high social costs\nand a threat to regional economies depending on cotton farming and associated textile industries.\nHowever, cotton farming requires large amounts of water, (varying from 7-29 tons per kg of raw cotton\nfibres [6].\nModernisation of the textile industry is rather slow and a lot of manufacturers are still using old\ninefficient technology[7]. Energy consumption is on the rise due to modern machines and inefficient\nusage of equipment. The energy cost contributes 15-20% of production cost, next only to raw material\ncost [8]. However, textile energy studies make up a relatively small share of all industrial energy studies.\nMore energy studies in this sector will help to identify the energy efficiency potential for the industry\nitself as well as for the other similar industries.[9]\nParticulate matter (PM) is the primary air pollutant emitted from cotton ginning. [10] Ismail et al.30\nevaluated the energy usage and greenhouse gas (GHG) emission inside cotton gins in Australia. Results\nshowed the electricity and gas usage of 61% and 39% of total energy use, respectively. It was estimated\nthat 60.38 kg CO2 were produced by energy usage for ginning one bale of cotton. [11]. Hughs et. al (2013)\nhave calculated the energy requirement in U.S. gins to be in the range 33.07kWh/bale to\n41.37kWh/bale.[12]\nElectricity is the major type of energy used in spinning plants, especially in cotton spinning set ups. In the\nspinning plant electricity is a major energy source, which is used in humidification in cold weather.\nPresent study analyses the life cycle energy inputs and emissions related to production of 1 kg cotton\nyarn in its Cradle to Gate boundary. The study provides relevant recommendations for internal\nimprovements and decisions on pollution prevention, resource conservation, and waste minimization\nopportunities. Therefore, there is a need for detailed information of energy and water requirement and\nemissions profile of cotton yarn.\nMETHODOLOGY\nScope and Boundary of the analysis\nLCA includes all four stages of a product or process life cycle: raw material acquisition, manufacturing,\nuse/reuse/maintenance. This paper presents the energy inputs and emissions associated with cotton\nyarn production system in Indian context, in its Cradle to Gate boundary. Since cotton yarn is an\nintermediate product and it has a wide variety of applications, the scope of the study was limited to\nfactory gate. The cotton yarn production chain is divided into following processes (i) The agriculture\nphase includes fuels or energy-intensive material inputs such as fertilizers, herbicides, seeds, diesel fuel,\nand electricity for irrigation, machinery and labour used for agriculture.(ii) The transportation phase\nincludes moving of agricultural product from farm to the ginning site by truck, (iii) Ginning process which\ninvolves electricity consumption, and from there (iv) transport to spinning plants.(v) Conversion of fibre\nto yarn consists of mainly electricity and water for humidification (vi) Packaging. Life Cycle stages within\nthe boundary under consideration have been modelled in Figure1 illustrates the phase in the life cycle of\ncotton yarn. Retail Consumer Use and disposal phase have not been considered because of the\nintermediate nature and unavailability of data.\nLife-cycle system boundary: Cotton yarn\nResource inputs:\nfrom nature and\ntechnosphere\nEmission: to air\nCotton farming\nTransport\nGinning\nto ginning plant\nTransport to\nspinning plant\nSpinning\nPackaging\nEmission: to water\nat all stages\nPesticides\nFertilisers\nDiesel\nElectricity\nDiesel\nWater\nElectricity\nLDPE\nbags\nHDPE\nbags\nCardboard\nbox\nWater\nEmission: to land\nFigure 1: Analysis boundary of the cotton yarn production\nData Sources and Assumptions\nThis work faced challenges in obtaining relevant data, since data for India are still to be included in LCI\ndatabases. The fact, that the relevant data is scattered, and is not available in literature and databases,\nIAAST Vol 7[1] March 2016\n7 | Page\n\u00a92016 Society of Education, India\nNigam et al\ncompelled the data, to be acquired from various sources like journal articles, publications of Ministry of\nStatistics, Ministry of Agriculture, India Statistical databases. Farming data has wide spatial and temporal\nvariation across the country. The lack of completeness of data and ignorance about the resources used in\nthe farming phase regionally necessitates quantifying the variability. The data from diverse sources have\nbeen included in this study, and the mean, mode, median, high value, and low values for each input and\nyield have been calculated. Energy inputs need to be allocated between cotton and cottonseed as\ncottonseed is a by-product and used for edible oil extraction and the residue is further used in animal\nfodder. In the absence of primary data secondary information was obtained from literature, previous\nstudies, and reports of environment and audit agencies. Results from some foreign studies have been\nused in absence of Indian data leading to the assumption that the energy consumption scenarios are the\nsame between India and the country from which data is taken.\nFunctional unit:\nFunctional unit of analysis, for present study is 1 kg of cotton yarn at spinning unit. As the yarn can have\nvarious applications, we have analysed the life cycle impact of 1 kg of cotton yarn till factory gate\nboundary. The analysis can be extended based on its using further processing.\nAllocation:\nAllocation is to partition the input or output flows of a process or a product system under study and one\nor more other product systems. The inputs and outputs shall be allocated to the different products\naccording to clearly stated procedures that shall be documented and explained together with the\nallocation procedure [14]. Cotton production yields two valuable outputs, namely cotton fibre and cotton\nseed. Various allocation methods are available for life cycle studies\n1.\n2.\nMass based\nSubstitution and\n3. Economy based\nThus, the environmental burden can be allocated to respective products at different stages but present\nstudy does not compute effects of allocation.\nRESULTS AND DISCUSSION\nThe metrics calculated in this work are energy inputs and resulting GHG emissions related to cotton yarn\nproduction, which have been described in the rest of this section. This metric will help in evaluating and\ncomparing the life cycle GHG emissions of various cotton yarn systems. All the phases of production\nleading to emissions are highlighted, so that focus on the specific inputs can be given to improve the\noverall sustainability by reducing the energy inputs or by trying alternative inputs or processes.\nEnergy and emissions analysis of cotton farming:\nFarming energy and emission analysis has been carried out to find out which inputs contributes more for\nenergy and emissions at farming stage so that we can focus on specific inputs to improve the\nsustainability of the farming process. Figure 2 shows the percentage energy distribution at the farming\nstage for the mean values of the Inputs. From the figure, it is evident that fertilizer consumes major\nportion of the energy inputs along with diesel used for farming operation and electricity for irrigation.\nFarming energy distribution for cotton production in\nIndia(MJ/ha)\nIAAST Vol 7[1] March 2016\n39%\n6%\n9%\n2%\n4%\n3%\n14%\n11%\n\u25a0 Human\n\u25a0 Electricity\n\u25a0 Fertilizers\n\u25a0Canal\n3%\n9%\n\u25a0 Animal\nDiesel\nSeeds\n\u25a0 Chemicals\nFarmyard manure\n\u25a0 Machinery\nFigure 2: GHG emissions from Cotton farming\n8 | Page\n\u00a92016 Society of Education, India\nNigam et al\nFarming emission analysis shows that the farming stage has 2.1076 kg CO2 eq/kg cotton yarn without\nallocation and 0.8430 CO2eq/kg cotton yarn on mass basis and 1.8336 CO2eq/kg cotton yarn on economy\nbasis. The contribution of emissions from the various inputs has revealed that for the cotton production\nthe highest contribution is from the N2O Emissions from fertilizer application. Along with this, fertilizer\nproduction, diesel used and electricity also contribute largely, which suggests that, there is a need of\nimprovement in the irrigation practices as well as minimal use of organic fertilizers.\nGHG (in CO2 Equivalent) emission for cotton farming\n8%\n7%\n29%\n11%\n21%\n0%\n3%\n\u25a0 Diesel\n\u25a0Fertilizers\n\u25a0 Electricity\n\u25a0Chemicals\n21%\n\u25a0Machinary\n\u25a0N20 from N Fertilizer application\n\u25a0Canal\n\u25a0CO2 from N fertilizer application\nFigure 3: GHG emissions from Cotton farming\nAs the data has been collected from various sources and regions of India, to capture the data variation of\ncotton farming stage, we have calculated the mean high and low value inputs and outputs which have\nbeen used in further calculation to include the sensitivity of the variation.\n12000\n10000\nInputs(MJ/ha) output Yield(kg/ha) for Cotton Farming\n8000\n6000\n4000\n2000\nHuman\nAnimal\nDiesel\nElectricity\nSeeds\nFarmyard manure\nFertilizers\nChemicals\nMachinery\nCanal\nYield (kg/ha)\nFigure 4. Inputs and Outputs of Cotton Farming\nEnergy and emissions analysis of transport from farming to ginning plant:\nDistance of transportation of harvested cotton from farm to ginning plants is estimated to be 200 km\nroundtrip. The energy inputs of transport have been calculated to be 0.21933 MJ/kg cotton yarn without\nallocation. 0.087733 MJ/kg cotton yarn with mass allocation and 0.190725 cotton yarn with economy\nallocation cotton yarn which leads to an emission of kgCO2 eq/kg cotton yarn.\nThe analysis has revealed that the transport farming to ginning plant leads to 0.015083 CO2eq/kg cotton\nyarn without allocation and 0.0060 CO2eq/kg cotton yarn on mass basis and 0.013115 CO2eq/kg cotton\nyarn on economy basis.\nEnergy and emissions analysis of Ginning:\nIt is well known that as much as 60-70% of seed is available from seed cotton during ginning. Average\nenergy inputs required for ginning were calculated for the functional unit. Manpower required to process\nIAAST Vol 7[1] March 2016\n9 | Page\n\u00a92016 Society of Education, India\nNigam et al\n1 kg of cotton is calculated to be 0.002 man h/kg. Electrical energy required for the processing of 1 kg of\nyarn is 0.5863 MJ/kg.\nBased on electrical energy consumed for ginning GHG emissions are calculated to be 0.131025 CO2eq/kg\nCotton yarn [15]. Indian national emission factors have been used for the analysis.\nEnergy and emissions analysis of transport for spinning plant:\nTransport data has been obtained by communicating with the firm personnel from Uttarakhand district\nbetween 2012-2015 periods. Transport of ginned cotton, plastic and paper cone and packaging material\nlike H.D.P.E., L.D.P.E. and Cardboard box are considered. Employee transport for local and business\ntransport have been considered. The energy inputs from transport are found to be 1.7894 MJ/kg Cotton\nyarn. Of this major portion is contributed by transport of raw material which is 1.709 MJ/kg Cotton yarn.\nThe emission analysis shows that the transport of raw materials and employee transport related to\nspinning plant is 0.1452 CO2eq/kg cotton yarn. This analysis does not include Air and rail transport.\nEnergy and emissions analysis of spinning:\nFor the spinning process of cotton yarn, only electrical power is important for the LCA calculation (the\nmaintenance of the machine can be neglected, as well as the making of it) [16]. The total electrical energy\nconsumption is 8.84 MJ/kg cotton yarn. Out of this 97.7% of electrical power is from grid and 2.3% is\nfrom DG set. Figure 5 illustrates the percentage distribution of total electricity in various phases of\nspinning, within the plant. It takes up 44% of energy in spinning step, consisting of both Ring and Rotor\nspinning, followed by Pre-draw, Lap and combing which consume 16% of energy. This shows that\nmaximum electricity is consumed in spinning phase. Humidification and air conditioning takes 16% of the\nenergy.\nElectricity consumption distribution in spinning plant\n1%\n1% -3%\n2%\n9%\n14%\n9%\nOpening, cleaning mixing\n\u25a0Carding\n16%\n44%\n2%\nI Predraw, lap, comb\n\u25a0Drawing\n\u25a0Roving- Ring/Rotor spin\nI Humidification\nConditioning\nStorage\nFigure 5: Distribution of electricity emissions in various stages of spinning\nThe GHG emissions have been attributed to grid electricity and electricity generated from DG set. The\nanalysis shows that grid electricity contributes to 2.137 kgCO2eq/kg of cotton yarn and diesel\ncombustion is responsible for 0.090 kgCO2eq/kg cotton yarn. Figure 3 shows the contribution of grid and\nDG electricity to total GHG emissions. The emission analysis shows that the spinning emits GHGs which is\n2.2273 CO2eq/kg cotton yarn.\nEnergy and emissions analysis of Packaging\nThe total energy inputs from packaging have been found to be 0.1549 MJ/kg of cotton yarn produced.\nThis major energy input of packaging has been contributed by cardboard (0.1432 MJ/kg cotton yarn),\nHDPE bag (0.0075 MJ/kg cotton yarn), Paper cones 0.0038 MJ/kg cotton yarn) and LDPE packaging\n(0.0002 MJ/kg cotton yarn).\nThe energy and emission analysis of packaging material has revealed that this phase contributes to\n0.0075 CO2eq/kg cotton yarn and cardboard has the highest contribution to GHG emissions. Reuse and\nrecycling scenarios of packaging materials have not been considered in this analysis.[17]\nBased on mass allocation the Life Cycle analysis of cotton yarn production, Figure 6 shows that Life cycle\nGHG emissions for cotton yarn are highest in spinning stage. It is followed by farming phase which is the\nnext highest contributory phase. The next most important phases are ginning and transport to spinning\nplant.\nIAAST Vol 7[1] March 2016\n10 | Page\n\u00a92016 Society of Education, India\n5\n4.5\n4\n3.5\n3\n2.5\n2\n1.5\n1\n0.5\n0\nMass\nNigam et al\nEconomic\nWithout allocation\nPackaging\n\u25a0Spinning\n| Transportation to\nspinning\n\u25a0Ginning\n\u25a0Transportation from\nfarming to ginning\n\u25a0Farming\nFigure 6: Distribution of electricity in various stages of life cycle stages of cotton yarn\nRECOMMENDATIONS\nThe life cycle analysis of cotton yarn production has revealed that manufacturers need to focus on\nspinning and farming phases in order to reduce overall emissions from cotton yarn production. The\nmanufacturers can reduce the carbon footprint in spinning stage by the following measures.\n\u2022\n\u2022\n\u2022\n\u2022\nElectrical meters to quantify the optimal power consumption based on output of machine.\nUse of DG waste heat will lead to reduced thermal and air pollution as less flue gases of high\ntemperature will be released.\nInstallation of Solar power system to meet partial load during peak or daytime\nOverall equipment efficiency of machine based on ideal cycle time.\nMachine modification or technological upgradation\nImproving the loading of motors, as the quantum of energy saving will depend on the extent of\nloading.\nEfficient blowers should be used for higher CFM.\nThe analysis revealed that sustainability of farming phase can be improved by using modern better\nmanagement practices such as drip irrigation and use of organic farming to reduce the overall impacts.\nReduced fertiliser application will substantially control the GHG emissions.\nThe prospect of energy recovery from cotton ginning waste can lead to substantial emission reduction.\nAlmost half the plants thermal requirements can be met by a bioenergy unit in the plant.\nTransport emission contribution can be reduced by sourcing raw materials from nearby suppliers and\nusing larger vehicles to reduce the number of trips.\nCONCLUSION\nThere is a general consensus among scientific community that cotton is a better option as it is a natural\nfibre. Therefore, there is a need for detailed information of energy and emissions profile of cotton yarn.\nResults for the Cotton yarn production pathway aim to highlight the life cycle GHG emissions in Indian\ncontext. This is the first life cycle study of Indian textile products/processes pathways. Findings from the\ncurrent emissions analysis of yarn production are expected to improve the supply chain by focusing on\nthe phases which have the higher impact, which in turn will help to enhance decision making in the textile\nproduction processes. This study will give a benchmark for comparing cotton yarn with other fibre yarns.\nNOMENCLATURE\nLCI\nLife Cycle Inventory\nMJ/kg - Mega Joules/kg\nCarbon dioxide equivalent\nLCA\nLife Cycle Assessment\nCO2eq\nISO\n- International Standards Organization\nREFERENCES\n1.\nR. Velavan, R. Rudramoorthy, and S. Balachandran, (2009). \u201cCO2 Emission Reduction Opportunities for Small and\nMedium Scale Textile Sector in India,\u201d J. Sci. Ind. Res., vol. 68, no. July, pp. 630\u2013633.\nIAAST Vol 7[1] March 2016\n11 | Page\n\u00a92016 Society of Education, India\n2.\nNigam et al\nC.Prakash, T.Maruthavanan, and C. Parvathi, (2009). \"Environmental impacts of textile industries,\u201d Indian Text.\njournal., vol. CXVII, no. 2, pp. 22-26.\nT. Publication and H. a S. B. Published, (2011). \"Cotton Market and Sustainability in India.\"\n3.\n4.\nB. Jeffries, (2013). \u201cCutting cotton carbon emissions ~ Findings from Warangal, India,\u201d.\n3456\n5.\n6.\n7.\n8.\n9.\nP. Ton, A. Asterine, and M. Knappa, \u201cCotton and Climate Change- Impacts and Options to Adapt,\u201d 2011.\nE. M. Kalliala and P. Nousiainen, \"Life Cycle Assessment Environmental profile of cotton and polyster-cotton\nfabrics,\" AUTEX Res. J., vol. 1, no. 1, pp. 8-20, 1999.\nM. S. Bhaskar, P. Verma, A. Kumar, and N. Delhi, \u201cBureau of Energy Efficiency, Ministry of Power, New Delhi,\"\nvol. 4, no. 2231, pp. 36-39, 2013.\nY. Dhayaneswaran and L. Ashokkumar, \u201cA Study on Energy Conservation in Textile Industry,\u201d J. Inst. Eng. Ser. B,\nvol. 94, no. 1, pp. 53-60, Aug. 2013.\nA. Agha and D. P. Jenkins, \u201cEnergy analysis of a case-study textile mill by using real-time energy data,\u201d pp. 223-\n231.\n10. S. Virginia, N. Carolina, and S. Carolina, \"EPA - Ginning Regulations,\" October, pp. 1\u20139, 1995.\n11. S. H. Pishgar-Komleh, P. Sefeedpari, and M. Ghahderijani, \u201cExploring energy consumption and CO2 emission of\ncotton production in Iran,\" J. Renew. Sustain. Energy, vol. 4, no. 3, 2012.\n12. P. A. Funk, R. G. H. Iv, S. E. Hughs, and J. C. Boykin, \u201cChanges in Cotton Gin Energy Consumption Apportioned by\n10 Functions,\" vol. 183, pp. 174-183, 2013.\n13. T.-M. Choi, \u201cSupply Chain Management in Textiles and Apparel,\" J. Text. Sci. Eng., vol. 02, no. 02, 2012.\n14. ISO/TC/207, \u201cEnvironmental management - Life Cycle Assessment - Principles and Framework,\" Int. Organ.\nStand., vol. 1997, 2006.\n15. P. A. Funk, R. G. H. Iv, S. E. Hughs, and J. C. Boykin, (2013). \"Changes in Cotton Gin Energy Consumption\nApportioned by 10 Functions,\" vol. 183, pp. 174\u2013183.\n16. N. M. Van Der Velden, M. K. Patel, and J. G. Vogtl\u00e4nder, (2014). \"LCA benchmarking study on textiles made of\ncotton, polyester, nylon, acryl, or elastane,\u201d Int. J. Life Cycle Assess., vol. 19, no. 2, pp. 331-356.\n17. D. C. Edwards and J. M. Fry, (2006). \"Life Cycle Assessment of Supermarket Carrier Bags: A review of the bags\navailable in 2006,\" Environmental Agency Bristol.\nIAAST Vol 7[1] March 2016\n12 | Page\n\u00a92016 Society of Education, India\n"}, "expected_output": {"claims": [{"unit": "kWh/kg", "value": 8.84, "evidence": ["Transport data has been obtained by communicating with the firm personnel from Uttarakhand district\nbetween 2012-2015 periods. Transport of ginned cotton, plastic and paper cone and packaging material\nlike H.D.P.E., L.D.P.E. and Cardboard box are considered. Employee transport for local and business\ntransport have been considered. The energy inputs from transport are found to be 1.7894 MJ/kg Cotton\nyarn. Of this major portion is contributed by transport of raw material which is 1.709 MJ/kg Cotton yarn.\nThe emission analysis shows that the transport of raw materials and employee transport related to\nspinning plant is 0.1452 CO2eq/kg cotton yarn. This analysis does not include Air and rail transport.\nEnergy and emissions analysis of spinning:\nFor the spinning process of cotton yarn, only electrical power is important for the LCA calculation (the\nmaintenance of the machine can be neglected, as well as the making of it) [16]. The total electrical energy\nconsumption is 8.84 MJ/kg cotton yarn. Out of this 97.7% of electrical power is from grid and 2.3% is\nfrom DG set. Figure 5 illustrates the percentage distribution of total electricity in various phases of\nspinning, within the plant. It takes up 44% of energy in spinning step, consisting of both Ring and Rotor\nspinning, followed by Pre-draw, Lap and combing which consume 16% of energy. This shows that\nmaximum electricity is consumed in spinning phase. Humidification and air conditioning takes 16% of the\nenergy.\n", "The total electrical energy consumption is 8.84 MJ/kg cotton yarn. Out of this 97.7% of electrical power is from grid and 2.3% is from DG set.", "For the spinning process of cotton yarn, only electrical power is important for the LCA calculation (the maintenance of the machine can be neglected, as well as the making of it) [16].", "Functional unit of analysis, for present study is 1 kg of cotton yarn at spinning unit."]}]}, "metadata": {"product_category": "Textiles, footwear & apparel", "request_id": "req_1d2e8482f64c2c52"}}
{"id": "2121c6e36f7e3cb9cc02a123", "input": {"query": "What is the fuel mix (percentage of energy from fuels like natural gas and fuel oil, divided by total energy inclusive of electricity) for US flat glass plants?", "source_url": "https://www.energystar.gov/sites/default/files/tools/Industry_Insights_Flat_Glass_2017.pdf", "document_text": "ENERGY STAR\nENERGY STAR\u00ae\nINDUSTRIALINSIGHTS\u2122\nUS Flat Glass Plants\nNAICS 327211\n25 Flat Glass Plants1\n7 Companies1\n16 States & Territories with Plants1\n0 ENERGY STAR Plants\nEnergy Use Profile\nFlat Glass Manufacturing\nFlat glass manufacturing plants process sand, soda ash, and other raw materials into sheet\nand float glass for building products, automotive applications, tabletops, mirrors, and other\nuses. It is the second largest segment of the glass industry in the United States (based on\nproduction) and is one of the most energy-intensive industrial processes. Additionally, it is a\nsector where energy costs comprise a significant percentage of operating costs.\nThe US Environmental Protection Agency's ENERGY STAR partnership has worked with\nthe glass industry since 2006 to promote energy efficiency and energy management best\npractices within the sector through the ENERGY STAR Industrial Focus initiative.\nNatural gas and electricity are the dominant energy sources\nused in flat glass manufacturing.2\nEnergy Use Distribution\n31%\n69%\nDistribution of Energy Performance\nEPA, through the ENERGY STAR Glass Manufacturing\nIndustrial Focus, has benchmarked the energy performance of\nflat glass plants. The curve below, generated from the ENERGY\nSTAR Flat Glass Plant Energy Performance Indicator (EPI)\nbenchmarking tool, shows the normalized distribution of energy\nperformance for a representative plant. A dashed line\ncorresponding to the performance of an average plant is\nprovided for reference. An Energy Performance Score (EPS) of\n75 or higher is defined by EPA as the threshold for efficient\nplants.\n100\n90\n80\n\u25a0 Fuels \u25a0 Electricity (Source Btus)\nElectricity and natural gas use vary by plant size and product\nmix. The table below provides an estimate of total energy use for\neach plant size category.\u00b3\nEnergy Performance Score\nPlant Energy Use\nTotal Energy (MMBtu)\n10\nSmall\n~534,000\n0\nMedium\n~1,597,500\n20\n40\n60\n100\nMMBtu per $1,000 of Product Shipped\n80\n120\nLarge\n~3,115,000\nFuels are the largest energy cost, representing 64% of total\nenergy costs. In 2013, flat glass plants spent over $218 million\non fuels and over $125 million on electricity.4\nElectric Costs (36%)\nFuel Costs (64%)\nThis curve reveals a narrow range of energy performance\nbetween the majority of plants. The largest energy efficiency\nopportunities are in plants in the lowest percentiles of energy\nperformance (<30%). For these plants, the greatest efficiency\ngains will likely be made during scheduled rebuilds of glass\nfurnaces. Increasing the use of cullet (recycled glass) can\nimprove the energy performance of all plants.\nMajor Energy Uses\nFlat glass manufacturing is an energy-intensive process that\ninvolves melting raw materials at extreme heat followed by\nmultiple finishing processes. The table below outlines major\nenergy using processes.5\nUse / Process\nBatch (raw material) preparation\nMelting and refining furnaces.\nForming\nFinishing (tempering, coating, etc.)\nShare of Energy\n3%\n63%\n15%\n19%\nFurnaces used to melt and refine raw materials into glass are\ndesigned to operate continuously for 8 to 10 years before\nstopping for major maintenance and rebuilding. Consequently,\nsignificant efficiency improvements to furnaces usually occur\nonly during rebuilds. Increasing the use of cullet (recycled glass),\nwhich requires much less energy to melt than other raw\nmaterials, will improve efficiency. In furnaces, smaller efficiency\ngains can be made by optimizing oxygen levels, upgrading\nburners, and increasing insulation. Compressed air is used\nextensively in the forming and finishing phases, making it a focus\narea for energy management.\nENERGY STAR Resources\nThe ENERGY STAR Glass Manufacturing Focus, a collaborative\neffort between EPA and the industry, has developed the\nfollowing materials for energy efficiency in flat glass plants:\n\u2022\nEnergy Performance Indicator (EPI): Benchmarks and rates plant\nenergy performance.\nEnergy Guide: Technical guidance on energy saving opportunities.\nENERGY STAR Certified Plants\nEPA's ENERGY STAR program certifies flat glass plants that\ndemonstrate energy performance in the top quartile nationally\nusing the Flat Glass Plant Energy Performance Indicator (EPI).\nGreenhouse Gas (GHG) Emissions\nDirect GHG emissions from flat glass plants are produced from\nboth fuel use and the chemical reactions in the glass making\nprocess. Twenty-five plants reported direct emissions to EPA's\nGreenhouse Gas reporting program in 2013, totaling over 3.1\nmillion metric tons of CO2e (MMTCO2e). 6 As shown below,\nemissions ranged from 241,977 to 63,163 metric tons of CO2e\n(mtCO2e) and averaged around 124,526 mtCO2e.6\nMetric Tones CO\u2082e\nDirect Emissions Distribution\n300,000\n250,000\n200,000\n150,000\n100,000\n50,000\n0\n5\n10\nIndirect emissions\nfrom\n15\n20\n25\nFlat Glass Plants\nelectricity purchases were\napproximately 1.1 MMTCO2e in 2013.7\nTotal GHG emissions from flat glass plants were approximately\n4.2 MMTCO2e in 2013.7,8 Just over half of the GHG emissions\nin flat glass plants are from fuels (natural gas) used for heating\nand melting processes, as shown in the graphic below.\nGHG Emissions by Source\n24%\n56%\nReferences:\n1. Plant, company, and state counts from the EPA FLIGHT Database (ghgdata.epa.gov).\n2. Fuels from 2013 EPA FLIGHT Database (ghgdata.epa.gov). Electric from 2013 Annual\nSurvey of Manufacturers.\n3. 2010 Manufacturers Economic Census Survey, Table 6.3.\n4. 2013 Annual Survey of Manufacturers.\n5. ENERGY STAR Energy Efficiency Improvement and Cost Saving Opportunities for the\nGlass Industry.\n6. EPA Greenhouse Gas Reporting Program Database (ghgdata.epa.gov).\n7. Estimate calculated from purchased electricity reported in the 2013 Annual Survey of\nManufacturers.\n8. Estimate calculated by combining direct emissions from the EPA Greenhouse Gas\nReporting program with estimates from the 2013 Annual Survey of Manufacturers.\nJuly 2017\n20%\n\u25a0Fuels\n\u25a0 Process Emissions\n\u25a0 Electricity\nEPA\nUnited States\nEnvironmental Protection\nAgency\n"}, "expected_output": {"claims": [{"unit": "percent", "value": 69, "evidence": ["69%", "\u25a0 Fuels \u25a0 Electricity (Source Btus)", "Natural gas and electricity are the dominant energy sources used in flat glass manufacturing.2", "Energy Use Distribution"]}]}, "metadata": {"product_category": "Metal, mineral, plastic & glass products", "request_id": "req_78bc254026f923ba"}}
{"id": "20fb08d235e22196cac01826", "input": {"query": "What is the energy consumption for secondary aluminum remelting and casting? Please provide: Energy in GJ per tonne or MJ per kg.", "source_url": "https://aluminiumtoday.com/content-images/news/BIR.pdf", "document_text": "A Report on the environmental\nbenefits of recycling\n- A critical review of the data for aluminium\nThe Bureau of International Recycling (BIR) commissioned Imperial College, London to obtain the energy\nrequirements and carbon footprint impact for the production of primary and secondary metals. Data\npresented in the report enables an estimated saving of 136MtCO2 in 2006 to be derived by recycling\naluminium scrap based on mean energy figures, but that primary production generates 80% more CO2\nwhen compared with primary production of steel. By Editor Aluminium International Today\nThe Report, 'Report on the environmental\nbenefits of recycling' prepared by the Centre\nfor Sustainable Production and Resource\nEfficiency (CSPRE) of Imperial College\nLondon, provides an extensive review of\nenergy requirements and associated CO2\nemissions for the production of aluminium,\nsteel, copper, lead, nickel, tin and zinc -\nalthough surprisingly not magnesium.\nThe data brings together numerous sources\nto present energy requirements for primary\nproduction of these metals from their ores\nand secondary production from recycled\n(scrap) metal and uses these results to\ncalculate an average for the carbon footprint\nof each process route.\nThis article reviews the data for aluminium\nproduction and finds some to be at odds with\nthe industrial accepted norms and draws on\ncomments from representative organisations\nof these industries in an attempt to reconcile\nthe data. A comparison is also made with\ndata from the Report for primary and\nsecondary production of steel.\nMuch of the anomaly between the report's\nconclusions and the industrial accepted\nvalues arise from the limited data sources\nused in the report and the lack of any\nweighting to account for the global share of\nvarious production methods cited when\ncalculating mean values.\nText taken from the BIR Report is\npresented in italics and comments are made\nin Roman text. All Tables use data taken from\nthe BIR Report except Tables 3 and 8 which\nare from the International Aluminium\nInstitute (IAI). Also Tables 9 and 10 are\nderived from the BIR data but are not\npublished in the format presented in this\narticle. For ease of reference, the tables are\nreferred to by number in the text although\nthe BIR report does not number the tables.\nPrimary aluminium\nThe gross energy requirement for primary\naluminium production is estimated at 120MJ/kg\nAl based on using hydroelectricity with 89%\nenergy efficiency. As alternatives to\nCoal (c.e. 35%)\nSource\nMJ/kg Al\nNotes\nNorgate\n211\nNorgate\n150\nGas (c.e. 54%)\nNorgate\n120\nHydro (c.e. 89%)\nCambridge\n260\nAus Alu Council\nGrant\nChoate and Green\n(c.e.-refers to conversion efficiency)\n182-212\n207\n133\nCoal (c.e. 35%)\nCoal (c.e. 35%)\nCoal (c.e. 35%)\nUS average\nTable 1 Energy requirements Bayer Hall H\u00e9roult route\n(alumina and aluminium production)\nElectricity benchmark\nSource\nSchwarz\nMJ/kg Al\nNotes\n47\nIAI\n54\nElectricity average\nNorgate\n66\nElectricity max\nNorgate\n46\nIAI\n69\nCambridge\n55\nCambridge\n160\nCambridge\n50\nChoate and Green 56\nElectricity benchmark\nElectricity max\n95% Hydro efficiency\n35% Coal efficiency\n100% efficient\nUS average\nTable 2 Energy requirements of Hall H\u00e9roult electrolysis\nprocess alone\nStage\nMining\nTypical Energy\n0.15\nRefining\n16\n1.9\nMultiplier Total Energy\n5\n0.75\n30\nAnode\nSmelting\nTotal Primary\n9\n0.44\n4\n117\n1.02\n120\n155\nRemelting\n10\n1\n10\nTable 3 IAI global energy averages for primary and\nsecondary (remelted) aluminium production (MJ/kg)\nSource IAI\nhydroelectricity, use of black coal for electricity\ngeneration with an efficiency of 35% or natural\ngas with an efficiency of 54% would give gross\nenergy estimates of approximately 211 and\n150MJ/kg Al respectively. The data in the\nfollowing table (Table 1) are the gross energy\nrequirements that have been quoted in various\npublications for production of primary\n4 ALUMINIUM INTERNATIONAL TODAY BUYERS' DIRECTORY 2010\naluminium by the Bayer-Hall H\u00e9roult route, (ie\ntaking the energy to refine alumina from bauxite\nore into account) along with the assumptions\nthat the authors made on the fuel used.\nThe electricity consumption in the Hall\nH\u00e9roult process is the most energy-demanding\naspect of primary production of aluminium. The\nenergy requirements reported in the literature for\nthe Hall H\u00e9roult process alone (ie for conversion\nof treated ore to metal) are in Table 2 along\nwith the assumptions made on the fuel used.\nElectrolysis alone\nThe Report accepts that the electricity\nconsumption in the Hall H\u00e9roult process\nalone (ie the electrolysis of alumina dissolved\nin cryolite) is the most energy-demanding\naspect of primary production of aluminium.\nComparing Tables 1 & 2, there is\nsometimes a discrepancy between the\nReport's statement that the electrolysis stage\nis the most energy intensive part of the\nprocess. For example, taking the US average\nfigure given by Choate & Green from Table 1\nfor alumina plus aluminium production of\n133MJ/kgAl and subtracting from this their\nUS average figure of 56MJ/kgAl for the\nelectrolysis stage alone (Table 2) we obtain\n77MJ/kgAl for the refining stage (Bayer\nprocess) alone, ie 58% of the total energy\nrequirement.\nFor alumina (not aluminium) production,\nthe International Aluminium Institute (IAI)\nreports a weighted average energy\nconsumption of 11.7MJ/kgAl2O3 (see\nhttp://stats.world-aluminium.org/iai/stats_\nnew/formServer.asp?form=8). This averaged\nvalue is derived from the performance of 80%\nof the world's refiners constituting low\ntemperature digestion (Av 10.9 MJ/kgAl2O3),\nhigh temperature digestion (Av\n13.7MJ/kgAl2O3) and Bayer-Sinter (Av 27.9\nMJ/kgAl2O3). Plants treating Nepheline ores\nare excluded. Taking the approximation that,\nby mass, it requires 1.9 times the amount of\nalumina to produce one unit of aluminium\nmetal, this equates to 22.2 MJ/kgAl. The BIR\nreport figure of 77MJ/kgAl for the alumina\nstage of production is thus 3.46 times higher\nthan the IAI which reports values ranging\nfrom 10.74 to 14.64 MJ/kg Al2O3 (~20.4 -\n27.8MJ/kgAl metal).\nBacking up the IAI evidence is data from\nUC Rusal's Aughinish refinery in County\nLimerick, Ireland which reports a total\nenergy requirement of 10.5MJ/kg of alumina\nproduced of which 6.6MJ/kg is used in the\ndigestion process, 3.3MJ/kg in calcining and\n0.7MJ/kg for plant power (see AIT Nov/Dec\n2007 P 40).\nThe European Aluminium Association\n(EAA) reports an average thermal\nenergy\nrequirement of 10GJ/t of alumina plus\n230kWh/t alumina (64MJ) making a total of\n10.064GJ/Al2O3 (=MJ/kg).\nIn some other examples in the BIR report,\ngreater energy is seen to be required for the\nelectrolysis stage, but in all cases the energy\nrequirement for refining is far higher than\nthat stated by the IAI. For example, if we\ncompare the worst case reported in Table 1\nwhich is for coal based power generation at\n35% efficiency as reported by Cambridge we\nobtain: 260 - 160 = 100MJ/kgAl for the\nalumina stage of production, ie less than the\nelectrolysis stage as expected but over three\nand a half times greater than the IAI worst\ncase reported of 14.64MJ/kgAl2O3 (ie\n27.8MJ/kgAl metal).\nIn correspondence, the authors of the BIR\nreport claim the IAI figure to be an under\nestimate and consider their sources to be\nmore reliable. The IAI in turn dispute this,\npointing out that their data is collected\ndirectly from refineries which account for\napproximately 80% of global production and\ndates back to 1985. Also, close to 100% of US\nrefineries\nreport their\nenergy consumptions\nto IAI. The only major region absent from the\nIAI figures is China. The BIR data they point\nout is from secondary sources.\nThe data from the commercial operating\nrefinery at Aughinish backs the IAI figure as\nbeing the more realistic as does the EAA data.\nThe author of the BIR report confirms that\nthe data in Table 1 excludes energy required\nto mine and transport the bauxite, which,\nanyway \u0399\u0391\u0399 say, accounts for less than 0.5% of\nthe total energy to produce aluminium.\nIndeed, the IAI calculates a global average\ntotal energy of 155MJ/kg of aluminium\nproduced taking into account mining,\nrefining, anode production, and smelting.\nEach contributor is given a weighting related\nto the volume consumed eg x5 for mining\nsince it takes about four units of bauxite to\nproduce one of aluminium metal plus\ntransporting this quantity from the mine to\nthe refinery. The IAI calculation is presented\nin Table 3. This table is not presented in the\nBIR report.\nDespite the much higher BIR figure at the\nrefining stage the average total energy\nrequirement reported in Table 1 of 133MJ/kg\nfor the US average is just 1.4% less than the\nIAI global figure of 155MJ/kg (Table 3).\nIn addition, the weighted average energy\nSource\nNorgate\nAnode, Low Temp\nElectrolyte, Natural\nAnode, Low Temp\nElectrolyte,\nHydroelectricity 89%\nAverage IAI\ntCO2/t Al Energy Source\nSource\ntCO2/t Al\nNotes\n22.4\nCoal\nNorgate\n7.2\nDrain Cathode, Inert\nGrant\n18.2\nCoal\nKvande\n24\nCoal\nIAI\n20\nCoal\nGas 54%\nIAI\n9.8\nHydro 57%,\nNorgate\n4.6\nDrain Cathode, Inert\nCoal 28%, Natural\nGas 9%, Nuclear 5%,\nOil 1%\nChoate and Green\nChoate and Green\n9.11\nUS Average\nIAI\n5.48\nInert Anode, Wetted\nCathode, ACD 2cm\nChoate and Green\n8.56\nCarbothermic\nReaction\nproduction only\nChoate and Green\n6.71\nWetted Cathode and\nProcess\nMean in MJ/kg\nBenchmark in MJ/kg\nACD of 2cm\nRemelting\n4.5\n2.1\nChoate and Green\n8.95\nChloride Reduction\nof Kaolinite Clays\nCasting\nTotal\n0.5\n0.3\n5.0\n2.4\nTable 4 Carbon footprint Bayer Hall H\u00e9roult route for\nalumina plus primary production of aluminium\nrequirement for electrolysis to the metal\nalone reported by IAI is similar to that\nreported in the BIR Report. IAI reports\nenergy consumption in terms of kWh/t Al\nproviding a weighted global average across\nregions of 15384kWh per metric tonne\n(15.38kWh/kg). Using a conversion factor of\n1MJ = 3.6kWh this equates to 55.4MJ/kg\nclose to the 56MJ/kg stated in the BIR Report\nas the US average (Table 2), but is some 9MJ\nhigher (19%) than the 47MJ/kg electricity\nbenchmark figure presented in that Table\nwhich the BIR Report authors use to\ncalculate the carbon footprint. In reality, the\nIAI average figure of 15.38kWh/kg Al is\nbettered by modern high amperage smelters\nwhich can achieve power consumptions of\nthe order of 13kWh/kg equating to 47MJ/kg\nin line with the BIR reported benchmark\nfigure in Table 2.\nThe European Aluminium Association\nreport an average electricity consumption for\nEuropean smelters of 14.914kWh/kg\naluminium with a range of 13 to 18kWh/kg.\nIn terms of MJ/kgAl the average equates to\n53.6MJ/kg. The calculation takes into\naccount: - Rectifying loss; - DC power usage;\n- Pollution control equipment; - Auxiliary\npower (general plant use); and - Electric\ntransmission losses of 2%.\nSurprisingly the BIR Report does not\nattribute any data from the EAA despite that\nAssociation publishing a very extensive 72\npage report on all stages of aluminium\nproduction and fabrication. One reference to\nthe EAA is provided in the bibliography.\nEmissions\nThe BIR Report states: For the purpose of\ncomparison of the energy requirements and\nassociated carbon emissions for primary\naluminium production with data for secondary\naluminium production, the Report assumes that\nthe benchmark process would involve an\nelectricity benchmark figure of about 47MJ/kg.\nThe literature data on the carbon footprint for\n7.7\nUS Average (Typical)\nChoate and Green 3.83\nTable 5 Carbon footprint Hall H\u00e9roult aluminium\nTable 6 Energy requirement of secondary processes for\naluminium production from scrap\nprimary production of aluminium following the\nBayer-Hall H\u00e9roult route and for the Hall\nH\u00e9roult process alone are given in Tables 4 & 5\nrespectively, along with the assumptions made\nby the authors on the fuel used.\nThe CO2 emissions presented in the BIR\nReport for alumina plus aluminium\nproduction (Table 4) are in line with the\nindustrial accepted average norm of close to\n10tCO2 per tonne aluminium. As reported in\nTable 4, the IAI, for example, estimate a\nvalue of 9.8tCO2/tAl based on a power\ngeneration mix of Hydro 57%, Coal 28%,\nNatural Gas 9%, Nuclear 5%. The US average\nis given as 9.11tCO2/tAl. The European\naverage (not quoted) is reported by EAA as\n8.566tCO2/tAl in 2005. Of this, 1.804t is\ngenerated in the cell from consumption of\nthe anode and PFC CO2 equivalent\nemissions, 4.584t from electricity generation\nand 1.758t as thermal energy. Auxiliary\ndemands and transport account for a further\n353 and 68kgCO2/tAl respectively.\nEmissions for the electrolysis stage alone\nreported in Table 5 are low for the US\naverage at 3.83tCO2/tAl compared with the\nIAI average of 7.7tCO2/tAl. The data quoted\nfor Norgate using inert anodes (a process not\nyet commercially developed) would expect to\nsee a reduction of the order of one tonne less\nCO2/tAl assuming a typical net carbon\nconsumption in a conventional pre-baked\ncarbon anode of around 400kg/tAl producing\n32/12 x 1.06tCO2.\nSecondary production\nThe BIR Report says: It has been reported that\nthe production of one tonne aluminium from\nscrap requires only 12% of the energy required\nfor primary production. Energy savings of\nbetween 90 and 95% have also been reported for\nsecondary aluminium production compared with\nprimary production, starting with mining the ore\nand not with as-received concentrate.\nThe energy requirement to recycle aluminium\nhas been calculated at between 6 and 10MJ/kg\nALUMINIUM INTERNATIONAL TODAY BUYERS' DIRECTORY 2010\n5\nLQ\nassuming efficiencies of 60-80% in the recycling\nprocess.\nThe energy requirement data for secondary\naluminium production are reported in Table 6\nas mean values for melting and casting and\nbenchmark values for melting and casting. The\ncarbon footprint data presented in Table 7 have\nbeen calculated on the basis of these energy\nrequirement data, using the carbon emission\nfactor for the UK.\nThe IAI average energy consumption for\nremelting scrap is 10GJ/t (10MJ/kg) (Table 3)\nwhich accords with the range of values stated\nin the text of the BIR Report but is\nsignificantly above the mean values quoted in\nTable 6 of 5.0MJ/kg (melting plus casting)\nand well above the benchmark value of\n2.4MJ/kg which the authors use to calculate\nthe carbon footprint.\nBIR report summary\nIn its summary findings, the Report presents\nits benchmark findings per 100kt of\naluminium produced as:\nEnergy requirement (for 100kt) primary\nproduction: 4700TJ (therefore 47GJ/t)\n-\n\u2013 Energy requirement (for 100kt) secondary\nproduction: 240TJ (therefore 2.4GJ/t)\nConfirming the industrial accepted value of a\n95% saving in energy by the secondary\nremelting route.\nUsing this energy data, the carbon\nfootprints for primary (electrolysis stage\nonly) and secondary production of\naluminium on the same basis are:\n- Carbon footprint for primary production:\n383kt CO2 (3.83tCO2/Al)\nCarbon footprint for secondary production:\n29kt CO2 (0.29tCO2/tAl)\nRepresenting a 92% saving in CO2 emissions\nbetween the two methods.\nHowever, in terms of absolute emissions\nthe figure for primary production\ncorresponds only to the US typical average\nfigure (Table 5) and is well below the other\nvalues quoted in Table 5 eg IAI = 7.7t\nCO2/tAl. Also, this figure is for the\nelectrolysis stage of smelting only. If the\nrefining of bauxite to alumina is also\nincluded, the US average rises to 9.11t and\nthe IAI figure to 9.8t for hydro generated\npower and 20t for coal generation (Table 4).\nThe IAI figure of 9.8tCO2/tAl is an 'ore to\nmetal' figure taking into account emissions\ncontributed by mining, refining, anode\nproduction, smelting and casting and\nincludes a weighting factor for each stage of\nproduction eg 1.9 for alumina production\nsince on average it requires 1.9 units of\nalumina to produce one unit of aluminium. It\nProcess\nCO2 Mean\nCO\u2082 Benchmark\nCO\u2082/tAl\ntCO2/t\nRemelting\n0.54\n0.25\nCasting\n0.06\nTotal\n0.6\n0.04\n0.29\nTable 7 Carbon footprint for the secondary processes for\nthe production of aluminium from scrap\nBauxite\nAlumina\nAnode\nPrimary\nPrimary\nTotal Mine\nMining\nRefining\nProduction\nSmelting\nCasting\nto Ingot(2)\nProcess(1)\n0\n0\n402\n1557\n0\n1763\nElectricity(3)\nFossil Fuel\n1\n64\n66\n5225\n42\n5529\n4\n707\n150\n0\n82\n1530\nPFCs\nTotal\n0\n0\n0\n970\n0\n989\n5\n771\n617\n7752\n125\n9812\nMult Factor\n5.272\n1.923\n0.435\n1.02\n1.00\nTable 8 Contribution of CO2 equivalent emissions for each stage of aluminium production (kgCO2/t of product)\nSource IAI\nNotes: (1) Contribution at process stage eg for Primary smelting CO2 and CO2 equivalents arising from net carbon\nconsumption of anode + CO2eq from fluoride emissions (PFCs are detailed separately)\n(2) Sum of each production stage after multiplying by its respective contributing factor\n(3) Hydro 57%, Coal 28%, Nat Gas 9%, Nuclear 5%, Oil 1%.\nOutput 06\nShare (%)\nCO\u2082/t\nTotal O\u2082/t\nCO\u2082Saving/t\n34.0\n68\n9.11\n309.74\n16.0\n32\n0.29\n4.64\n8.82\nPrimary\nSecondary\n(benchmark)\nSecondary (Mean)\n16.0\n32\n0.60\n9.6\nTotal & CO2\n50.0\n8.51\n141.1 bench\n136.2 mean\nSaving (Mt)\nTable 9 CO2 saving resulting from secondary production\nalso includes the CO2 equivalent emissions\narising from PFC emissions during smelting\nas well as the contribution from the different\nfuels employed ie electricity and fossil fuel.\nTable 8 from the IAI -which is not included\nin the BIR Report - summarises the\ncontribution of each of these inputs.\nThe secondary production route footprint\nof 0.29tCO2/tAl represents the benchmark\nfigure in Table 7 and is only about half the\naverage o.6otCO2 emissions to remelt and\ncast also presented in Table 7.\nPrimary v Secondary Al\nThe BIR Report estimates that in 2006, 16Mt\nof secondary aluminium was produced from\nscrap and 34Mt of primary metal. Dross\nlosses are estimated at 2.5% for secondary\nproduction (in Europe) and 2-4% for\nprimary production.\nThe 34Mt primary output may be a little\nhigh. IAI reports primary production from\ndata covering 64% of total production of\n24Mt in 2007 equating to a global primary\noutput 32.64Mt. The most significant\nabsence from IAI production data is China\n(and some other less significant regions).\nChina is estimated to produce close to\n10Mt/y of primary metal.\nThe secondary production figure is more\nlikely to be an underestimate as it cannot\naccount for 'unrecorded' remelters which are\ncommon in parts of the world such as India\nand China. An estimate made at the Alcastek\n2008 conference puts unrecorded secondary\nproduction in India to be over 0.5Mt, or\nnearly half the total secondary production.\n(See AIT May/June 2008 P44).\nUsing the BIR Report data, close to one\nthird of the total 50Mt of aluminium\nproduced in 2006 was from secondary\nsources. Using their Carbon footprint figures\n(Tables 4 & 7) a saving of 141.1MtCO2 was\nachieved that year if the benchmark figure is\nused for secondary production or 136.2Mt if\nthe mean value is taken, just 3.5% more than\nthe benchmark (Table 9).\nIt is worth noting that moving from the\nmean energy value to the benchmark value\nresults in a saving of just 3.4% in CO\u2082\nemissions or 4.9 MtCO2 a year.\nCarbon footprint Al vs Steel\nThe BIR report also compares steelmaking\nfrom ore by the blast furnace (BF) oxygen\nsteelmaking (BOF) route and the Direct\nReduced Iron (DRI) Electric Arc Furnace\n(EAF) route as well as from melting scrap in\nthe EAF.\nAs with some of the data for aluminium,\nthere are figures presented for steel which are\nnot consistent with the industry norm, in\nparticular the BIR\ndata which suggests\nenergy\nthat there is only a 16% saving in energy\nwhen melting scrap in an EAF compared\nwith the BF-BOF route. The impact of the\nsaving in terms of the carbon footprint is\nmore in line with the industry accepted\nfigure, the BIR Report concluding a 35%\nreduction (Table 10) compared with the\nindustry's estimate of 39% as estimated\nbelow:\nIn its report '2008 Sustainability Report of\nthe world steel industry' the World Steel\nAssociation (formerly International Iron &\nSteel Institute) - whose members represent\n85% of total world production - state: 'More\nsteel is recycled worldwide annually than all\nother materials put together, with an\nestimated 459Mt being recycled in 2006,\nabout 37% of the crude steel produced that\nyear. Recycling this steel avoided 827Mt of\nCO2 emissions, saved 868Mt of iron ore, and\nsaved the energy equivalent of 242 Mt of\n6\nALUMINIUM INTERNATIONAL TODAY BUYERS' DIRECTORY 2010\nanthracite coal.' They also conclude that each\ntonne of crude steel produced 1.7tCO2 on\nweighted average (69% BOF 30% EAF,\n1%OH). In 2006, 1.25bnt of crude steel was\nproduced thus emitting 2.125bnt CO\u2082. Thus\nrecycling of scrap resulted in a saving of\n827/2125 =38.9%.\nThe significant effect on the carbon\nfootprint of recycling scrap is evident and in\naddition there is a substantial reduction in\nCO2 emitted due to the removal of 868Mt of\nore and 242Mt of hard coal from the\nprocessing route.\nThe BIR Report assumes a much lower\nenergy requirement for the BF - BOF route\nof 14GJ/t for their carbon footprint\ncalculation despite some of their own data\nshowing an average of 21.9GJ/t (with a\nstandard deviation of 5.1) as total energy\nrequirement from the ore.\nThe World Steel Association in their 2008\nSustainability Report for the 2006 Fiscal Year\npresents an average energy intensity value of\n20.6GJ/t of crude steel produced. This is a\nweighted average including both the BF-BOS\nroute and the EAF route from 38 member\ncompanies and two industry associations\n(including a further 77 companies) with 70%\nBOF, 29% EAF and 1% OHF production\nroute spread. Together, these companies\nproduced 42% of the crude steel output\nworldwide in 2006. Unfortunately, the data is\nnot broken down between the BF-BOF and\nEAF routes but, since, by the BIR Report's\nown figures, the average energy requirement\nfor the EAF route alone is 11.7GJ/t, the BF-\nBOF contribution to the average must be\ngreater than the average 20.6GJ/t and be\nestimated in the order of 22.3GJ/t.\nmay\nBased on their energy figures of 14GJ/t for\nBF-BOF and 11.7GJ/t for the EAF route, the\nBIR Report attributes the mean carbon\nfootprint per tonne of steel to be 1.97t CO2\nfor the BF-BOF route and 0.70tCO2 for the\nEAF route (Table 10).\nComparing the carbon footprints of\naluminium and steel, BIR Report data shows\nthat primary aluminium production emits\n7.14t more CO2/t metal than steel but 0.41t\nCO2/t less for the secondary route (Table 10).\nThe generally accepted industry figure for\nthe BF-BOF route alone is close to 2tCO2/t\nAluminium\nPrimary\n(Bayer &HH)\nPrimary (HH)\nSecondary\nMax Min Mean\n22.4 5.48 3.83\nNote\nUS Av\n7.7 3.83 7.7\n0.60 0.29 0.29\nUS Av\nBenchmark\nSteel\nPrimary\nBF+BOF\n2.30 1.32 1.97\n(SD 0.30)\nPrimary\n3.31 0.7 1.76\n(SD 0.96)\nDRI+EAF\nSecondary\n(scrap)\n1.18 0.54 0.70\n(SD 0.27)\nTable 10 Comparison of carbon footprint for production\nof aluminium and steel (tCO2/tmetal)\nsteel and the BIR carbon footprint is thus in\nreasonable agreement for the BF-BOF\nprimary route for steel production and\nsomewhat conservative for aluminium\nprimary production accepting a mean figure\nof 9.11tCO2/t some 7% lower than the IAI\naverage of 9.8tCO2/t.\nComparing the representative Association\nfigures for aluminium and steel production of\n9.8 and 2.0 respectively we must conclude\nthat primary aluminium production has a\ncarbon footprint nearly 80% greater than that\nof steel, but in contrast, using the BIR figures\nfor secondary production, there is a 41%\nreduction in CO2 emissions per tonne of\nmetal produced when melting aluminium\nscrap compared with steel scrap.\nIn terms of volume production, since the\ndensity of aluminium is 2.70 compared with\nthat of steel of 7.87, approximately two-thirds\ngreater volume of aluminium results per\ntonne compared to steel. Thus by volume, the\ncarbon footprint for primary production of\naluminium reduces to 9.11 x 0.33 =\n3.otCO2/m\u00b3 while that for steel remains 1.97\ntCO2/m\u00b3 ie the difference falls to 34%.\nHowever, it should be noted that the lower\nyield strength and modulus of aluminium\nrequires thicker sections than an equivalent\nsection in steel to achieve the same load\nbearing capacity, hence replacement of steel\nby aluminium is not on a one for one basis.\nIn balance, the BIR Report has come up\nwith the broadly accepted conclusions of\nindustry regarding the advantages of\nrecycling but by using selected data based on\nbenchmark rather than mean industrial\nvalues.\nThe 'Report on the Environmental benefits of\nRecycling' is available from the Bureau of\nInternational Recycling (BIR), Avenue Franklin\nRoosevelt 24, 1050 Brussels, Belgium.\nTel +32 2 627 5770 Fax +32 2 627 5773\nemail bir@bir.org, website www.bir.org\nSources\nFrom BIR Report\n1) Blomberg, J and Hellmer, S Short-run demand and\nsupply elasticities in the West European market for\nsecondary aluminium Resources Policy 26 (2000) 39-\n50\n2) Xiao, Y, Reuter, MA Recycling of distributed\naluminium turning scrap Minerals Engineering, 15/11\n(2002) 963-970\n3) Developments in Mineral Processing, 16 (2005)\n391-451\n4) Energy efficiency and greenhouse gas reduction\npotentials in the aluminium industry A workshop in\nthe framework of the G8 Dialogue on climate change,\nclean energy and sustainable development\nInternational Aluminium Institute 2007\n5) Aluminium story International Aluminium\nInstitute Online access 05 April 08 http://www world-\naluminium org/About+Aluminium/Story+of\n6) Samuel, M A new technique for recycling\naluminium scrap Journal of Materials Processing\nTechnology 135/1 (2003) 117-124\n7) Schwarz, H G Technology diffusion in metal\nindustries: driving forces and barriers in the German\naluminium smelting sector Journal of Cleaner\nProduction, 16/1 (2008) 37-49\n8) Coal to remain top energy source for China\nShenzhen Daily/Agencies Updated: 2005-03-30 [Cited\n04/04/08] http://www chinadaily com cn/\nenglish/doc/2005-03/30/content_429544 htm\n9) Life Cycle Inventory data for aluminium\nproduction and transformation processes in Europe\nEnvironmental Profile Report for the European\nAluminium Industry European Aluminium\nAssociation [cited 15/04/08]\nwww eaa net/upl/4/en/doc/ EAA_Environmental_\nprofile_report_Mayo8 pdf\n10) Norgate, T E, Rankin, WJ Greenhouse gas\nemissions from aluminium production a life cycle\napproach CSIRO Minerals [cited 14/04/08]\nAvailable online: www minerals csiro\nau/sd/CSIRO_Paper_LCA_Al htm\n11) Aluminium production: an example calculation of\nthe energy saved by recycling Department of\nMaterials Science and Metallurgy, University of\nCambridge Online accessed: 23 Feb 2008\nwww doitpoms ac uk/tlplib/recycling-\nmetals/aluminium_production php\n12) Norgate, T E Metal recycling: An assessment using\nLife Cycle Energy Consumption as a sustainability\nindicator (Report), CSIRO Minerals, 2004\n13) Life Cycle Assessment of aluminium: inventory\ndata for the primary aluminium industry Year 2005\nupdate International Aluminium Institute\nSeptember 2007\n14) Norgate, T E, Jahanshahi, S and Rankin, W J\nAssessing the environmental impact of metal\nproduction processes Journal of Cleaner Production\n15/8-9 (2007) 838-848\n15) Logozar, K Radonjic, G and Bastic, M\nIncorporation of reverse logistics model into in-plant\nrecycling process: A case of aluminium industry\nResources, Conservation and Recycling 49/1\n(2006) 49-67\n16) Choate, WT and Green, J A US Energy\nrequirements for aluminium production: Historical\nperspective, theoretical limits and new opportunities\nUS Department of Energy, Energy efficiency and\nrenewable energy February 2003\nFrom Reviewer\nIAI\n17) IAI Statistics http://www.world-aluminium.org/\nStatistics/Current + statistics\n18) IAI Life Cycle Assessment of Aluminium:\nInventory Data for the Primary Aluminium Industry\n(2005 Update)\nhttp://world-aluminium.org/cache/flo000166.pdf\n19) IAI Carbon Footprint Guidance Document\nhttp://world-aluminium.org/cache/fl0000169.pdf\n20) Aluminium for Future Generations/2008 update\nwww.world-aluminium.org\nEAA\n21) 'Environmental Profile Report for the European\nAluminium Industry' April 2008. European\nAluminium Association www.eaa.net/\nworldsteel\n22) 2008 Sustainability Report of the world steel\nindustry'\nwww.worldsteel.org/pictures/publicationfiles/Sustaina\nbility%20Report%202008_English.pdf\nALUMINIUM INTERNATIONAL TODAY BUYERS' DIRECTORY 2010 7\n"}, "expected_output": {"claims": [{"unit": "MJ/kg", "value": 2.4, "evidence": ["assuming efficiencies of 60-80% in the recycling\nprocess.\nThe energy requirement data for secondary\naluminium production are reported in Table 6\nas mean values for melting and casting and\nbenchmark values for melting and casting. The\ncarbon footprint data presented in Table 7 have\nbeen calculated on the basis of these energy\nrequirement data, using the carbon emission\nfactor for the UK.\nThe IAI average energy consumption for\nremelting scrap is 10GJ/t (10MJ/kg) (Table 3)\nwhich accords with the range of values stated\nin the text of the BIR Report but is\nsignificantly above the mean values quoted in\nTable 6 of 5.0MJ/kg (melting plus casting)\nand well above the benchmark value of\n2.4MJ/kg which the authors use to calculate\nthe carbon footprint.\n"]}, {"unit": "MJ/kg", "value": 5, "evidence": ["assuming efficiencies of 60-80% in the recycling\nprocess.\nThe energy requirement data for secondary\naluminium production are reported in Table 6\nas mean values for melting and casting and\nbenchmark values for melting and casting. The\ncarbon footprint data presented in Table 7 have\nbeen calculated on the basis of these energy\nrequirement data, using the carbon emission\nfactor for the UK.\nThe IAI average energy consumption for\nremelting scrap is 10GJ/t (10MJ/kg) (Table 3)\nwhich accords with the range of values stated\nin the text of the BIR Report but is\nsignificantly above the mean values quoted in\nTable 6 of 5.0MJ/kg (melting plus casting)\nand well above the benchmark value of\n2.4MJ/kg which the authors use to calculate\nthe carbon footprint.\n"]}, {"unit": "MJ/kg", "value": 6, "evidence": ["The energy requirement to recycle aluminium has been calculated at between 6 and 10MJ/kg", "assuming efficiencies of 60-80% in the recycling process.", "The BIR Report says: It has been reported that\nthe production of one tonne aluminium from\nscrap requires only 12% of the energy required\nfor primary production. Energy savings of\nbetween 90 and 95% have also been reported for\nsecondary aluminium production compared with\nprimary production, starting with mining the ore\nand not with as-received concentrate.\nThe energy requirement to recycle aluminium\nhas been calculated at between 6 and 10MJ/kg\n"]}, {"unit": "MJ/kg", "value": 10, "evidence": ["The energy requirement to recycle aluminium has been calculated at between 6 and 10MJ/kg", "assuming efficiencies of 60-80% in the recycling process.", "The BIR Report says: It has been reported that\nthe production of one tonne aluminium from\nscrap requires only 12% of the energy required\nfor primary production. Energy savings of\nbetween 90 and 95% have also been reported for\nsecondary aluminium production compared with\nprimary production, starting with mining the ore\nand not with as-received concentrate.\nThe energy requirement to recycle aluminium\nhas been calculated at between 6 and 10MJ/kg\n"]}]}, "metadata": {"product_category": "Metal, mineral, plastic & glass products", "request_id": "req_95bdcd7c940f3e66"}}
{"id": "21f461f37b2be45cc349d9d5", "input": {"query": "What is the energy consumption in kWh per kg for film extrusion processes? Provide the specific energy consumption values for extrusion of plastic films.", "source_url": "https://tangram.co.uk/wp-content/uploads/BPF-Energy-Management-in-Plastics-Processing.pdf", "document_text": "BPF\nEnergy\nManagement\nin Plastics\nProcessing\nA Signposting\nGuide by\nThe British\nPlastics\nFederation\nTANGRAM\nTECHNOLOGY\nConsulting\nEngineers\nBPF\nENERGY\nSponsored by\nWritten By\nDr Robin Kent\nWHAT\nEnergy Management in Plastics Processing\nSignposting Guide\nForeword\nThis Guide has been commissioned by the British Plastics Federation in order to assist\ncompanies in meeting energy efficiency targets associated with the Federation's Climate\nChange Agreement. It is brought to you by BPF Energy the company managing the\nAgreement.\nThis guide has been designed as a starting point for companies looking at reducing their\nenergy usage and will provide companies with an initial list of projects and actions to\ntake.\nFor a more detailed guide, the BPF commissioned 'Controlling Energy Use in Plastics\nProcessing' (also written by energy expert Dr Robin Kent) is available free of charge to\nall BPF Members and companies taking part in the Federation's Climate Change\nAgreement.\nPeter Davis\nDirector-General\nBritish Plastics Federation\nIntroduction\nThis guide is designed to provide British plastics processors with a 'first primer' in\nenergy management. It gives the first actions that processors should take to reduce\nenergy use and associated costs. Using this guide, plastics processors should be able to\nreview their operations and take some basic actions to reduce their energy use.\nThis guide is not an exhaustive list of projects or actions, it is designed to stimulate\ncompanies to action. A more comprehensive list of projects is given in the larger BPF\nEnergy publication 'Controlling energy in plastics processing' (available from the BPF -\nwww.bpf.co.uk) and it is recommended that companies obtain a copy of this.\nEnergy management is potentially one of the most cost-effective actions that a\ncompany can take to reduce both carbon emissions and costs. There is no conflict,\nimproving environmental performance and reducing costs are synergistic. It is not only\npossible to be 'green' and save money, becoming 'green' almost always reduces costs.\nDr Robin Kent\nTangram Technology Ltd.\nHitchin\nAugust 2011\nNote:\nWhilst all reasonable steps have been taken to ensure that the information contained within this guide is correct, the\ncontent is necessarily general in nature.\nAccordingly, BPF Energy and Tangram Technology Ltd. can make no warranties or representations of any kind as to\nthe content of this guide and, to the maximum extent permitted by law, accept no liability whatsoever for the same\nincluding without limit, for direct, indirect or consequential loss, business interruption, loss of profits, production,\ncontracts, goodwill or anticipated savings.\nAny person making use of this guide does so at their own risk.\n\u00a9 BPF Energy 2011 (Issue 1).\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n2\nEnergy management\nWhy do we need energy management?\nEnergy costs are rising and there is no reason to believe that they will decrease in the\nfuture. This is driven by factors such as:\n\u2022 Increasing use of taxation and other financial instruments.\n\u2022\nIncreasing supply and distribution shortages.\n\u2022 Decreasing security of supply.\nIncreasing importance of environmental issues and the public perception of these.\n\u2022 Increasing importance of corporate social responsibility.\nFor many plastics processing sites, energy costs are approaching the cost of direct\nlabour and energy costs are almost always higher than the actual profits of the site.\nExperience shows that for typical sites, where little action has been taken in the past,\nover 30% of the energy use is 'discretionary' - this means that the cost is incurred\nbecause the site management has either decided to take no action or because it has not\nrecognised the opportunities for improvement. In most cases, energy use and costs can\nbe reduced by over 30% and these savings add directly to the site profits.\nWhere are the savings?\nEnergy costs can be reduced by:\nManagement actions that typically cost less than\n\u00a31,000.\n30% energy cost savings\nMaintenance actions that typically cost less than\n\u00a31,000.\n\u2022 Capital investment actions that typically cost\nmore than \u00a31,000.\nManagement\n(10%)\nMaintenance\n(10%)\nCapital\ninvestment\n(10%)\nWhat are the returns?\nThe returns from energy management actions are quick, certain and need only internal\neffort. The payback for almost all management and maintenance actions is 6-9 months\nand for almost all capital investment actions is less than 4 years.\nThe returns from energy management are much better than the returns from increasing\nsales.\nWhat do we need to start?\nImplementing energy management requires an energy management system and this\nmust cover:\n\u2022\nThe company energy policy.\nPerformance assessment.\n\u2022 Targets for short and long-term performance.\n\u2022 Reporting Systems must show results to get resources.\n\u2022 Auditing.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n3\nBase and variable loads\nEnergy management requires both measurements and an understanding of the process.\nThe measurements are very simple to obtain and can come from most standard\naccounts packages.\nThe measurements need very little treatment to give vital information on the site and\nprocess operations.\nEnergy use is not fixed and uncontrollable, it is variable and controllable and is directly\nrelated to the production volume of the site.\nThe Performance Characteristic Line (PCL)\nThe basic information is the Performance Characteristic Line (PCL).\nGet energy and production volume data\nfor at least 12 months. Use a\nspreadsheet to plot the energy use for\nthe month (kWh) versus the production\nvolume for the month (kg) as a scatter\nchart and use the same spreadsheet to\ninsert a linear best fit trend line to\ngenerate to PCL.\nThe equation for the PCL gives the 'base'\nand the 'process' loads of the site and\nthe correlation coefficient (R2) indicates\nhow well the PCL fits the data.\nThe base load\nEnergy use (kWh)\n800,000\n700,000\n600,000\n500,000\n400,000\n300,000\n200,000\n100,000\nBase load\nBase and variable loads (injection moulding)\nIntersection with axis = 152,440 kWh\nProcess load = Slope of line = 1.5751 kWh/kg\nCorrelation coefficient = 0.9397\nkWh = 1.5751 x Production volume + 152,440\nR\u00b2 = 0.9397\n0\n0\n50,000 100,000 150,000 200,000 250,000 300,000 350,000 400,000\nProduction volume (kg)\nThe base load of a site is the intersection of the best fit line with the vertical axis. It is\nthe 'energy overhead' and for plastics processing will range between 20 and 40% of the\ntotal load of the site.\nA low base load generally indicates good management control of energy at the site and\na high base load generally indicates poor management control of energy at the site.\nReducing the base load is easy to carry out, low cost and has rapid payback. Savings in\nthe base load are very profitable because the base load is largely a dead weight that is\nunrelated to production output.\nThe process load\nThe process load of a site is the slope of the best fit line and is the energy needed to\nrun the process. Reducing the process load is more difficult to achieve because it\ngenerally (but not always) requires more fundamental process improvements. The\nprocess load depends on the type of process being used at the site.\nTypical process loads are:\n\u2022 Injection moulding: 0.9 to 1.6 kWh/kg.\n\u2022 Extrusion: 0.4 to 0.6 kWh/kg.\n\u2022\nExtrusion blow moulding: 2.0 to 2.6 kWh/kg.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n4\nPerformance and budgets\nEnergy consumption, savings and budgets need to be expressed in terms that the\naccounts function can recognise and deal with.\nAssessing performance\nThe PCL can be used to assess a site's performance:\nSet up a spreadsheet to calculate the predicted kWh\nfor a given production volume using the PCL.\nProduction\nvolume\nkWh\n\u00a3/month\n0\n152,440\n\u00a315,244\n50,000\n231,195\n\u00a323,120\n100,000\n309,950\n\u00a330,995\n\u2022\nDetermine the volume of material processed in the\nmonth (through the nozzles) and calculate the\npredicted energy usage.\n150,000\n388,705\n\u00a338,871\n200,000\n467,460\n\u00a346,746\n250,000\n546,215\n\u00a354,622\n\u2022 Determine the actual energy usage for the month.\n300,000\n624,970\n\u00a362,497\n\u2022\nCompare the predicted energy usage to the actual\nenergy usage.\n350,000\n703,725\n\u00a370,373\n400,000\n782,480\n\u00a378,248\n\u2022\nIf the actual energy usage is less than the predicted energy usage then the site\nperformed better than it has done historically - find out what the site did right and\ndo more of it.\nIf the actual energy usage is more than the predicted energy usage then the site\nperformed worse than it has done historically - find out what the site did wrong and\ndo less of it.\nWeekly data collection gives faster feedback to production departments on how to\nimprove.\nFor plastics processing, it is best to use `kg' as a reference for the production volume\nbut it is also possible to use 'parts' or any other convenient measure of production\nprovided the product mix doesn't vary greatly.\nRecording kWh/kg as a process efficiency measure on a monthly basis will lead to\ninaccurate conclusions due to the effect of production volume changes.\nBudgeting\nThe PCL can also be used to provide a\nmodel for energy budgeting.\nMonth\nForecast\nproduction\nForecast\nenergy use\nForecast\nenergy cost\nvolume (kg)\n(kWh)\n(\u00a3)\nUse the forecast sales volume to produce a\nforecast production volume (in kg) and use\nthe PCL to predict the energy usage.\nThis links energy use directly to the sales\nbudget and to the accounting system.\nIntegrating energy use into the accounting\nsystem is a vital element of controlling\nenergy costs.\nJanuary\n182,421\n439,772\n\u00a343,977\nFebruary\n197,897\n464,148\n\u00a346,415\nMarch\n248,742\n544,233\n\u00a354,423\nApril\n204,228\n474,120\n\u00a347,412\nMay\n212,716\n487,490\n\u00a348,749\nJune\n225,239\n507,214\n\u00a350,721\nJuly\n217,864\n495,598\n\u00a349,560\nAugust\n207,615\n479,454\n\u00a347,945\nSeptember\n347,845\n700,331\n\u00a370,033\nOctober\n343,468\n693,436\n\u00a369,344\nNovember\n311,174\n642,570\n\u00a364,257\nDecember\n147,378\n384,575\n\u00a338,458\nTotals\n2,846,588\n6,312,941\n\u00a3631,294\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n5\nMonitoring and targeting\nTargeting and cost drivers\nMonitoring and targeting is used to set targets based on the PCL of the site. Simply\nassessing performance provides an incentive for improvement but setting targets\nprovides a better incentive for improvement.\nTargets can be set on the basis of simple charts, e.g. CUSUM charts are very sensitive\nto changes in performance and a 'challenging but achievable' performance target can be\nset from the data used to generate the PCL. This is based on the best possible historic\nperformance of the site.\nTargeting energy use needs an understanding of what drives energy use. Energy use\ncan be 'activity' driven (by production volume) or `condition' driven (generally by the\nweather). Measuring and understanding the relevant cost drivers allows cost\nassignment to the relevant areas and ownership of the costs can be created.\nEnergy costs are not `somebody else's problem' and assigning ownership is often the\nquickest way to reduce costs.\nReporting\nEnergy management needs a formal reporting structure to ensure that targets are met\nand translated into real financial performance improvements.\nTo be effective, reporting must:\n\u2022 Be regular - ideally part of the monthly management accounts for management\npurposes and posted on notice boards for all staff to see.\n\u25cf Be concise and effective - reports should fit onto 1 A4 page at most.\nBe suitable for the audience - simple graphs are the key to attracting and retaining\nthe audience's attention.\n\u25cf Be focused on real improvements in performance and the financial implications.\nExternal targets (sites and machines)\nTargets based on external benchmarking at the site level are possible using industry\ndata but the results are only relevant for a specific process and production rate.\nTargets based on external benchmarking at the machine level are possible using\nindustry data but the results are only relevant for a specific process and production\nrate.\nInvestment decisions\nInvestment in improving energy usage performance can change the rules of energy use\nand make energy cost reduction automatic.\nInvestment in improving energy efficiency is often neglected because of the lack of a\nrecognisable income stream from the investment.\nInvestment in capital equipment should consider the whole life cycle of the equipment\nand particularly the energy costs over the life cycle.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n6\nPower supply\nUnderstanding billing information\nUnderstanding the energy consumption (electricity or gas) is a key task in energy\nmanagement and many companies do not understand how to read their bill or fail to do\nso. Simply knowing how to read energy bills can save money by revealing areas for\npotential cost reduction.\nActions to take are:\n\u2022\nLearn how to read the energy bills. Contact the supplier if in doubt about any\nelements of the bill.\nRead and check the energy bills every month. Bills are not always correct and often\ncontain errors.\n\u25cf Record energy billing data in a spreadsheet every month.\n\u2022\n\u2022\nCross-check the energy billing data with manual reading of the relevant meters.\nMake the person responsible for energy use (generally the Production Manager)\nresponsible for signing off the energy bills each month.\nEnergy costs are often in the region of 8% of production costs and most companies do\nnot spend enough time looking at the bills.\nAvailable capacity and maximum demand\nThe Available Capacity is the amount of power that a site is allowed to draw from the\nsupply network (in kVA). This is a set amount and is the limit of power that can be\ndrawn without penalty charges being applied.\nThe Maximum Demand is the actual monthly maximum power drawn from the network.\nIf the Available Capacity is too high there will be high fixed monthly charges and if the\nAvailable Capacity is too low there will be penalty charges.\nSetting the correct Available Capacity is a strategic decision for management.\nPower factor correction\nThe power factor (or cos p) is the ratio of 'useful power' to 'apparent power'.\nA low power factor increases losses and improving the power factor with Power Factor\nCorrection (PFC) equipment will reduce the maximum demand (in kW) and allow\nreductions in the Available Capacity.\nInterval data\nMost sites will have a recording meter and 12-hour data should be available from the\nsupplier. This is a vital tool and simple plots of energy use versus time will reveal\nabnormal events and allow these to be investigated.\nEnergy mapping\nAn energy map' of a site is easily prepared and will show where most of the energy is\nbeing used at a site. This is often not where it is expected to be! The \u2018energy map' can\nbe used to target efforts to the most rewarding areas and to decide on sub-metering\narrangements.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n7\nMotor management\nMotor management policy and motor register\nMotors are the largest energy user in most plastics processing and motor management\nis a necessity for modern plastics processing.\nSites should create a simple Motor Management policy and decision matrix for the\npurchase and maintenance of all electric motors that covers:\n\u2022\n\u2022\n.\n\u2022\nRepair and replacement based on lifetime costing.\nSpecification of high efficiency motors for all new motors.\nA 'rewind' policy (rewound electric motors are less efficient).\nA 'motor register' to manage the motors on the site.\nMinimise the demand\nMinimizing the demand is the first step to\nmanaging any service and should be\nMinimise the demand\nOptimise the supply\ncompleted before optimizing the supply. It is\nnot economic to optimize the supply based on\nexcessive demand.\n|\n|\nStep 1\nTurn it off\nTurn it off\nTurning motors off is one of the most effective\nmethods of reducing energy usage. This can\nbe done with timers, condition sensors,\nsequenced operation or by linking downstream\nequipment to the main processing machine.\nReduce the load\n|\n|\nStep 2\nReduce the load\nStep 3\nSelect the correct size\nmotor\nStep 4\nImprove motor\nefficiency\nStep 5\nSlow the motor down\nStage 1\nStage 2\nReducing the load at source can be done by cleaning the systems or lowering pressures.\nThe transmission systems (belt drives or gearboxes) should also be examined, e.g.\ntoothed belts are 2-3% more efficient than standard V-belts.\nSelect the correct size motor\nOperating motors at the maximum efficiency means getting the size right. Many motors\nare too large for the actual application and using a large motor for a small load\nincreases energy use.\nOptimise the supply\nImprove motor efficiency\nHigh efficiency motors offer significant energy savings for a small additional cost.\nSlow the motor down\nVariable speed drives (VSDs) allow motors to be slowed down to match the demand\nand offer energy savings and improved process control. VSDs are one of the most\nimportant tools available to plastics processors to reduce energy use and costs in\npumps and fans.\nSimply slowing a motor down by 20% with a VSD reduces the energy use by 49%.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n8\nCompressed air management\nCompressed air is NOT free, it is a very expensive resource. In most sites, compressed\nair is approximately 10% of the total energy use.\nMinimise the demand\nReduce leakage\nCompressed air leakage is an avoidable waste.\nIn the average site, leaks use from 20-40% of\nthe compressed air generated.\nLeaks are very expensive and a 3mm\ndiameter hole @ 7 bar can cost up to \u00a31,500/\nyear in lost air. A `rule of thumb' is that if you\ncan hear a compressed air leak in a quiet site\nthen it is costing more than \u00a3200/year.\nI\n| Minimise the demand\nOptimise the supply\nStep 1\nReduce leakage\nStep 2\nReduce usage\nStep 3\nReduce generation\ncosts\nStep 4\nReduce treatment\ncosts\nStep 5\nImprove distribution\nImplement a regular check of all systems and\nconsider purchasing an ultrasonic detector to\nfind leaks at noisy sites.\nStage 1\nStage 2\nA compressed air map of the site is a vital tool to locate leaks and poor usage. Map\nwhere compressed air goes and how it is used.\nReduce usage\nCompressed air usage should be reduced by using other means of power where\npossible. Almost any other method of doing a job will be cheaper than using\ncompressed air. Areas to examine are:\n\u2022 Robots using compressed air to provide vacuum.\n\u2022 Assembly areas with bowl feeders and air lines for product movement.\n\u2022\nAir-operated power tools which cost 10 times more than direct electric drives to\noperate.\nOptimise the supply\nReduce generation costs\nFeeding cold air to the compressor inlet will reduce the cost of compressing the air (it is\nalready more dense).\nCompressed air costs can be reduced by reducing the system pressure to the minimum\nrequired to actually operate the process.\nConsider purchasing a VSD compressor to dramatically reduce generation costs.\nReduce treatment costs\nCompressed air treatment is expensive and the bulk of air should be treated to the\nminimum quality necessary, e.g. 40-micron filters are usually sufficient. Filters should\nbe tested regularly to make sure that the pressure drop does not exceed 0.4 bar and\nelectronic condensate traps should be used instead of manual condensate traps.\nImprove distribution\nA ring main and good smooth bore piping should be used to reduce distribution costs.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\nCooling water management\nCooling water (chilled and cool) is a major hidden cost for plastics processors and in\nmost sites, cooling water air is approximately 10-15% of the total energy use.\nMinimise the demand\nReduce heat gains\nInsulate chilled water piping wherever possible\nto reduce parasitic heat gains. The\ntemperature that counts is the temperature at\nthe point of application not the temperature at \u00a6\nthe point of generation.\n| Minimise the demand\nOptimise the supply\nStep 1\nReduce heat gains\nStep 3\nReduce cooling costs\nGood insulation reduces parasitic heat gains\nStep 2\nIncrease temperatures\nStep 4\nReduce distribution\ncosts\nand allows increased generation temperatures\nwith no effect on the process. Typically\nStage 1\nStage 2\ninsulation projects will have a payback of less than 1 year.\nIf you can see condensation on pipes during summer then it is obvious that there is a\nneed for insulation on the piping.\nIncrease temperatures\nIncreasing water temperatures will reduce energy costs. For a chiller system (chillers\nare basically compressors), a 1\u00b0C increase in the flow temperature will decrease the\ngeneration costs by \u2248 3%.\nSites should find out what the flow temperatures are and ask why they are set at this\nlevel. Increasing the flow temperature by 4\u00b0C will decrease chiller operating costs by >\n10%.\nOptimise the supply\nReduce cooling costs\nFor cooling water (16\u00b0C to 30\u00b0C), cooling towers are widely used and offer good\nopportunities for energy saving through low-cost actions such as VSDs for fan control.\nThe total cost of cooling when using cooling towers is often increased substantially\nbecause of the need for Legionella controls. Many modern sites use air blast or free\ncooling to remove the need for cooling towers and the associated testing.\nFor chilled water (5\u00b0C to 16\u00b0C), chillers are almost always used to produce the chilled\nwater. If the flow temperatures are high enough (> 12\u00b0C) it is possible to use air blast\nor free cooling to act as pre-cooling for the return water. In this case, when the\nexternal temperature is < 9\u00b0C the pre-cooler will provide the complete cooling load and\nthe chiller will not operate. This arrangement will reduce cooling costs dramatically and\nhave a payback of less than 2 years.\nReduce distribution costs\nMost sites use fixed speed pumps for distribution of the cooling and chilled water. This\nis a perfect application for VSDs. A VSD can be used to control the speed of the pump\nbased on the temperature of the return water from the process - as the water gets\nwarmer the pump slows down to adjust to the needs of the process. VSDs in this type\nof application often generate paybacks of less than 6 months.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n10\n\n\nGeneral processing\nEverything is more efficient\nPlastic processing technology is rapidly improving in energy efficiency and old machines\nare inevitably less energy efficient than new machines. Processors using old machinery\nare not saving money, they are paying more to run their process than their competitors\nand they may well be putting themselves at a permanent cost disadvantage.\nCompared to the machines available in 1996, modern machines are at least 20% more\nenergy efficient and in the case of injection moulding machines where all-electric\nmachines are the new technology the new machines are up to 60% more efficient than\nthe standard 1996 hydraulic machine.\nThere is no conflict!\nMost processing methods offer significant opportunities for energy management and\nenergy efficiency improvements and it is important for processors to understand that\nthere is no conflict between energy efficiency and productivity - both can be achieved.\nIn fact, increasing the production rate of most plastics machinery decreases the kWh/kg\nbecause the base load of the machine is amortised into a greater process load.\nSpecifying new machines\nWhen considering purchasing new machines, sites need to consider the 'whole life cost'\nof the machine rather than the simple 'initial cost'.\nThe cost of operating a machine over a 10-year life will always be greater than the\ninitial purchase cost. In addition, this cost will increase as energy prices increase - a\ncheap initial cost' machine can easily be the most expensive machine over a 10-year\nlife.\nSpecifiers should look particularly for large motors that are not used to their design\nspecification on small machines.\nMachine monitoring\nMachine energy monitoring can be rapid and\nlow cost and allows processors to see inside\nthe machine cycle and to adjust the settings\nto get the most energy efficient settings for\nthe job.\nThe chart at right shows the energy\nconsumption of an injection moulding\nmachine. At the left hand side of the chart the\nmachine is idling with no production and\ndrawing 80% of the full load power.\n002-\n00:34\nIdling machines in any process are not free - they are costing large amounts of money\n(up to 90% of the full running costs) but are often ignored by site management.\nMachine monitoring is also a sensitive indicator of the general condition of a machine.\nAs a general rule, when producing the same product under the same conditions,\nincreasing energy use indicates a need for maintenance.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n11\nInjection moulding\nGet the right machine\nAlways check that the machine is right for the job. Large machines making small\nproducts are energy inefficient and will increase costs. For some smaller machines the\nuse of accumulators can reduce transient power requirements.\nProcess setting and controls\nProcess setting is the key to energy efficiency in injection moulding. Optimised process\nsettings will increase productivity and reduce energy use.\nBarrel heating and insulation\nBarrel heating in injection moulding machines is a major energy user and the energy\nuse can easily be reduced by fitting barrel insulation. Barrel insulation will reduce\nenergy use in heating with payback periods of around 1 year. Barrel insulation also\nreduces Health and Safety concerns with hot surfaces. At sites which also use air\nconditioning, barrel insulation will also reduce the air conditioning energy use.\nEven if barrel insulation is not used it is essential that barrel heaters are 'bedded-in' and\nuse a conductive metal compound between the heater and the barrel for good heat\ntransfer. Sites should also ensure that barrel heater thermostats are accurate and can\ncontrol the heater.\nAll-electric machines\nInjection moulding machines have made a huge leap in efficiency with the introduction\nof all-electric machines and these can give processors a permanent advantage over\ncompetitors using conventional hydraulic machines. All-electric machines not only use\nup to 60% less energy in operation but also have lower standing losses, are easier to\nmaintain and are more accurate in operation.\nRetro-fitting VSDs\nRetro-fitting VSDs to hydraulic machines is very cost effective if the machine\nparameters are right. VSDs will save energy by slowing down or stopping the hydraulic\nmotor when the cycle does not need it. Investigate if the machines are suitable (large\nfixed displacement motors, long cycle times and long operating hours).\nMould temperature controllers\nMould temperature controllers (MTCs) are a hidden cost in injection moulding and the\nneed to use them should be examined carefully. Insulate piping between MTCs and\ntooling to reduce parasitic heat gains or losses.\nMould design\nInitial mould design can affect energy use and designers need to be aware of the cost\nof their decisions. Areas to look at are:\nInvestigate 'conformal cooling' to reduce cooling times and energy use.\n\u2022\nMinimise sprue and runner sizes to minimise the material processed in the cycle.\nSprues and runners are not free even if they are regranulated.\n\u2022\nHandling systems should operate 'on-demand' and use gravity if possible. It is free.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n12\nExtrusion\nGet the right machine\nAs with injection moulding, using large extruders for small profiles wastes energy and\ncosts money. It is often possible to switch extruder motors to match the size of the job.\nThis is sometimes a cost-effective operation if the cost of changing the motor is less\nthan the extra energy used.\nAC motors and VSDs\nExtrusion costs can be reduced by replacing DC motors with high efficiency AC motors\nand VSDs.\nThe energy savings will be in the region of 4% but the main advantages are the\nincreased reliability, decreased maintenance load and ease of motor replacement.\nMotor sizing\nCheck the loading on extruder motors and modify gear ratios to optimise the energy\nusage. Extruder motor gear ratios can be managed to optimise the motor load and\nmaximise energy efficiency.\nWhere belt drives are used then replacing the standard V-belt drives with toothed belts\ncan give 2-3% energy savings.\nInsulation\nBarrel insulation in extrusion is not generally needed and can lead to a 'runaway\nprocess' because shear heating should supply most of the heating load. Shear heating\nis also much more energy efficient than electrically applied heating. In most cases,\nextruders will need barrel blowers to remove excess shear heat generated. It is\nimportant to check the heating and cooling controls to make sure that heating and\nblowing are not fighting one another.\nInsulation is cost effective in areas where shear heating is low such as the first zone\nwhere the incoming material absorbs a lot of heat and for most areas forward of the\nscrew tips where there is little shear heating. Areas forward of the screw tips suitable\nfor insulation include:\n\u2022 Transfer pipes from secondary to primary extruders.\nHot oil pipes.\n\u2022 Melt pumps and filters.\n\u2022\nDies.\nDownstream equipment\nFor profile extrusion using vacuum calibration, the vacuum tanks should be regularly\nchecked for seal efficiency and the vacuum pumps should be fitted with VSDs to control\nthe amount of vacuum generated.\nFor sheet extrusion the method of treating any edge trim should be carefully examined.\nFor blown film extrusion the fans for the chilling bubble should be VSD controlled to\ngive energy savings and improved control.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n13\nInjection and extrusion blow moulding\nInjection and extrusion blow moulding are similar to standard injection and extrusion\nand many of the actions listed for injection moulding and extrusion should be carried\nout.\nAll-electric machines\nAll-electric machines are now available for both injection and extrusion blow moulding.\nThese have all the benefits of all-electric injection moulding machines and can reduce\nenergy use significantly.\nBarrel insulation\nShear heating in blow moulding does not contribute greatly to the heat input to the\nmelt and barrel insulation can be very profitable for both injection and extrusion blow\nmoulding.\nMelt temperature\nCooling is a major part of the cycle time for both types of blow moulding and a major\nenergy use (at the chillers). Minimising the melt temperature to the minimum needed\nwill reduce the cooling demand and will improve both cycle times and energy use.\nParison control\nGood parison control in extrusion blow moulding will improve product quality, process\nefficiency and reduce energy use. Investment in improved parison control will have a\ngood payback.\nCompressed air\nThe blowing step is a large user of compressed air in both injection and extrusion blow\nmoulding and good compressed air management is vital, particularly for injection blow\nmoulding of PET pre-forms where the compressed air pressure is generally much higher\n(40 bar).\nActions to take are:\n\u2022 Minimise the blowing pressure to just enough to get full blowing.\n\u25cf\nReduce compressed air pressures when holding after initial blowing.\n\u2022 Use compressed air recovery systems to recover the high-pressure air (40 bar) for\nuse in the low-pressure (7 bar) system.\nTops and tails\nTops and tails management is a key energy issue in extrusion blow moulding. The\nmaterial from the tops and tails may be recycled but the energy and production\ncapacity is lost forever. Setters should be given targets for tops and tails (< 20% by\nweight) and machines should not be released for production until this target is\nachieved.\nRecrystallization\nRecrystallization of regrind from PET pre-forms or mouldings can be combined with\ndrying through the use of infra-red drying to give good energy savings.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n14\nThermoforming\nExtrusion is a first stage of thermoforming and processors should carry out the actions\nlisted for extrusion. This section only considers the thermoforming process from sheet\n(either from roll or in-line).\nPre-warming ovens\nWhere pre-warming ovens are used these should be well sealed and insulated to\nprevent excessive heat losses. Examination with a thermal camera will quickly reveal\ndegraded seals and areas of heat leakage.\nMinimising the distance between the pre-warming oven and the actual thermoformer\nwill minimise heat losses.\nInsulation and seals\nPrimary heating ovens should be well sealed and insulated.\nEntrances and exits\nMinimising the size of entrances and exits on pre-warming and primary ovens will\nminimise heat losses. If the sheet being warmed is only 0.5 mm thick then the entrance\nand exit gaps do not have to be more than 4 mm.\nHeater banks\nThermoformer heater banks should use heating elements that match the emissivity of\nthe heater to the absorption of the material.\nHeaters should be kept clean to ensure good emissivity at the right wavelength for the\nmaterial.\nCatalytic flameless gas heaters are a new technology that can remove the need to\nelectric heating of the sheet.\nHeater banks should be sealed and insulated where possible to reduce heat losses from\nradiation, convection and conduction.\nCooling\nCooling is often a limiting factor in cycle times and the sheet temperature should be the\nminimum required for good forming.\nWeb handling\nThermoforming generates a large amount of web waste that is reground and recycled.\nThe regrinders and blowers should be linked to the thermoformer operation so that they\nstop operating when the thermoformer is not producing.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n15\nRotational moulding\nGas use data\nRotational moulding is the one plastics processing method where gas use is higher than\nelectricity use. It is possible to create a Performance Characteristic Line (PCL) for both\nelectricity and gas use for rotational moulding and these can be combined to give a\ntotal PCL for the site.\nGas meters do not normally provide interval data but it is possible to fit an interval data\nmeter to the main gas feed. This is strongly recommended for rotational moulding sites.\nThe energy efficiency of individual ovens can be examined by fitting standard or interval\ngas meters to ovens.\nBurners\nGas burner efficiency is a key driver of efficiency in rotational moulding and all burners\nshould be monitored to give complete combustion of the gas.\nCombustion efficiency can be improved by a variety of methods and these should be\ninvestigated.\nTooling\nThe rotational moulding process does not need a hot mould, it needs a hot plastic that\nwill flow and the quicker the heat can be got into (and out of) the mould the faster the\nprocess and the better the process efficiency. Increasing the heat flow can be achieved\nby:\n\u2022\nReducing the thermal mass of the mould by using high heat transfer rate materials.\nUsing heat pipes and other techniques to get heat into the mould quicker.\nInsulation\nRotational moulding ovens are often poorly sealed and insulated and seals and\ninsulation will always degrade with time. Examination of ovens with a thermal camera\nwill quickly reveal degraded seals and areas of heat leakage through poor insulation.\nProcess settings\nProcess settings for rotational moulding are often poorly defined and offer good\nopportunities for improvement.\nActions to take are:\n\u2022 Reduce door opening times to minimise heat losses from the oven.\nMinimise the size of the door opening to minimise heat losses.\nVSDs for pumps and fans\nRotational moulding uses many fans for recirculation and exhaust of the hot combustion\ngases. These are ideal applications for VSDs to improve process control and minimise\nenergy use.\nThe cooling phase of rotational moulding uses fans and sometimes water pumps for\nspraying water onto the mould. These are ideal applications for VSDs to reduce energy\nuse.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n16\nOperations\nOperations is where the technical improvements are put into practice - it is where the\n'rubber meets the road'. Operations depends on people, improvements are cheap but\ncan often be difficult to implement and sustain. Training and motivation are the key\nissues in operations.\nControlling machine operations (start-up, stand-by and shut-down) is a key factor in\nreducing energy costs.\nProcess setting\nOptimized settings for production consistency and cycle time reduction will also give\noptimised energy use.\nThe initial process setting needs a scientific approach to find the best setting and give\nthe best results. These setting must be adequately recorded and used to be effective.\nNo settings should be changed without justification and high-level approval.\nStart-up\nStart-up should follow setting sheets at all times. Correct sequencing of operations and\na simple time-line approach to start-up will reduce energy use dramatically, e.g. in\ninjection moulding there is no logic in turning on the main motor (the major energy\nuse) until the latest possible time.\nFast tool changes will reduce energy costs in idling machines.\nStand-by\nMachines should be set into 'stand-by' when they are not going to be used for a short\ntime, e.g. main motor off, downstream, equipment off. If machines are not going to be\noperated for more than 4 hours then they should be shut-down. Automation can help\nhere.\nShut-down\nShut-down should take the high energy loads off-line as soon as possible, e.g. in\ninjection moulding then the main motor should be turned off as soon as possible.\nAutomation can help here.\nTraining\nStaff training is one of the quickest and most profitable actions in energy efficiency and\ncan reduce energy use by up to 20%. Training should explain the process, motivate\nstaff and be relevant to the role of the staff. Empowering staff to switch machines/\nprocesses/services off can lead to remarkable results.\nMaintenance\nMaintenance is a key issue in achieving and sustaining the energy efficiency of\nprocesses and machines. The maintenance function has a key role to play in both\nachieving and maintaining energy management.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n17\nBuildings\nBuilding energy costs are not always a significant percentage of the total energy costs\nin plastics processing and at the typical site they are 7-8% of the total energy costs.\nDespite this, they are almost always the first area to be considered and improving\nbuilding energy efficiency can reduce costs, improve staff comfort and improve work\noutput.\nLighting\nLighting only represents around 5% of the energy use at a typical plastics processing\nbut some lighting projects are strongly recommended as a visible sign of management\ncommitment to energy management.\nLighting can be divided into 'ambient' and 'task' lighting - they are very different.\nAmbient lighting is to allow safe movement, task lighting is to allow completion of a\nspecific task. The lighting levels are very different, recognising this and taking action to\nseparate them can reduce costs.\nA 'lighting map' is vital in reducing lighting energy use. Map the lights, switches and\ncontrols on the site to identify areas for improvements.\nInvestment in controls such as PIRs, timers and push switches can automatically reduce\nlighting costs without affecting product or lighting quality.\nHeating\nHeating energy use in buildings is 'condition' driven, i.e. the driver is the external\ntemperature. Monitoring and targeting for heating use should be carried out using\nHeating Degree Days (HDD = a measure of how cold it is). It is possible to create a PCL\nfor heating use and HDD to set targets and assess performance.\nQuality and comfort heating are very different. Taking action to separate them can\nreduce costs.\nReducing the heating load is the first task and heating levels should be set to match the\nactivity.\nInvestment in heating controls can reduce heating costs but they must be set correctly\nand be tamper-proof.\nAir conditioning\nAir conditioning is a rapidly rising energy user but is mostly 'comfort cooling' for a few\ndays of the year, this can cost as much as the yearly heating bill. Air conditioning\ncontrols are often tampered with by staff.\nBuilding fabric\nBuilding fabric improvements can reduce both heating and air conditioning loads and\nreduce costs.\nThe main tasks are to reduce air leakage and to improve building insulation. Air leakage\nis not the same thing as ventilation. Insulation can be improved through simple local\nmeasures.\nOffice equipment\nThe energy use of office equipment is generally low but can be reduced through simple\nstaff measures, energy saving setting and simple 24/7 timers.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n18\nSite surveys\nSite surveys are a key part of energy management. They identify the status of a site\nand are a reference point for future progress.\nFollow the data\nInformation is the key to an effective site survey.\nBasic energy consumption data, and an energy map are needed for a site survey. These\nallow targeting of the largest energy usage areas to provide the greatest rewards.\nEquipment\nThe equipment needed for a basic site survey is minimal. It is possible to carry out a\nsite survey with virtually no equipment. The equipment needed for an advanced site\nsurvey is inexpensive but allows more value to be added.\nPlanning\nSite surveys should be planned and carried out during normal production and also, if\npossible, during shut-down periods.\nProject generation\nSite surveys should produce a range of clearly defined projects that will pay back\nrapidly and should report in financial terms to gain top management support.\nSite surveys should be regularly repeated to check progress, to report success, to close\nout completed projects and to generate new projects.\nNon-conformance reports\nSite surveys should drive action to reduce energy usage and effective non-conformance\nreports are a key to action. Most sites have a quality management system, use this as a\nmodel for non-conformance reporting.\nENERGY MANAGEMENT IN PLASTICS PROCESSING\n19\nBPF\nThe British Plastics Federation\n5-6 Bath Place\nRivington Street\nLondon\nEC2A 3JE\nTel: 020 7457 5000\nFax: 020 7457 5020\nE-mail: reception@bpf.co.uk\nwww.bpf.co.uk\nStronger\nTogether\n"}, "expected_output": {"claims": [{"unit": "kwh/kg", "value": 0.4, "evidence": ["Extrusion: 0.4 to 0.6 kWh/kg.", "The process load of a site is the slope of the best fit line and is the energy needed to\nrun the process. Reducing the process load is more difficult to achieve because it\ngenerally (but not always) requires more fundamental process improvements. The\nprocess load depends on the type of process being used at the site.\n", "The process load depends on the type of process being used at the site.", "\u2022 Extrusion: 0.4 to 0.6 kWh/kg."]}, {"unit": "kwh/kg", "value": 0.6, "evidence": ["Extrusion: 0.4 to 0.6 kWh/kg.", "The process load of a site is the slope of the best fit line and is the energy needed to\nrun the process. Reducing the process load is more difficult to achieve because it\ngenerally (but not always) requires more fundamental process improvements. The\nprocess load depends on the type of process being used at the site.\n", "The process load depends on the type of process being used at the site.", "\u2022 Extrusion: 0.4 to 0.6 kWh/kg."]}]}, "metadata": {"product_category": "Paper and plastic products", "request_id": "req_9cfe25cede501bc8"}}
{"id": "8cc61f002986578d3bd8b014", "input": {"query": "What is the specific energy consumption (SEC) in MJ/kg or kWh/kg for injection molding of thermoplastic parts, excluding upstream polymer production?", "source_url": "http://web.mit.edu/ebm/www/Publications/Thiriez_ISEE_2006.pdf", "document_text": "An Environmental Analysis of Injection Molding\nAlexandre Thiriez and Timothy Gutowski\nDepartment of Mechanical Engineering\nMassachusetts Institute of Technology\nCambridge, MA, USA.\nAbstract This environmental analysis of injection molding\nhighlights a few important points. The choice of injection\nmolding machine type (hydraulic, hybrid or all-electric) has a\nsubstantial impact on the specific energy consumption (SEC).\nThe SEC values for hydraulic, hybrid and all-electric machines\nanalyzed are 19.0, 13.2 and 12.6 MJ/kg respectively (including\nauxiliaries, compounding and the inefficiency of the electric grid).\nFor hydraulic and hybrid machines SEC seems to exhibit a\ndecreasing behavior with increasing throughput. This derives\nfrom spreading fixed energy costs over more kilograms of\npolymer as throughput increases. For all-electric machines SEC\nis constant with throughput. When the polymer production stage\nis included in the analysis, the energy consumption values\nincrease up to 100 MJ/kg. The overall injection molding energy\nconsumption in the U.S. in a yearly basis amounts to 2.06 x 108\nGJ. This value is of similar magnitude to the overall U.S. energy\nconsumption for sand casting, and to the entire electricity\nproduction of some developed countries\nKeywords-Injection Molding; Hydraulic; Hybrid; All-electric;\nSpecific Energy Consumption (SEC); Life Cycle Inventory (LCI).\nI. INTRODUCTION\nPlastic components are integral parts in electrical and\nelectronic (E&E) products. 8.5% of the plastic production is\ndedicated to this market [1]. Although this number might seem\nsmall it is larger than the amount of plastic used for the\nautomotive industry (8%) [1]. In E&E products plastic can\nrepresent from 3% of the total weight in medical equipment to\n33% in small house appliances and 42% in toys [1]. The\nmajority of this plastic used for E&E products is injection\nmolded in order to attain the specific geometric requirements.\nInjection molding involves melting polymer resin together with\nadditives and then injecting the melt into a mold. Once the\nresin is solidified, the mold opens and the part is ejected. At\nfirst glance, injection molding may appear to be a relatively\nbenign process with respect to the environment due to its low\ndirect emission levels and apparently low energy consumption.\nHowever, when calculating the environmental cost of injection\nmolding one must also take into account the ancillary processes\nand raw materials used in the process. Aside from the raw\nmaterial production stage which has substantial emissions, the\nmain metric in the whole injection molding process is energy\nconsumption. The large scale of the injection molding industry\nmakes the environmental impacts of this process especially\ncritical. In other words, a small increase in the efficiency of the\nprocess could lead to substantial savings for the environment.\nThis paper investigates injection molding from an\nenvironmental standpoint, yielding a system-level\nenvironmental analysis of the process. It provides a transparent\nprocess model that includes all major steps involved in the\nproduction of injection molded products and shows the\ndependency of injection molding on the most important process\nparameters. This paper presents a summary of our findings\nalong with detail on four major issues:\n1. Relationship between\nthroughput.\nenergy consumption and\n2. Differences in environmental performance between\nhydraulic and all-electric machines.\n3. Role of secondary/subsidiary process in the energy\naccounting.\n4. Environmental scale of injection molding.\nII.\nBACKGROUND\nWith regards to injection molding life cycle inventories\n(LCI), much effort has gone into studying the production of\nraw materials (polymers) as well as the product end-of-life\naspects, such as disassembly separation and recycling.\nAmongst the researchers in this area, it is worth mentioning Ian\nBoustead, who developed a set of \u201ceco-profiles,\" or LCI's, of\nthe most consumed polymers in the plastic industry. He also\ncreated life cycle inventories for injection molded PVC and\ninjection molded polypropylene. The former LCI studied two\ninjection molding facilities in France that produce PVC fittings\nfor pipe drainage systems [2]. The latter LCI studied one\nfacility in the U.K. that produces 12 to 76 g polypropylene\ncomponents [3]. These studies are product specific, narrowing\non one application and one set of processing parameters. In an\neffort to obtain a range of values typical in injection molding,\nand thus more breadth of data, this study incorporates\nmeasurements from machines processing different products\nand materials. It also provides a transparent outline of all the\nsub-processes that make up the injection molding process\ntogether with their environmental performance.\nOther contributors to the field of injection molding include\nMattis et al. 1996 and Boothroyd et al. 2002. Mattis et al. used\na 3-D solid modeling environment and numerical analysis to\nexplore the influence of mold design, part design, and some\nprocess parameters on the process efficiency [4]. Boothroyd et\nal., whose goals were to develop design-for-manufacturing\nguidelines, developed a set of empirical equations predicting\nmachine size and processing time of each stage in the injection\nmolding cycle [5].\nIf the reader is not familiar with injection molding\ntechnology please refer to [6].\nIII.\nMACHINE ENERGY CONSUMPTION\nThe main division in injection molding machinery lies in\nhow the drives in these machines are powered. The oldest and\nmost common injection molding machine type is the hydraulic\npowered machine. This machine uses one or more hydraulic\npumps to power all of the machine's motions. One can have a\npump for each drive, a centralized pump driving all motion, or\na combination thereof. There are two obvious inefficiencies\nwith hydraulic machines. First, for most machines, pumps\ncontinue running even while the machine is idle, consuming\npower that does not get used in production and thus wasting\npower. Secondly, there is an intrinsic inefficiency in the\narchitecture of the system. An electric pump transfers work to\nthe hydraulic circuit, which in turn transfers work to the\nmechanical components. Each transfer of work entails\ninefficiencies. Why not eliminate one of these transfers? This\nis where all-electric powered injection molding machines come\ninto place. As their name indicates, these machines use servo\nmotors to power each of the mechanical drives. Basically, one\nservo motor runs the rotation of the screw, another moves the\nscrew along the injection axis, and a third moves a toggle\nclamp to close the mold. Aside from the above mentioned\nmain servos, there might be others that run secondary\nfunctions. These machines exhibit superior efficiency on\naverage, but are not applicable for high clamping force\napplications due to the instabilities in the toggle clamp\nconfiguration. This is where the hybrid powered machines\ncome into place. A hybrid machine uses both servo motors and\nhydraulic pumps. The most common configuration is using the\nhydraulic pump for clamping and servo motors for screw\nmovement. These machines sacrifice some of the all-electric\nefficiency for the precision of hydraulic clamps. Fig. 1 portrays\nthe power requirement for a hybrid and an all-electric machine\nboth running the same part with a cycle time of 14 seconds.\nSimple inspection reveals substantial energy savings from\nusing all-electric over hybrid technology. The reader must note\nthat the curve for a hydraulic machine would be even higher\nthan that of the hybrid.\nThus the choice of machine has a substantial impact on the\nspecific energy consumption' (SEC), or energy consumption\nper kilogram of polymer processed. More than 100 energy\nmeasurements and calculations were examined from the three\ntypes of machines. This analysis yields average SEC values for\nhydraulic, hybrid and all-electric machines of 3.39, 1.67 and\n1.46 MJ/kg respectively (without accounting for the efficiency\nof the electric grid).\nFor hydraulic and hybrid machines SEC seems to exhibit a\ndecreasing behavior with increasing throughput, as portrayed in\nFig. 2. This derives from spreading fixed energy costs over\nmore kilograms of polymer as throughput increases.\npower in a hydraulic and hybrid can be described as:\n-\nP = Po + km\nwhere,\nPo = fn(hydraulic pumps, computer, etc..)\nk = extra SEC to process the polymer\nThe\n(1)\nwhere Po is the fixed power requirement (power required when\nthe machine is on, but not processing any polymer), \u1e41 is the\nthroughput or process rate, and k is a processing constant. In\nterms of SEC, this formula can be expressed as:\nP E\nm m\nPo\nSEC = +k\nm\n(2)\nAs throughput increases, SEC approaches the constant k as\nobserved in Fig. 2.\nAll-electrics on the other hand have very low fixed energy\ncosts (ex: running the computers), and their SEC stays constant\nas throughput increases, as portrayed in Fig. 3.\nName = Magna MM550 with and without e-drive.\nShot Weight 0.68 kg for the PS shots. Unknown kor others.\nPower Required (kW)\n120\n100\nPlasticize\n80\n60\n40\n20\n20\nTon\nBuildup\nCool\nClamp open-\nclose\nInject high\n2.6\n550 hybrid press (PP)\n2.4\nSEC (MJ/kg)\n2.2\n550 hybrid press (HDPE)\n550 hybrid press (PS)\n550 hydraulic Press (PS)\n2.0\n1.8\n1.6\nInject low\n1.4\n40\n90\n140\nThroughput (kg/hr)\n190\n0\n0\n1\n2\n3\n4\n5\n6 7 8 9\nTime (seconds)\n10 11 12 13 14\nMM 550 Hybrid\nNT 440 All-Electric\nFigure 2. SEC vs. throughput for a Magna MM550, hydraulic and\nhybrid. There is no inclusion of the efficiency of the electric grid.\nSource: [8, 9].\nFigure 1. - Energy consumed in the injection molding cycle of a hybrid\n(electric screw drive) and an all-electric machine. Source: [7].\nHere we use a common expression, while in fact energy is not consumed but\ntransformed. A more rigorous but less understood definition for SEC would\nbe \"specific exergy consumed\".\nSEC (MJ/kg)\n9\n\u2022 8765+\n4\n3\nAll-Bectric 85 tons\nHydraulic - 85 tons\nMaterial: Polypropylene (PP)\n2\n1\n0\n0\n5\n10\n15\nThroughput (kg/hr)\n20\nFigure 3. SEC vs. throughput for an all-electric and a hydraulic machine.\nThere is no inclusion of the efficiency of the electric grid. Source: [8].\nIV.\nSUMMMARIZED LIFE CYCLE INVENTORY\nIn order to develop a successful life cycle inventory (LCI) it\nis first necessary to establish the boundaries of the system to be\nanalyzed. In the case of injection molding, the overall process\nstarts at the polymer production stage. This stage takes raw\nmaterials from the earth and transforms them, with the addition\nof energy, into polymers. The raw polymer is then shipped in\nbulk to the compounder which mixes it with additives in order\nto bestow the polymer with the required properties for its future\napplication. The polymer is then shipped to the injection\nmolder which transforms the polymer into a finished product.\nThe injection molder might also add some additives in the\nprocess, such as coloring. After being injection molded and\npackaged, the product is ready to be used by the consumer (and\neventually disposed). The scope of this analysis is \"cradle to\nfactory gate\" with the exclusion of packaging. Thus it\nencompasses everything from the creation of the raw materials\nfor polymer production to the injection molding of the product.\nThe system boundaries are portrayed by the dashed square in\nFig. 4.\nClose to 100 sources were consulted in order to develop\nthis LCI. Most of these sources are not listed in the references\nsection but can be obtained from [10]. The results of the LCI\nare summarized diagrammatically in Fig. 5. The reader must\nnote that with the exception of the polymer production stage\nwhen energy data exhibited variation with type of polymer it\nwas averaged according to the relative amount of polymer\ninjection molded in the U.S.\nIt is interesting to note how even though the energy\nconsumption for injection molding machinery seems low, when\nother stages in the process are included the figure becomes\nsubstantial. Considering the energy consumption of all stages\nfrom the compounder to the injection molder (not including\npolymer production), hydraulic, hybrid and all-electric\nmachines yield average values for SEC of 19.0, 13.2 and 12.6\nMJ/kg respectively. These values take into account the energy\nCRADLE\nNaphtha, Oil.\nNatural Gas\nAncilliary Raw\nMaterials\nAdditives\nCompounder\nInternal Transport\nDrying\nThermoplastic Production\nPolymer\nDelivery\nExtrusion\nPelletizing\n(Boustead)\n+\nEmissions\nto air,\nwater &\nland\nEmissions to\nBuilding (lights, heating, ect..)\nair, water, &\nPolymer Delivery\nland\nInjection Molder\nEnergy Production Industry\n\u2611 Internal Transport\nDrying\nEmissions\nInjection Molding\nEmissions to air, water, & land\nto air,\nwater &\nland\nScrap\nAnciliary Raw\nMaterials\nBuilding (lights, heating, ect..)\n---\nPackaging\nNote to Reader:\nFocus of this Analysis\nAlso included in the Paper\n1 kg of Injection Molded Polymer\nService Period\nWaste Management\nFigure 4. Injection Molding System Boundaries\nFACTORY GATE\nENERGY CONSUMPTION BY STAGE in MJ/kg of shot\nThermoplastic Production\nGeneric by Amount\nExtras\nHDPE LLDPE LDPE PP\nPVC\nPS\nConsumed Inj. Molded PC\nPET\navg\nlow\n89.8 79.7 73.1 83.0 59.2\n77.9 79.7 64.6 64.0\n87.2\n81.2\n74.6\n95.7\n78.8\n52.4\n70.8\n69.7\n62.8\n78.2\n59.4\nhigh\n111.5 79.7 92.0 111.5 7\n79.5\n118.0\n102.7\n97.6\n117.4 96.0\navg\n0.19\nPolymer Delivery\nlow\n0.12\nhigh\n0.24\nCompounder\nInternal\nBuilding (lights,\nTransport\nDrying\nExtrusion\nPelletizing\nheating, ect..)\navg\n0.09\n0.70\n3.57\n0.16\n0.99\nlow\n0.30\n1.82\n0.06\nhigh\n1.62\n5.00\n0.31\nSubtotal\navg\n5.51\nlow\n3.25\nhigh\n8.01\navg\n0.19\nPolymer Delivery\nlow\n0.12\nhigh\n0.24\nInjection Molder\nInternal\nTransport\nInjection Molding\nDrying\nScrap (Granulating)\n(look below)\navg\n0.04\n0.70\n0.05\nBuilding (lights,\nheating, ect..)\n0.99\nlow\n0.30\n0.03\nhigh\n1.62\n0.12\nInjection Molding - Choose One\nHydraulic\nHybrid\nAll-Electric\navg\n11.29\n5.56\n4.89\nlow\n3.99\n3.11\n1.80\nhigh\n69.79\n8.45\n15.29\nSubtotal\navg\n13.08\n7.35\n6.68\nlow\n5.35\n4.47\n3.17\nhigh\n72.57\n11.22\n18.06\nTOTAL w/\nHydraulic\nHybrid\nAll-Electric\nGeneric Inj. avg\n93.60\n87.87\n87.20\nMolded\nlow\n71.65\n70.77\n69.46\nPolymer\nhigh\n178.68\n117.34\n124.18\navg\n18.97\n13.24\n12.57\nTOTAL w/o\nlow\n8.84\n7.96\n6.66\nPolymer Prod\nhigh\n81.04\n19.70\n26.54\nNotes Drying - the values presented assume no knowledge of the materials' hygroscopia. In order words, they are\naverages between hygroscopic and non-hygroscopic values. For hygroscopic materials such as PC and PET\nadditional drying energy is needed (0.65 MJ/kg in the case of PC and 0.52 MJ/kg in the case of PET)\nPelletizing - in the case of pelletizing an extra 0.3 MJ/kg is needed for PP\nGranulating a scarp rate of 10 % is assumed\n-\nFigure 5 - Overall System Diagram. The values above account for the efficiency of the electric grid. Multiple sources. Refer to [10] for an extended bibliography.\nburden associated with producing the electricity to power the\nmanufacturing processes\u00b2. When the polymer production stage\nis included in the scope of the LCI, the energy consumption\nvalues increase up to 100 MJ/kg. In the whole LCI, producing\nthe polymer has the greatest impact on the environment. After\nthe polymer production, injection molding machinery and\nextrusion have the greatest impact.\nWith regards to emissions, the majority of emissions come\nfrom the polymer production stage. Please refer to [11] if\ninterested in these emissions. In the rest of the LCI, emissions\ncan divided into: energy related emissions and processing\nemissions. Energy related emissions refers to those emissions\noriginated from the generation of electricity necessary to run\nthe processes. Table 1 presents energy related emissions for\nthe compounder and the injection molder.\nProcessing emissions arise at the polymer processing sites.\nThese kinds of emissions are small compared to energy related\nones. For instance extruding polypropylene generates 0.185 g\nof volatile organic compounds (VOC's), 0.030 g of particulate\nmatter, 0.0099 g of ketones, 0.0022 g of aldehydes,, and\n0.0018 g of organic acids per kg of polymer extruded [12].\nV. ENVIRONMENTAL SIGNIFICANCE\nWhen compared to other conventional manufacturing\nprocesses, injection molding appears to be on the same order of\nmagnitude in terms of energy consumption. For instance,\nprocesses such as sand and die casting have similar energy\nrequirements (11-15 MJ/kg) [13, 14]. However, when\ncompared to processes used in the semi-conductor industry,\nsuch as chemical vapor deposition and atomic layer deposition,\nthe impact of injection molding seems insignificant. This is far\nfrom the truth, though. In order to understand the real impact\nof a manufacturing system one has to understand how\nwidespread its use is in the economy. Injection molding is one\nof the predominant manufacturing processes, and its use is\nincreasing daily in growing economies like China and India.\nTable 2 presents an estimate of the current quantities of\npolymer injection molded in the U.S. and in the world. With\nthese values and distribution of the different machine types, the\ntotal energy spent in injection molding can be estimated.\nAccording to Snyder, in 2002 29% of the machines sold in\nthe U.S. were electric based rather than hydraulic [15]. With\n6 Main Thermoplastics\nAll Plastics\nInj. Molded - Million kg/yr\nU.S. Only\nGlobal\n5,571\n12,031\n23,899\n38,961\nTable 2. Injection molded polymer totals in kg/year. The subdivision\n6 main thermoplastics refers to HDPE, LDPE, LLDPE, PP, PS and\nPVC. The complete calculation can be found at [1]. The sources used\nare [16, 17, 18].\nCompounder and\nInjection Molder\n6 Main Thermoplastics\nAll Plastics\nU.S.\nGJ/year\nGlobal\nGJ/year\n9.34E+07\n2.06E+08\n4.01E+08\n6.68E+08\nTable 3. Total energy used in Injection Molding. The subdivision 6\nmain thermoplastics refers to HDPE, LDPE, LLDPE, PP, PS and PVC.\nThe complete calculation can be found at [1].\nthis information, we assume that 70% of the injection molding\nmachines are hydraulic, 15% are hybrids and 15% are all-\nelectric. Table 3 shows the results of the U.S. and global\nenergy estimate.\nAs can be observed, the overall injection molding energy\nconsumption in the U.S. in a yearly basis amounts to 2.06 x 108\nGJ. This value includes all steps in the LCI, except polymer\nproduction. Including polymer production would increase this\nnumber by an order of magnitude. This value (2.06 x 108 GJ)\nis of similar magnitude to the overall U.S. energy consumption\nfor sand casting (1.62 x 108- 2.28 x 108 GJ, [14]). For the\nreader to comprehend the scale of the U.S. injection molding\nenergy consumption, Table 4 provides values of the entire\nelectricity production of several countries. Without accounting\nfor the electric grid, the overall injection molding, energy\nconsumption in the U.S. amounts to 6.19 x 10' GJ/year\u00b3. This\nvalue can be compared with the values in Table 4.\nIt seems imperative for industry to keep improving the\nefficiency of the process, since small savings anywhere in the\nLCI can lead to tremendous energy savings on a national scale.\nThis seems an intelligent move in a time of raising energy\nprices.\nACKNOWLEDGMENT\nThis research was supported by the National Science\nFoundation Award DMI 0323426.\nEnergy Related Emissions\nStage\nCompounder\nSEC\n(MJ/kg)\n5.51\nCO2\ng\ng\ng\n284.25 1.26 0.51\nSO2 NOX CH4 Hg\ng\nmg\n10.32 0.01\nInjection Modler\nHydraulic\nHybird\n13.08\n7.35\n674.82 2.98 1.22 24.49 0.01\n379.33 1.68 0.68 13.77 0.01\n6.68 344.57 1.52 0.62 12.50 0.01\nAll-Eletric\nTable 1 - Energy-related air emissions for the \"compounder\" stage and the\n\"injection molder\" stage. Multiple sources. Refer to [10] for calculation\n2 The electric grid in the U.S. is 30% efficient. Visit [10] for more details.\n3 Equivalent to electricity consumption.\nAnnual Electricity Production\nWithin 1 Order of Magnitude\nSmaller than U.S. Injection Molding Totals\nHonduras\nIceland\nJamaica\nCountry\nGJ/year Country\nAfghanistan 1.71E+06 Jordan\nGuatemala 2.22E+07 Nicaragua\n1.37E+07 Nigeria\n2.84E+07 Panama\n2.26E+07 Slovenia\nGJ/year Country\n2.55E+07 Austria\nto U.S. Injection Molding Totals\nGJ/year\nCountry\nGJ/year\n2.19E+08 Iran\n4.47E+08\n8.41E+06 Belgium\n2.68E+08\nNetherlands\n3.18E+08\n6.25E+07\n1.78E+07\nBulgaria\nCzech Rep.\n1.49E+08 New Zealand\n1.39E+08\n2.52E+08\nPoland\n4.87E+08\n4.92E+07\nDenmark\n1.27E+08\nPortugal\n1.59E+08\nFinland\n2.56E+08 Saudi Arabia\n4.65E+08\nGreece\n1.80E+08\nSwitzerland\n2.47E+08\n1.36E+08\nIndonesia\nHungary\n3.46E+08\n1.24E+08\nUAE\nVenezuela\n3.15E+08\nTable 4 - Selected countries with smaller or similar order of magnitude electricity production to the amount of energy spent injection molding (compounder\n+ injection molder) in the U.S. Source: [19].\nREFERENCES\n[1] M. M. Fisher, F. E. Mark, T. Kingsbury, J. Vehlow, and T. Yamawaki,\n\"Energy recovery in the sustainable recycling of plastics from end-of-\nlife electrical and electronic products,\" 2005 IEEE International\nSymposium on Electronics and the Environment, May 2005.\n[2] I. Boustead, Eco-profiles of the European plastics industry: PVC\nconversion processes, Brussels: APME, 2002. Visited: 25 Feb. 2005\n<http://www.apme.org/dashboard/business_layer/template.asp?url=http:/\n/www.apme.org/media/public_documents/20021009_123742/EcoProfile\n_PVC_conversion_Oct2002.pdf>\n[3] I. Boustead, Eco-profiles of the European plastics industry: conversion\nprocesses for polyolefins, Brussels: APME, 2003. Visited: 25 Feb. 2005\n<http://www.apme.org/dashboard/business_layer/template.asp?url=http:/\n/www.apme.org/media/public_documents/20040610_153828/Polyolefin\nsConversionReport_Nov2003.pdf>\n[4] J. Mattis, P. Sheng, W. DiScipio, and K. Leong, A framework for\nanalyzing energy efficient injection-molding die design, Technical\nReport (CSM Report-96-09), California: University of California,\nBerkeley, 1996.\n[5] G. Boothroyd, P. Dewhurst, and W. Knight, Product design for\nmanufacture and assembly, New York: Marcel Dekker, 2002.\n[6] Dominick V. Rosato, Donald V. Rosato, and M.G. Rosato, Injection\nmolding handbook, 3rd ed. Springer-Verlag, 2000. Visited: 05 Aug.\n2005\n<http://www.knovel.com/knovel2/Toc.jsp?SpaceID=0&BookID=357&\nVerticalID=0&NodeID=1884960566&ClickedNodeID=1884960566&Fr\nomSearch=true&Action=expand&FilterMode=false&View=4&Collapse\nAll=false#node1884960566>.\n[7] Data file: cm test graphs. Sent by Mark Elsass (Cincinnati Milacron\nSupervisor for Technical Service). Received: 18 Apr. 2005.\n[8] Ferromatik Milacron, \"A communications update for ferromatik\nmilacron sales professionals,\u201d North America: 2001, unpublished. Sent\nby Mark Elsass (Cincinnati Milacron Supervisor for Technical Service).\nReceived on: Apr. 2005.\n[9] The previous reference was received as a word document. In it was\nimbedded an EXCEL file.\n[10] A.Thiriez, \"An Environmental Analysis of Injection Molding\u201d, Masters\nThesis at Massachussetts Institute of Technology, Cambridge, MA,\nUSA: 2006.\n[11] I. Boustead, Eco-profiles of the european plastics industry, Brussels:\nAPME, 2002-2003 Visited: 20 Mar. 2005 <http://www.apme.org/>.\n[12] K. Adams, J. Bankston, A. Barlow, M.W. Holdren, J. Meyer, and V.J.\nMarchesani, \"Development of emission factors for polypropylene\nprocessing,\" Journal of Air & Waste Management Association, 49\n(1999): 49-56.\n[13] S. Dalquist and T. Gutowski, Life cycle analysis of conventional\nmanufacturing techniques: die casting, Cambridge, USA: 2004.\nUnpublished. Available at: <http://web.mit.edu/ebm/publications.htm>.\n[14] S. Dalquist and T. Gutowski, \u201cLife cycle analysis of conventional\nmanufacturing techniques: sand casting,\" ASME International\nMechanical Engineering Congress and RD&D Expo, California, USA,\nNov. 2004. Available at: <http://web.mit.edu/ebm/publications.htm>.\n[15] M.R. Snyder, \"Electric injection machines inspire respect, loyalty,\"\nPlastics Machinery&Auxiliaries, Mar.-Apr. 2002. Visited: 18 Jan. 2005\n<http://www.pma-magazine.com/articles/2002/March/02>.\n[16] J.A. Brydson, Plastics Materials, 7th ed. Oxford, Great\nBritain: Butterworth-Heinemann, 1999.\n[17] Probe Economics, Inc, SPI Economic Report 2000,\nWashington, DC: The Society of the Plastics Industry, Inc., 2000.\n[18] I.I.\nRubin, Injection Molding Theory and Practice, New\nYork: John Wiley & Sons, 1972.\n[19] Energy Information Administration, International energy annual 2002,\nWashington: Department of Energy, 2002. Visited: 23 Feb. 2005\n<http://www.eia.doe.gov/iea/contents.html>\nNOTE: for the LCI results not all references that were used in the\ncalculations were listed. If the reader desires to obtain the references please\nrefer to [10].\n"}, "expected_output": {"claims": [{"unit": "MJ/kg", "value": 1.46, "evidence": ["This analysis yields average SEC values for hydraulic, hybrid and all-electric machines of 3.39, 1.67 and 1.46 MJ/kg respectively (without accounting for the efficiency of the electric grid).", "All-electrics on the other hand have very low fixed energy costs (ex: running the computers), and their SEC stays constant as throughput increases, as portrayed in Fig. 3."]}, {"unit": "MJ/kg", "value": 1.67, "evidence": ["This analysis yields average SEC values for hydraulic, hybrid and all-electric machines of 3.39, 1.67 and 1.46 MJ/kg respectively (without accounting for the efficiency of the electric grid).", "For hydraulic and hybrid machines SEC seems to exhibit a decreasing behavior with increasing throughput, as portrayed in Fig. 2. This derives from spreading fixed energy costs over more kilograms of polymer as throughput increases."]}, {"unit": "MJ/kg", "value": 3.39, "evidence": ["This analysis yields average SEC values for hydraulic, hybrid and all-electric machines of 3.39, 1.67 and 1.46 MJ/kg respectively (without accounting for the efficiency of the electric grid).", "For hydraulic and hybrid machines SEC seems to exhibit a decreasing behavior with increasing throughput, as portrayed in Fig. 2. This derives from spreading fixed energy costs over more kilograms of polymer as throughput increases."]}, {"unit": "MJ/kg", "value": 12.6, "evidence": ["The SEC values for hydraulic, hybrid and all-electric machines analyzed are 19.0, 13.2 and 12.6 MJ/kg respectively (including auxiliaries, compounding and the inefficiency of the electric grid).", "Considering the energy consumption of all stages from the compounder to the injection molder (not including polymer production), hydraulic, hybrid and all-electric machines yield average values for SEC of 19.0, 13.2 and 12.6 MJ/kg respectively."]}, {"unit": "MJ/kg", "value": 13.2, "evidence": ["The SEC values for hydraulic, hybrid and all-electric machines analyzed are 19.0, 13.2 and 12.6 MJ/kg respectively (including auxiliaries, compounding and the inefficiency of the electric grid).", "Considering the energy consumption of all stages from the compounder to the injection molder (not including polymer production), hydraulic, hybrid and all-electric machines yield average values for SEC of 19.0, 13.2 and 12.6 MJ/kg respectively."]}, {"unit": "MJ/kg", "value": 19, "evidence": ["The SEC values for hydraulic, hybrid and all-electric machines analyzed are 19.0, 13.2 and 12.6 MJ/kg respectively (including auxiliaries, compounding and the inefficiency of the electric grid).", "Considering the energy consumption of all stages from the compounder to the injection molder (not including polymer production), hydraulic, hybrid and all-electric machines yield average values for SEC of 19.0, 13.2 and 12.6 MJ/kg respectively."]}]}, "metadata": {"product_category": "Machinery & equipment", "request_id": "req_bfbad4eaf98f3469"}}
{"id": "ea7a1854187b49293b532e5f", "input": {"query": "What is the TOTAL manufacturing energy consumption (thermal energy) for producing Trosifol PVB film per kg or per m\u00b2 (inclusive of energy used in the upstream manufacturing processes?", "source_url": "https://www.trosifol.com/fileadmin/user_upload/about_us/sustainability/environmental-product-decleration-as-per-iso-14025-and-en-15804_a2-pvb-film-trosifol.pdf", "document_text": "ENVIRONMENTAL PRODUCT DECLARATION\n\n\nas per ISO 14025 and EN 15804+A2\n\n\nOwner of the Declaration Kuraray Europe GmbH\n\n\nInstitut Bauen und Umwelt e.V. (IBU)\n\n\nPublisher\n\n\nInstitut Bauen und Umwelt e.V. (IBU)\n\n\nProgramme holder\n\n\nEPD-KUR-20230072-CCI1-EN\n\n\nDeclaration number\n\n\nIssue date\n\n\n12/05/2023\n\n\nValid to\n\n\n11/05/2028\n\n\nPVB film (Trosifol\u24c7)\nKuraray Europe GmbH\n\n\nInstitut Bauen\nund Umwelt e.V.\n\n\nwww.ibu-epd.com | https://epd-online.com\n\n\nECO PLATFORM\n\n\nEPD\n\n\nVERIFIED\n\n\nALIEK\n\n\nkuraray\n\n\nGeneral Information\n\n\nKuraray Europe GmbH\nProgramme holder\n\n\nPVB film (Trosifol\u24c7)\n\n\nOwner of the declaration\nKuraray Europe GmbH\nPhilipp-Reis-Str. 4\n\n\nIBU - Institut Bauen und Umwelt e.V.\n\n\nHegelplatz 1\n10117 Berlin\n\n\n65795 Hattersheim\nGermany\n\n\nGermany\n\n\nDeclaration number\n\n\nDeclared product / declared unit\n\n\nTrosifol\u00ae PVB.\n\n\nEPD-KUR-20230072-CCI1-EN\n\n\nThe declared unit is 1 m\u00b2.\n\n\nThis declaration is based on the product category rules:\n\n\nScope:\n\n\nTrosifol PVB, manufactured in Troisdorf based on Mowital\u24c7 resin from\nFrankfurt.\n\n\nPlate glass for construction and interlayers, 01/09/2022\n(PCR checked and approved by the SVR)\n\n\nThe owner of the declaration shall be liable for the underlying information\nand evidence; the IBU shall not be liable with respect to manufacturer\ninformation, life cycle assessment data and evidences.\n\n\nIssue date\n\n\n12/05/2023\n\n\nThe EPD was created according to the specifications of EN 15804+A2. In\nthe following, the standard will be simplified as EN 15804.\n\n\nValid to\n\n\nVerification\n\n\n11/05/2028\n\n\nThe standard EN 15804 serves as the core PCR\nIndependent verification of the declaration and data according to ISO\n14025:2011\n\n\ninternally \u2611\n\n\nexternally\n\n\nNam Paten\n\n\nDipl.-Ing. Hans Peters\n\n\n(Chairman of Institut Bauen und Umwelt e.V.)\n\n\nHam Peter\nDipl.-Ing. Hans Peters\n\n\n\u041c\u0438\u043b\u0438\n\n\nDr. Matthew Fishwick,\n(Independent verifier)\n\n\n(Managing Director Institut Bauen und Umwelt e.V.)\n\n\nEnvironmental-Product Declaration - Kuraray Europe GmbH - PVB film (Trosifol\u24c7)\n\n\nkuraray\n\n\nProduct\n\n\nschueren/\n\n\nProduct description/Product definition\n\n\nConstructional data\n\n\nKuraray's polyvinyl butyral (PVB) thermoplastic films are tough,\nresilient safety interlayers used in laminated architectural safety\nglass. These Trosifol\u24c7 PVB interlayers offer safety advantages\nby retaining dangerous shards in case of glass breakage. They\nare commonly used as safety glass interlayers available\nworldwide.\n\n\nUnit\n\n\nValue\n\n\nName\n\n\nRefractive index acc. to DIN EN ISO 489\n\n\n1.48\n\n\n0.21 W/mK\n\n\nThermal conductivity acc. to DIN EN 993-15\nThermal expansion coefficient acc. to ISO\n11359-2\n\n\n0.00017 K-1\n\n\nSpecific heat capacity\n\n\n1.9 kJ/kgK\n1E+13 \u03a9\n\n\nThis EPD covers all Trosifol\u24c7 PVB products produced in\nTroisdorf based on Mowital\u24c7 resin produced in Frankfurt.\nProduct codes Trosifol\u24c7 B2XX and Trosifol B8XX.\n\n\nSurface resistivity acc. to DIN 53482\nTensile strength acc. to ISO 527-3\nElongation at break acc. to ISO 527-3\n\n\nN/mm\u00b2\n\n\n20\n\n\n250\n\n\n%\n\n\nTg acc. to DMA, 3K/min, 1Hz\n\n\nFor the use and application of the product, the respective\nnational provisions at the place of use apply, in Germany, for\nexample, the building codes of the federal states and the\ncorresponding national specifications.\n\n\n\u00b0C\n\n\n32\n\n\nPerformance data of the product with respect to its\ncharacteristics in accordance with the relevant technical\nprovision (no CE-marking).\n\n\nApplication\n\n\nBase materials/Ancillary materials\n\n\nTrosifol\u24c7 PVB film needs to be laminated between two pieces\nof glass. This sandwich arrangement is called laminated safety\nglass according to EN ISO 14449. Special Trosifol\u24c7 PVB\ngrades offer additional decorative, acoustic, UV managing and\nstructural properties.\n\n\nThe main constituents of Trosifol\u24c7 PVB film are (in mass\npercentages):\n\n\n-PVB resin ~72 %\n\n\n-Plasticizer ~27-28 %\n\n\n-Additives and water <1%\n\n\n1) This product contains substances listed in the candidate list\n(date: 17.01.2023) exceeding 0.1 percentage by mass: no\n\n\nTechnical Data\n\n\nFor calculating the light, solar and heat parameters of glazing\nspecifically containing films from the Trosifol\u24c7 & SentryGlas\u24c7\nproduct range, please go to:\n\n\n2) This product contains other Carcinogenic, Mutagenic,\nReprotoxic (CMR) substances in categories 1A or 1B which are\nnot on the candidate list, exceeding 0.1 percentage by mass:\n\n\nhttps://www.trosifol.com/winslt-tool/\n\n\nno\n\n\nSound Control data can be found here:\n\n\n3) Biocide products were added to this construction product or it\nhas been treated with biocide products (this then concerns a\ntreated product as defined by the (EU) Ordinance on Biocide\nProducts No. 528/2012): no\n\n\nhttps://www.trosifol.com/soundlab-ai/\n\n\nThe following data are valid for Trosifol\u24c7 Clear / Trosifol\u24c7\nUltraClear. Other product's data can be found in our laminator\nbrochure:\n\n\nReference service life\n\n\nThe reference service life is typically determined by the glass\nand not by the interlayer.\n\n\nhttps://www.trosifol.com/de/salessupport/downloads/produktbro\n\n\nLCA: Calculation rules\n\n\nproducts and energy, as well as waste processing up to the\nend-of-waste state or disposal of final residues during the\nproduct stage.\n\n\nDeclared Unit\n\n\nThis declaration refers to the declared unit of 1 m\u00b2 of PVB film\n(Trosifol\u24c7). The grammage of the PVB film is 0.775 kg/m\u00b2.\n\n\nThese modules consider the manufacturing of system\ncomponents/raw materials, the transport to the production site\nand the production processes of the products under study. The\nimpact of packaging materials is included.\n\n\nDeclared unit - PVB film (Trosifol\u24c7)\n\n\nName\n\n\nValue\n\n\nUnit\nm\u00b2\nkg/m^2\n\n\nDeclared unit\n\n\n1\n0.775\n0.00076\n1.07\n\n\nGrammage\n\n\nModule A5:\n\n\nLayer thickness\n\n\nm\n\n\nTreatment and disposal of packaging material. Credits for\npotential avoided burdens due to energy substitution of\nelectricity and thermal energy generation are declared in\nmodule D and affect only the rate of primary material (no\nsecondary materials).\n\n\nDensity\n\n\ng/cm^3\n\n\nSystem boundary\n\n\nThe type of EPD is cradle-to-gate with options, modules C1-\nC4, and module D (A1-A3, C, D and additional module A5). In\nthe following section, a detailed description of the specific\nsystem boundaries is given:\n\n\nModule C1 to C4:\n\n\nThe end-of-life scenarios are as follows:\n\n\nC1 Deconstruction/demolition: Dismantling is manual\n(no environmental burden).\n\n\n\u2022\n\n\nModule A1 to A3:\n\n\nThe product stage includes the provision of all materials,\n\n\nEnvironmental-Product Declaration - Kuraray Europe GmbH - PVB film (Trosifol\u24c7)\n\n\n2\n\n\nkuraray\n\n\n\u2022 C2 - Transport to treatment/disposal site: Average\ntransport distance from the demolition site to waste\ntreatment is assumed as 50 km to the landfill.\n\n\nLand or region, in which the declared product system is\nmanufactured, used or handled at the end of the product's\nlifespan: Europe\n\n\n\u2022 C3 - Waste processing for reuse, recovery or recycling:\nNo waste processing (no environmental burden).\n\u2022 C4 - Disposal: PVB film is 100% landfilled.\nModule D:\n\n\nComparability\n\n\nBasically, a comparison or an evaluation of EPD data is only\npossible if all the data sets to be compared were created\naccording to EN 15804 and the building context, respectively\nthe product-specific characteristics of performance, are taken\ninto account. Background datasets: GaBi ts 10.6 software\nsystem and GaBi Professional 2022.1 LCI database.\n\n\nFor the thermal and electrical energy generated in Module A5\ndue to the thermal treatment of packaging and product waste,\navoided burdens have been calculated by the inversion of the\nelectricity grid mix and thermal energy from natural gas, using\nEuropean datasets.\n\n\nGeographic Representativeness\n\n\nLCA: Scenarios and additional technical information\n\n\nEnd of life (C1-C4)\n\n\nCharacteristic product properties biogenic carbon\nBiogenic carbon is only present in the packaging (wooden\npallets and cartons).\n\n\nThe end-of-life scenarios are as follows:\n\n\nC1 The deconstruction of the PVB film is assumed to be done\nmanually. Therefore, no environmental loads for the dismantling\nof this product are considered.\nC2 - Transport to treatment/disposal site: Average transport\ndistance from the demolition site to waste treatment is assumed\nas 50 km to landfill.\n\n\nAssumed water content in wooden pallets (packaging): 18 %.\nAssumed carbon content: dry wood mass consists of 50 %\nbiogenic carbon and paper/cardboard 43 %.\n\n\nC4 - Disposal: The PVB film is 100% landfilled.\nName\n\n\nThe biogenic carbon content of the packaging is thus: 0.129 kg\npallet/declared unit * 0.82 * 0.5 kg C / kg pallet (abs. dry) +\n0.082 kg *0.43 kg C/kg cardboard= 0.08815 kg C/declared unit.\nInformation on describing the biogenic Carbon Content at\nfactory gate\n\n\nValue Unit\n\n\nCollected as mixed construction waste [PVB film\nper FU]\n\n\n0.775 kg\n\n\n0.775 kg\n\n\nLandfilling [PVB film per FU]\n\n\nName\n\n\nValue Unit\n0.08815\n\n\nBiogenic carbon content in accompanying\npackaging\n\n\nkg\nC\n\n\nReuse, recovery and/or recycling potentials (D), relevant\nscenario information\n\n\nFor the thermal and electrical energy generated in Module A5\ndue to the thermal treatment of packaging, avoided burdens\nhave been calculated by the inversion of the electricity grid mix\nand thermal energy from natural gas, using European datasets.\n\n\nThe following technical scenario information is required for the\ndeclared modules.\n\n\nInstallation into the building (A5)\n\n\nThe packaging material treatment and disposal are also\nconsidered in module A5.\n\n\nValue Unit\n\n\nName\n\n\nOutput substances following waste treatment on\nsite [packaging materials per FU]\n\n\n0.2301 kg\n\n\nEnvironmental-Product Declaration - Kuraray Europe GmbH - PVB film (Trosifol\u24c7)\n\n\n3\n\n\nkuraray\n\n\nLCA: Results\n\n\nDESCRIPTION OF THE SYSTEM BOUNDARY (X = INCLUDED IN LCA; ND = MODULE OR INDICATOR NOT DECLARED; MNR =\nMODULE NOT RELEVANT)\n\n\nBenefits and\nloads beyond\nthe system\nboundaries\n\n\nConstruction\nprocess stage\n\n\nProduct stage\n\n\nUse stage\n\n\nEnd of life stage\n\n\nOperational energy\n\n\nTransport from the\n\n\nOperational water\n\n\n2 Waste processing\nC3\n\n\nC1\n2 De-construction\n\n\nRefurbishment\n\n\n\u2611 gate to the site\nA4\n\n\nManufacturing\n\n\nRaw material\n\n\nReplacement\n\n\nMaintenance\n\n\nRecovery-\n\n\nRecycling-\n\n\ndemolition\n\n\nAssembly\n\n\nTransport\n\n\nTransport\n\n\nDisposal\n\n\npotential\n\n\nReuse-\n\n\nsupply\n\n\nRepair\n\n\nUse\n\n\nuse\n\n\nuse\n\n\nDX\n\n\nA2\n\n\nA3\n\n\nA5\n\n\nC2\n\n\nC4\n\n\nA1\n\n\nB1\n\n\nB2\n\n\nB3\n\n\nB4\n\n\nB5\n\n\nB6\n\n\nB7\n\n\nMND X MND MND MNR MNR MNR MND MND\n\n\n\u2717\n\n\n\u2717\n\n\nX\n\n\n\u2717\n\n\nRESULTS OF THE LCA - ENVIRONMENTAL IMPACT according to EN 15804+A2: 1 m\u00b2 PVB film (Trosifol\u24c7)\nParameter\n\n\nUnit\n\n\nA1-A3\n\n\nA5\n\n\nC2\n\n\nC3\n\n\nC1\n\n\nC4\n\n\nD\n\n\nkg CO2 eq\n\n\n3.26E+00\n\n\n2.37E-03\n\n\n-1.19E-01\n\n\nGWP-total\n\n\n3.77E-01\n\n\n0\n\n\n0\n\n\n5.49E-02\n\n\nkg CO2 eq\n\n\nGWP-fossil\n\n\n2.37E-03\n\n\n3.55E+00\n\n\n5.77E-02\n\n\n0\n\n\n0\n\n\n5.49E-02\n\n\n-1.18E-01\n\n\nGWP-biogenic\n\n\nkg CO2 eq\n\n\n3.2E-01\n\n\n-2.95E-01\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n-6.05E-04\n\n\nkg CO2 eq\n\n\n9.09E-06\n\n\nGWP-luluc\n\n\n5.4E-04\n\n\n5.85E-06\n\n\n0\n\n\n2.67E-05\n\n\n-1.3E-05\n\n\n0\n\n\nkg CFC11 eq\n\n\n1.91E-09\n\n\nODP\n\n\n5.63E-14\n\n\n0\n\n\n4.84E-16\n\n\n7.38E-14\n\n\n-7.99E-13\n\n\n0\n\n\n4.67E-03\n\n\nmol H+ eq\n\n\n1.63E-04\n\n\nAP\n\n\n7.13E-05\n\n\n0\n\n\n7.48E-06\n\n\n0\n\n\n-1.56E-04\n\n\nEP-freshwater\n\n\nkg P eq\nkg N eq\nmol N eq\nkg NMVOC\neq\nkg Sb eq\nMJ\nm\u00b3 world eq\ndeprived\n\n\n-1.63E-07\n\n\n2.99E-05\n\n\n1.62E-08\n\n\n0\n\n\n4.71E-09\n\n\n0\n\n\n1.03E-05\n\n\nEP-marine\nEP-terrestrial\n\n\n3.5E-06\n\n\n-4.22E-05\n\n\n1.47E-03\n\n\n2.45E-05\n\n\n0\n\n\n3.6E-05\n\n\n0\n\n\n3.28E-04\n\n\n1.64E-02\n\n\n3.95E-04\n\n\n0\n\n\n3.9E-05\n\n\n0\n\n\n-4.52E-04\n\n\n6.56E-05\n\n\nPOCP\n\n\n5.78E-03\n\n\n0\n\n\n6.82E-06\n\n\n0\n\n\n1.16E-04\n\n\n-1.18E-04\n\n\n-1.78E-08\n\n\nADPE\n\n\n1.61E-06\n\n\n1.45E-09\n\n\n0\n\n\n2.36E-10\n\n\n0\n\n\n3.81E-09\n\n\nADPF\n\n\n7.95E+01\n\n\n1.39E-01\n\n\n0\n\n\n3.13E-02\n\n\n7.79E-01\n\n\n-2.01E+00\n\n\n0\n\n\nWDP\n\n\n3.87E-02\n\n\n-7.12E-01\n\n\n0\n\n\n1.01E-05\n\n\n0\n\n\n-5.42E-04\n\n\n-1.26E-02\n\n\nGWP = Global warming potential; ODP = Depletion potential of the stratospheric ozone layer; AP = Acidification potential of land and water; EP =\nEutrophication potential; POCP = Formation potential of tropospheric ozone photochemical oxidants; ADPE = Abiotic depletion potential for non-fossil\nresources; ADPF = Abiotic depletion potential for fossil resources; WDP = Water (user) deprivation potential)\nRESULTS OF THE LCA - INDICATORS TO DESCRIBE RESOURCE USE according to EN 15804+A2: 1 m\u00b2 PVB film (Trosifol\u24c7)\nParameter\n\n\nA5\n45\n3.42E+00\n\n\nA1-A3\n1.03E+01\n\n\nUnit\nMJ\n\n\nC1\n\n\nC2\n\n\nC4\n\n\nD\n\n\nC3\n\n\n2.06E-03\n\n\nPERE\n\n\n0\n\n\n6.41E-02\n0\n\n\n-5.52E-01\n\n\n0\n\n\n3.39E+00\n\n\nPERM\n\n\n-3.39E+00\n\n\n0\n\n\nMJ\n\n\n0\n\n\n0\n\n\n0\n-5.52E-01\n-2.01E+00\n0\n-2.01E+00\n\n\n|PERT\n\n\n1.37E+01\n\n\nMJ\n\n\n3.22E-02\n\n\n0\n\n\n2.06E-03\n\n\n0\n\n\n6.41E-02\n\n\n2.5E+01\n\n\nPENRE\n\n\nMJ\n\n\n5.47E+01\n\n\n6.94E-01\n\n\n0\n\n\n3.13E-02\n\n\n0\n\n\n-2.43E+01\n\n\nPENRM\n\n\n2.48E+01\n\n\n-5.55E-01\n\n\n0\n\n\n0\n\n\n0\n\n\nMJ\n\n\nPENRT\n\n\n7.96E+01\n\n\n1.39E-01\n\n\n0\n\n\n3.13E-02\n\n\n0\n\n\n7.8E-01\n\n\nMJ\n\n\nSM\n\n\nkg\nMJ\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\nRSF\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\nNRSF\n\n\nMJ\nm\u00b3\n\n\n0\n1.64E-02\n\n\n0\n\n\n0\n\n\n0\n0\n\n\n0\n\n\n0\n\n\n0\n9.16E-04\n\n\n1.6E-06\n\n\n1.01E-05\n\n\n-5.31E-04\n\n\nFW\n\n\n0\n\n\nPERE = Use of renewable primary energy excluding renewable primary energy resources used as raw materials; PERM = Use of renewable primary\nenergy resources used as raw materials; PERT = Total use of renewable primary energy resources; PENRE = Use of non-renewable primary energy\nexcluding non-renewable primary energy resources used as raw materials; PENRM = Use of non-renewable primary energy resources used as raw\nmaterials; PENRT = Total use of non-renewable primary energy resources; SM = Use of secondary material; RSF = Use of renewable secondary fuels;\nNRSF = Use of non-renewable secondary fuels; FW = Use of net fresh water\n\n\nRESULTS OF THE LCA \u2013 WASTE CATEGORIES AND OUTPUT FLOWS according to EN 15804+A2:\n\n\n1 m\u00b2 PVB film (Trosifol\u00ae)\n\n\nParameter\n\n\nUnit\n\n\nA1-A3\n8.63E-08\n\n\nA5\n1.17E-11\n\n\nC3\n\n\nC1\n\n\nC2\n\n\nC4\n\n\nD\n\n\n1.2E-10\n7.72E-01\n9.58E-06\n\n\nkg\n\n\n0\n\n\n1.37E-13\n\n\n-2.73E-10\n\n\nHWD\n\n\n0\n\n\n-1.02E-03\n\n\nkg\n\n\n3.89E-02\n\n\n2.29E-02\n\n\n0\n\n\n5.1E-06\n\n\nNHWD\n\n\n0\n\n\n3.94E-08\n\n\nkg\n\n\n7.58E-04\n\n\n6.07E-06\n\n\n0\n\n\n0\n\n\n-1.58E-04\n\n\nRWD\n\n\nCRU\n\n\nkg\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\nMFR\n\n\nkg\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\nMER\n\n\nkg\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n5.32E-01\n\n\nEEE\n\n\nMJ\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\nEET\n\n\nMJ\n\n\n0\n\n\n9.58E-01\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\n0\n\n\nEnvironmental-Product Declaration - Kuraray Europe GmbH - PVB film (Trosifol\u24c7)\n\n\n4\n\n\nkuraray\n\n\nHWD = Hazardous waste disposed; NHWD = Non-hazardous waste disposed; RWD = Radioactive waste disposed; CRU = Components for re-use;\nMFR = Materials for recycling; MER = Materials for energy recovery; EEE = Exported electrical energy; EET = Exported thermal energy\n\n\nRESULTS OF THE LCA - additional impact categories according to EN 15804+A2-optional:\n\n\n1 m\u00b2 PVB film (Trosifol\u00ae)\n\n\nParameter\n\n\nUnit\nDisease\nincidence\n\n\nA1-A3\n\n\nA5\n\n\nC1\n\n\nC2\n\n\nC3\n\n\nC4\n\n\nD\n\n\nPM\n\n\n6.26E-08\n\n\n5.26E-10\n\n\n0\n\n\n4.26E-11\n\n\n0\n\n\n1.56E-09\n\n\n-1.29E-09\n\n\nIR\n\n\nkBq U235 eq\nCTUe\n\n\n7.89E-02\n\n\n9.02E-04\n\n\n0\n\n\n0\n\n\n1.41E-03\n\n\n-2.68E-02\n\n\n4E-06\n\n\n8.63E-02\n\n\n7.63E-01\n\n\nETP-fw\n\n\n6.53E+01\n\n\n0\n\n\n2.4E-02\n\n\n0\n\n\n-4.41E-01\n\n\n-2.03E-11\n\n\nHTP-C\nHTP-nc\n\n\nCTUh\n\n\n9.32E-09\n\n\n4.13E-12\n\n\n0\n\n\n4.81E-13\n\n\n0\n\n\n3.42E-11\n\n\nCTUh\nSQP\n\n\n4.94E-08\n4.91E+01\n\n\n2.77E-11\n\n\n2.79E-10\n3.71E-02\n\n\n0\n\n\n0\n\n\n2.87E-09\n\n\n-7.8E-10\n\n\nSQP\n\n\n0\n\n\n9.38E-03\n\n\n0\n\n\n5.61E-02\n\n\n-3.59E-01\n\n\nPM = Potential incidence of disease due to PM emissions; IR = Potential Human exposure efficiency relative to U235; ETP-fw = Potential comparative\nToxic Unit for ecosystems; HTP-c = Potential comparative Toxic Unit for humans (cancerogenic); HTP-nc = Potential comparative Toxic Unit for humans\n(not cancerogenic); SQP = Potential soil quality index\n\n\nDisclaimer 1 - for the indicator \"Potential Human exposure efficiency relative to U235\". This impact category deals mainly with the\neventual impact of low-dose ionizing radiation on human health of the nuclear fuel cycle. It does not consider effects due to possible\nnuclear accidents, occupational exposure or radioactive waste disposal in underground facilities. Potential ionizing radiation from the\nsoil, radon and from some construction materials is also not measured by this indicator.\n\n\nDisclaimer 2 - for the indicators 'abiotic depletion potential for non-fossil resources', 'abiotic depletion potential for fossil resources',\n'water (user) deprivation potential, deprivation-weighted water consumption', 'potential comparative toxic unit for ecosystems', 'potential\ncomparative toxic unit for humans - cancerogenic', 'Potential comparative toxic unit for humans - not cancerogenic', 'potential soil\nquality index'. The results of this environmental impact indicator shall be used with care as the uncertainties on these results are high\nas there is limited experience with the indicator.\n\n\nReferences\n\n\nEN ISO 14025:2011, Environmental labels and declarations \u2014\nType III environmental declarations - Principles and\nprocedures.\n\n\nStandards\n\n\nDIN 53482\n\n\nDIN 53482:1967-01, Testing of Insulating Materials;\nDetermination of Electrical Resistances Values.\n\n\nFurther References\n\n\nDIN EN ISO 489\n\n\nDIN EN ISO 489:1999-08, Plastics - Determination of the\nrefractive index (ISO 489:1999).\n\n\nCandidate list\n\n\nCandidate List of substances of very high concern for\nAuthorisation, published on ECHA website, latest version\n17.01.2023 (https://echa.europa.eu/candidatelist-table)\n\n\nDIN EN 993-15\n\n\nDIN EN 993-15:2005-07, Methods of test for dense shaped\nrefractory products - Part 15: Determination of thermal\nconductivity by the hot-wire (parallel) method.\n\n\nGaBi\n\n\nGaBi Software System and Database for Life Cycle\nEngineering, 1992-2021, Sphera Solutions GmbH, Leinfelden-\nEchterdingen, with acknowledgement of\nLBP University of Stuttgart, program version GaBi 10; database\nversion 2022.1.\n\n\nDIN EN ISO 527-3\n\n\nDIN EN ISO 527-3:2019-02, Plastics - Determination of tensile\nproperties - Part 3: Test conditions for films and sheets (ISO\n527-3:2018).\n\n\nGaBi documentation\n\n\nGaBi dataset documentation for the software system and\ndatabases, LBP, University of Stuttgart and Sphera Solutions\nGmbH, Leinfelden-Echterdingen,\n\n\nEN 15804\n\n\nEN15804:2012+A1:2013, Sustainability of construction works\n- Environmental Product Declarations Core rules for the\nproduct category of construction products.\n\n\n-\n\n\n2021. (http://www.gabi-\n\n\nsoftware.com/support/gabi/gabi[1] database-2021-lci-\n\n\ndocumentation/)\n\n\nEN 15804\n\n\nEN 15804:2012+A2:2019+AC:2021, Sustainability of\nconstruction works - Environmental Product Declarations -\nCore rules for the product category of construction products.\n\n\nIBU 2021\n\n\nInstitut Bauen und Umwelt e.V.: General Instructions for the\nEPD programme of Institut Bauen und Umwelt e. V., Version\n2.0, Berlin: Institut Bauen und Umwelt e. V., 2021 HYPERLINK\n\"http://www.ibu-epd.\" www.ibu-epd.com\n\n\nEN ISO 14449\n\n\nEN 14449:2005/AC:2005, Glass in building - Laminated glass\nand laminated safety glass - Evaluation of conformity/Product\nstandard.\n\n\nOrdinance on Biocide Products No. 528/2012\nRegulation (EU) No 528/2012 of the European Parliament and\nof the Council of 22 May 2012 concerning the making available\non the market and use of biocidal products\n\n\nISO 11359-2\n\n\nISO 11359-2:2021-11, Plastics - Thermomechanical analysis\n(TMA) - Part 2: Determination of coefficient of linear thermal\nexpansion and glass transition temperature.\n\n\nPCR Part A\n\n\nPCR Part A: Calculation rules for the Life Cycle Assessment\nand Requirements on the Background Report according to EN\n\n\nISO 14025\n\n\nEnvironmental-Product Declaration - Kuraray Europe GmbH - PVB film (Trosifol\u24c7)\n\n\n5\n\n\nkuraray\n\n\n15 804+A2:2019, Version 1.3, Institut Bauen und Umwelt e.V.,\n2020.\n\n\nProduct Category Rules for Building Products, Part B:\n\n\nRequirements on the EPD for plate glass for construction and\ninterlayers, version 1.6, 2022 www.bau-umwelt.de\n\n\nPCR Part B\n\n\nEnvironmental-Product Declaration - Kuraray Europe GmbH - PVB film (Trosifol\u24c7)\n\n\n6\n\n\nkuraray\n\n\nPublisher\n\n\nInstitut Bauen und Umwelt e.V.\n\n\n+49 (0)30 3087748-0\ninfo@ibu-epd.com\nwww.ibu-epd.com\n\n\nHegelplatz 1\n10117 Berlin\nGermany\n\n\nInstitut Bauen\nund Umwelt e.V.\n\n\nProgramme holder\n\n\n+49 (0)30 3087748-0\ninfo@ibu-epd.com\nwww.ibu-epd.com\n\n\nInstitut Bauen und Umwelt e.V.\nHegelplatz 1\n\n\n10117 Berlin\nGermany\n\n\nInstitut Bauen\nund Umwelt e.V.\n\n\nAuthor of the Life Cycle Assessment\n\n\nSphera Solutions GmbH\n\n\n+49 711 341817-0\ninfo@sphera.com\nwww.sphera.com\n\n\nsphera\u24c7\n\n\nHauptstra\u00dfe 111-113\n\n\n70771 Leinfelden-Echterdingen\nGermany\n\n\nOwner of the Declaration\n\n\nKuraray Europe GmbH\nPhilipp-Reis-Str. 4\n\n\n+49 69 305 85 300\ntrosifol@kuraray.com\nhttps://www.kuraray.eu/\n\n\nkuraray\n\n\n65795 Hattersheim\nGermany\n\n\nEnvironmental-Product Declaration - Kuraray Europe GmbH - PVB film (Trosifol\u24c7)\n\n\n7\n"}, "expected_output": {"claims": [{"unit": "MJ", "value": 65, "evidence": ["PERE", "A1-A3\n1.03E+01", "Unit\nMJ", "5.47E+01\n", "The type of EPD is cradle-to-gate with options, modules C1-\nC4, and module D (A1-A3, C, D and additional module A5). In\nthe following section, a detailed description of the specific\nsystem boundaries is given:\n", "Module A1 to A3:\n", "PENRE", "MJ"]}]}, "metadata": {"product_category": "Paper and plastic products", "request_id": "req_53c2ae0920346811"}}
{"id": "4cd8760411a0e2be20a6ca3d", "input": {"query": "What is the energy consumption for extrusion and pelletizing of PET in MJ/kg or kWh/kg?", "source_url": "https://environmentalclarity.com/wp-content/uploads/2024/05/Life-Cycle-Energy-Comparison-of-Different-Polymer-Recycling-Processes.pdf", "document_text": "Accepted: 17 October 2019\n\n\nReceived: 17 January 2019\n\n\nRevised: 2 October 2019\n\n\nDOI: 10.1002/amp2.10034\n\n\nJOURNAL OF\n\n\nADVANCED\n\n\nMANUFACTURING WILEY\n\n\nRESEARCH ARTICLE\n\n\nAND PROCESSING\n\n\nLife cycle energy comparison of different polymer\n\n\nrecycling processes\n\n\nMichael R. Overcash \u2460 | James H. Ewell | Evan M. Griffing\n\n\nEnvironmental Genome Initiative,\nRaleigh, North Carolina\n\n\nAbstract\n\n\nThis article is to demonstrate a consistent, transparent approach to comparing\nplastic recycling technologies. The uniform comparison is based on each recycling\ntechnology having the same input (waste PET), mass basis for processing, output\nas new product (new PET product or fuels), and the concept of the same multiple\n(two) closed loops of recycling. We seek to demonstrate, at the fundamental tech-\nnology level, how energy use differentiates plastic recycling technologies. Five\npolymer-recycling processes are examined using a uniform, quantitative compari-\nson of 1 kg PET bottles (about 100 single-serve 0.5 L water bottles): direct reuse,\n100% mechanical recycled content, depolymerization, and re-polymerization of\nnew resin and 100% to bottles, reclaiming energy value, and landfill. The life cycle\nenergy benefit for recycle technologies with varying product recycled content can\nbe determined by a single equation. All these recycling processes resulted in total\nenergy reduction per kg PET bottles compared to landfilling. The base case of\nthree cycles per 1 kg PET bottles is used to explore the influence of recycling\nloops. Direct reuse gave a 290% energy improvement with three cycles. Other\nprocesses, all at 100% recycle content, gave improvements: mechanical (250%),\ndepolymerization/repolymerization (150%), and energy recovery (120%). More\ninformation would improve the analysis of the depolymerization process assess-\nment. These preliminary data describe the analyses that are needed to quantify\nthe benefit of recycling any polymer using these recycling methods. The Environ-\nmental Genome (\u201cEGI\u201d) provides valuable information for these calculations as it\ncontains the polymers and supply chains for such evaluations.\n\n\nCorrespondence\n\n\nMichael R. Overcash, Environmental\nGenome Initiative, 2908 Chipmunk Lane,\nRaleigh, NC 27607.\n\n\nEmail: mrovercash@earthlink.net\n\n\nKEYWORDS\n\n\ndepolymerization, mechanical recovery, polymer recycling, recycle content, recycle loops\n\n\nnylon, polyethylene, [5] and PET. However, a quantita-\ntive, energy-based comparison of the major approaches to\nreuse/recycle would be essential to establishing the mea-\nsurable benefits of such technologies. In these polymer-\nrecycling reviews, the mechanical recycling (grinding,\nsorting) is covered for thermoplastics, which can be re-\nmelted and injection molded or extruded. Thermosets are\n\n\n[4]\n\n\n[1]\n\n\n1 INTRODUCTION\n\n\nThe field of chemical recycling for polymers has a number\nof existing commercial and developing technologies. These\nhave been reviewed qualitatively as separate technologies\nfor process descriptions and importance for a circular\neconomy with plastics. [1,2] Reviews cover polypropylene, [\u00b3]\n\n\n\u00a9 2019 American Institute of Chemical Engineers\n\n\nwileyonlinelibrary.com/journal/amp2\n\n\nJ Adv Manuf Process. 2020;2:e10034.\nhttps://doi.org/10.1002/amp2.10034\n\n\n1 of 13\n\n\nJOURNAL OF\n\n\nADVANCED\n-MANUFACTURING\nAND PROCESSING\n\n\nOVERCASH ET AL.\n\n\n2 of 13\n\n\nWILEY-\n\n\nsorted and ground into a fine material that is used as filler\nwith virgin resin. Reviews also describe depolymerization\nto monomers and co-monomers, followed by purification\nto obtain monomers to repolymerize into new resin with\nexactly the same quality as virgin resin. The article by\nGeyer, et al[6] provides a comprehensive evaluation of some\nof the central concepts that emerge from evaluating plastic\nrecycling technologies, but without quantitative data.\n\n\nThe objective of this article is to demonstrate a consis-\ntent, transparent approach to comparing plastic recycling\ntechnologies. The uniform comparison is based on each of\nthe recycling technologies having the same input (waste\nPET), mass basis for processing, output as a new product\n(new PET product or fuels), and the concept of the same\nmultiple loops of recycling [two]. We seek to show trans-\nparently, at the fundamental technology level, how energy\nuse differentiates plastic recycling technologies. The Envi-\nronmental Genome database is essential for the kind of\npolymer supply chain and recycling information needed\nfor such comparisons. This article is to quantitatively\nexamine the hypothesis: The alternatives in polymer\nrecycling demonstrate degrees of circularity in reuse, recy-\ncle, or recovery of chemicals or fuel value which then\nshow a quantitative difference in a continuum from low to\nhigh energy improvement. This continuum is also\ninfluenced by the number of recycle loops (eg, depolymeri-\nzation may give very large numbers of loops, while\nmechanical recycling might suffer from polymer degrada-\ntion and have fewer loops). For this preliminary study, the\nyield and reusability of the recycled products are assumed\nto be 100%. The complexity of collection and segregation\nare recognized as a common challenge for all the technolo-\ngies herein and hence is a common factor that is not\naddressed further. Refinement of less than 100% yield is\nvery company-specific and not within the scope of this ini-\ntial comparison. The high transparency herein permits a\ndirect assessment of actual recycling technology results\nand serves to support the collection of more in-depth data\nfor updating this preliminary comparison.\n\n\n2 APPROACH AND OBJECTIVES\n\n\nRecycling environmental benefits are linked to knowledge\nof chemical manufacturing of virgin inputs. The molecular\nbuilding processes beginning with natural resources in the\nearth produce virtually all the 100 000 chemicals used in-\ncommerce.\n[7] These processes require energy inputs (steam,\nDowtherm heating, furnaces, transport fuels, and potential\nenergy recovery) and operate at less than 100% mass effi-\nciency (losses as wastes or emissions to the environment).\nManufacturing process energies are made available by con-\nsumption of fuels, Table 1. For each mega Joule (MJ) of\nenergy put into the chemical manufacturing process, there\nare direct fuels needed to create that MJ of energy and fuel\nis also needed to deliver these fuels to the point of use, as\ndescribed in Table 1. In this article, the energy values in\nTable 1 are used and referred to as natural resource\nenergy (nre).\nMolecular building often occurs in separate chemical\nplants, over geographic and temporal domains. Value\nchains are assemblages of these chemical processes and\nwere discovered to have a repeatable pyramidal struc-\nture. There are various points into which a recovered\nchemical can be substituted for the same virgin chemical\n(in the same or a different product). If recycling is not\ninvolved, these supply chains are referred to as virgin\nchemicals.\n\n\n[7]\n\n\n| RESULTS\n\n\n3\n\n\nIn order to calculate an energy benefit for chemical\nrecycling we keep track of two energy categories:\n\n\nRelationship of MJ energy used in chemical manufacturing processes to MJ total natural resource energy (nre) consumed to\nproduce that energy[8]\n\n\nTABLE 1\n\n\nNon-transport\ndirect use\nof fuel\n\n\nHeat\n\n\nTransport potential\nfuel\nrecovery\n\n\nScale-up factors\n\n\nElectricity Dowtherm\n\n\nSteam\n\n\nPrecombustion factors, MJ fuel extracted and used\nper MJ delivered (This excess is consumed in\ndelivery)\n\n\n1.15\n\n\n1.15\n\n\n1.1\n\n\n1.15\n\n\n1.20\n\n\n1.15\n\n\nGeneration/combustion factors, MJ HHV fuel\n\n\n1.25\n\n\n1.25\n\n\n1.25\n\n\n3.13\n\n\n1.00\n\n\n1.00\n\n\ndelivered per MJ energy to process\n\n\nTotal scale up factor (precombustion times\n\n\n1.44\n\n\n1.20\n\n\n3.44\n\n\n1.44\n\n\n1.15\n\n\n1.44\n\n\ngeneration/combustion), MJ total fuel consumed\n\n\nfor this use per MJ into process, nre\n\n\naBased on United State energy grid.\n\n\nJOURNAL OF\n\n\nWILEY\n\n\nADVANCED\n\n\nOVERCASH ET AL.\n\n\n3 of 13\n\n\n-MANUFACTURING-\n\n\nAND PROCESSING\n\n\nLevel 9\n\n\nLevel 5\n\n\nLevel 4\n\n\nLevel 3\n\n\nLevel 0\n\n\nLevel 8\n\n\nLevel 7\n\n\nLevel 6\n\n\nLevel 2\n\n\nLevel 1\n\n\nPET melt,\nfrom TPA\n\n\nPET pellet,\n\n\nethylene\nglycol\n\n\noil (in ground)\n186\n\n\nbottle, PET from TPA from TPA\n1,000\n\n\nethylene oxide ethylene\n243\n\n\nnaphtha\n\n\n1,000\n\n\n336\n\n\n188\n\n\n1,000\n\n\n183\n\n\nair\n\n\noxygen (untreated)\n\n\n178\n\n\n179\n\n\nwater\n\n\nwater for rxn\n\n\n(untreated)\n\n\n94.3\n\n\n94.3\n\n\nnatural gas\n(unprocessed)\n\n\nterephthalic terephthalic\n\n\ncarbon\n\n\nnatural gas\n\n\nacid\n\n\nacid, crude\n\n\nacetic acid monoxide carbon dioxide\n\n\n30.2\n\n\n3.17\n\n\n858\n\n\n870\n\n\n57.5\n\n\n15.1\n\n\n3.11\n\n\nnitrogen from air air (untreated)\n5.81\n\n\n5.81\n\n\noxygen from air air (untreated)\n2.58\n\n\n2.58\n\n\nwater (untreated)\n\n\nwater for rxn\n\n\n4.03\n\n\n4.03\n\n\nnatural gas\n(unprocessed)\n\n\nnatural gas\n\n\n10.2\n\n\n10.4\n\n\nwater (untreated)\n\n\nwater for rxn\n\n\n6.17\n\n\n6.17\n\n\nnatural gas\n(unprocessed)\n\n\nnatural gas\n\n\nmethanol\n\n\n31.6\n\n\n16.2\noxygen from air air (untreated)\n16.2\n\n\n16.6\n\n\n16.2\n\n\nwater for rxn\n\n\nwater (untreated)\n\n\n29.0\n\n\n29.0\n\n\noxygen air\nfrom air\n\n\n(untreated)\n\n\n703\n\n\n703\n\n\noil (in ground)\n\n\np-xylene hydrogen naphtha\n\n\n582\n\n\n0.936\n\n\n3.33\n\n\n3.28\n\n\nair (untreated)\n\n\noxygen\n\n\n3.28\n\n\n3.28\n\n\noxygen from air air (untreated)\n2.15\n\n\n2.15\n\n\nwater for rxn\n\n\nwater (untreated)\n\n\n1.43\n\n\n1.43\n\n\npyrolysis gas naphtha\n\n\noil (in ground)\n\n\nxylenes\n\n\n580\n\n\n168\nreformate, from\nnaphtha\n\n\n172\n\n\n174\n\n\noil (in ground)\n\n\nnaphtha\n\n\n420\n\n\n413\n\n\n425\n\n\nSupply chain\n\n\nenergy by stage,\nMJ nre/1,000 kg\n\n\n3,009 10,305\n\n\n10,064\n\n\n6,768\n\n\n11,959\n\n\n2,854\n\n\nPET bottle\n\n\n180\n\n\n5,162\n\n\n1,935\n\n\nFIGURE 1 Supply chain from natural resources in the earth (salmon-colored cells), 1866 kg to make 1000 kg of PET bottles, numeric\nvalues are mass flows in kg (bottom row shows the energy of each level of supply chain, which total to 52.2 MJ nre/kg PET). PET,\npolyethylene terephthalate [8,9]\n\n\nchemical plant that occurs in the supply chain (natu-\nral resources in the earth to the final product). When\nevaluating the energy benefits of using recycled con-\ntent by a particular process, we subdivide this category\ninto two concepts\na. Energy required to make 1 kg of polymer product\nfrom virgin inputs (solid lines in Figures below).\nb. Energy to process waste polymer products at end-\nof-life into usable chemicals, materials, or fuels\n(dotted lines in Figures below).\n\n\n1. Category 1: The energy of fossil resources used to form\na polymer structure (feedstock-salmon colored cells,\nFigure 1). These are typically crude oil (45 MJ/kg) and\nnatural gas (53.5 MJ/kg) in various proportions (inor-\nganic feedstocks are not tabulated as category 1 energy).\nThese go into the polymer structure, and are thus fossil\nmaterials not available for other uses and thus are\nexpressed as consumption of fuel resources.\n2. Category 2: The energy needed for the entire set of\nunit processes (distillation, reactors, etc.) in each\n\n\nJOURNAL OF\n\n\nADVANCED\n-MANUFACTURING-\nAND PROCESSING\n\n\nOVERCASH ET AL.\n\n\n4 of 13\n\n\nWILEY-\n\n\nIt is categories 1 and 2a that can be accessed easily\nand transparently with the methods in the Environmen-\ntal Genome (www.environmentalgenome.org), but cate-\ngory 2b can use more specific technology data.\nThese recycling scenarios would warrant an article on\neach to explain the complexities of the process, process flow\ndiagrams, the calculation of the energies and mass losses,\nand application to multiple polymers. This article provides\njust a summary description with energy values that can be\nexplained and followed directly in the calculations. Issues\nnot directly related to a uniform comparison of recycling\ntechnologies are not included herein, such as mis-\nmanagement at end-of-life, public issues of trash or marine\npollution, ongoing recycling rates at the State or national\nlevel, detailed differences in equipment, social benefits of\njobs, and so forth. This comparison, begins with cleaned\nplastics, therefore, all the energy associated with the collec-\ntion, sorting, or cleaning prior to recycling are not included.\nFigure 1 shows the supply chain of each chemical\nneeded to produce 1000 kg of PET bottles, as found in the\nenvironmental genome database, while in this article 1 kg\nof PET bottles is used to reflect a more consumer-based\nperspective. Each stage represents the addition of new\nchemical inputs with the associated mass shown starting\nwith extraction of raw materials to finished PET bottles.\nAt the bottom of Figure 1 the sum of the energy per\n1000 kg of PET required for each stage is provided, which are\nthe values on each of the following Figures to create product\nA (1 kg of PET). For PET there are nine stages from the raw\nmaterials in the earth to the virgin product. The zero stage is\nnatural resources from the earth (salmon-colored cells). For\nPET bottles the total natural resources are about 1.88 kg to\n\n\nmake 1 kg of PET bottles (~100 single-serve 0.5 L water bot-\ntles). This is in the range of other chemical supply chain ratios\n(typical range 1.5-4.5). For PET, all products that do not use\nrecycled material must include fossil resources for new prod-\nucts and thus include the category 1 energies, 37.2 MJ/kg\nPET bottles (9) (calculated from Figure 1 as the mass of crude\noil and natural gas as natural resources that go into the final\nPET structure. The energy for crude oil (45 MJ/kg) and natu-\nral gas (53.5 MJ/kg) are in Appendix 1 and used to arrive at\nthe 37.2 MJ nre/kg PET bottle.\nThe scenarios that follow are examples from the four\nmain categories of plastics recycling.\n\n\n[9]\n\n\n3.1 | Scenario 1: Virgin inputs with\nbottle to landfill\n\n\nThe first, second, and third PET bottles are made identi-\ncally from fossil resources in the earth. This base case is\nPET bottles going to landfill or becoming land or ocean lit-\nter. The replacement bottle from the virgin supply chain\nbegins with the 37.2 MJ category 1 energy reflecting the\nfossil resources in the actual PET, Figure 2. The supply\nchain processes then convert the fossil resources into the\nmonomers (terephthalic acid and ethylene glycol) and\npolymerize these into 1 kg of polyethylene terephthalate\n(PET) resin, which is pelletized and injection molded into\n1 kg PET bottles. The category 2a energy for the PET bot-\ntles is 52.2 MJ nre/kg PET bottles. The first, second, and\nthe third PET bottles are each 89 MJ nre/kg PET bottle\n(3.4 kg CO2eq/kg PET bottles). The shape of the cumulative\nenergy supply chain curve, Figure 2, points to potential\n\n\nNon-recycle product case with bottle to landfill\n\n\n300\n\n\nCumulative energy (nre) from natural resources in the\n\n\n285\n\n\n270\n\n\n255\n\n\n240\n\n\nearth to PET bottles 1, 2 and 3, MJ\n\n\n225\n\n\n210\n\n\nProduct A3,\n89 MJ/kg\nPET product\n\n\n195\n\n\n180\n\n\n165\n\n\n150\n\n\n135\n\n\nProduct A2,\n89 MJ/kg\nPET product\n\n\n120\n\n\n105\n\n\n90\n\n\n75\n\n\n60\n\n\nProduct A1,\n\n\n45\n\n\n89 MJ/kg\n\n\n30\n\n\nPET product\n\n\n15\n\n\n0\n\n\n0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 89\nSteps of supply chain and recycling technology from natural resources in the\nearth through to PET bottles 1, 2, and 3\n\n\nFIGURE 2\n\n\nVirgin inputs Bottle\n\n\nto landfill. Note: the small energy for\n\n\nlandfilling is not included for simplicity\n\n\nJOURNAL OF\n\n\nMANUFACTURING\u2015 WILEY\n\n\nADVANCED\n\n\nOVERCASH ET AL.\n\n\n5 of 13\n\n\nAND PROCESSING\n\n\n0.9 MJ nre/kg PET, an astronomical 9800% improvement,\nan interesting theoretical limit.\n\n\nenergy benefits of replacing all or a portion of the second\nand third bottle cycles with recycled content.\n\n\n3.3 | Scenario 3: Mechanical recycling:\nReuse by grinding, extrusion, pelletizing,\nand injection molding to bottle\n\n\nScenario 2: Direct PET bottle reuse\nwith only transport or simple process to\n\n\n3.2 |\n\n\nnext user\n\n\nAn example of direct product reuse would be a PET wind-\nshield washing fluid reservoir container from a damaged\nautomobile removed and installed directly into another\nautomobile. Manufacture of one polymer product (89 MJ\nnre/kg PET bottles), then leads to essentially direct reuse of\nthe product in the same form. Transport of about 300 km\nis estimated at 0.9 MJ/kg PET bottles, using 0.003 MJ\nnre/kg km (Appendix 1). So two products are 89 +0.9 =\n90 MJ nre/2 kg PET bottles, Figure 3 and each bottle is now\nabout 45 MJ/kg PET bottles (1.8 kg CO2eq/kg PET bottles).\nProgressing to the third bottle gives us a third bottle for a\ncumulative energy of 91 MJ nre for three bottles or 30 MJ\nnre/bottle (1.3 kg CO2eq/kg PET bottles). This is probably\nthe lowest energy scenario because we are eliminating both\nthe category 1 and essentially all category 2a energy inputs\nwith recycling. At some point, usage or successive recycling\ndamages the PET polymer and the sequence of direct reuse\nmust stop, but it is unclear when that might be and so just\nat two bottle cycles, the energy (and carbon footprint) are a\n290% energy-related improvement. That is, landfilling is a\n290% greater impact on the environment based on energy\nuse alone. In the limit of very high numbers of recycling\nloops (such as 2000 cycles), a PET bottle could approach\n\n\nManufacture of one polymer product is 89 MJ nre/kg\nproduct, then mechanically grinding PET into flake\n(2.5 MJ nre/kg PET), extruding and pelletizing (2.2 MJ\nnre/kg PET and 0.18 MJ nre/kg, respectively), and injec-\ntion molding (10 MJ nre/kg PET) into the second bottle,\nFigure 4 (data in Appendix 1). Based on two bottle cycles\n(104 MJ nre/2 kg PET bottles), this is 52 MJ nre/kg PET\nbottle cycle (2.3 kg CO2eq/kg PET bottles). The category\n1 energy is eliminated and replaced by the grinding,\nextrusion, pelletizing, and injection molding. For the\nthird bottle cycle, we get 119 MJ nre/3 kg PET bottle, or\n36 MJ nre/kg PET bottle (1.9 kg CO2eq/kg PET bottles).\nAt the third cycle, mechanical recycling offers a 250%\nimprovement over all virgin energy inputs.\n\n\n3.4 | Scenario 4: Reuse by grinding,\ndepolymerizing to terephthalic acid and\nethylene glycol, repolymerizing, and\ninjection molding to bottle\n\n\nWhile there are different methods for depolymerizing\nPET, there is scant data about the energy use associated\n\n\nDirect reuse with transport or simple process to next user\n\n\n300\n285\n\n\nCumulative energy (nre) from natural resources in the\nearth to PET bottles 1, 2 and 3, MJ\n\n\n270\n\n\n255\n\n\n240\n\n\n225\n\n\n210\n\n\n195\n\n\nProduct A3,\n30 MJ/kg PET\nproduct\n\n\nProduct A1,\n89 MJ/kg PET\nproduct\n\n\n180\n\n\n165\n\n\n150\n\n\n135\n\n\n120\n\n\n105\n\n\n90\n\n\n75\n\n\n60\n\n\nProduct A2,\n45 MJ/kg PET\n\n\n45\n\n\n30\n\n\nproduct\n\n\n15\n\n\n0\n\n\n3\n\n\n5\n\n\n7\n\n\n8\n\n\n9\n\n\n0\n\n\n1\n\n\n2\n\n\n4\n\n\n6\n\n\nR1\n\n\nR1\n\n\nSteps of supply chain and recycling technology from natural resources in the\nearth through to PET bottles 1, 2, and 3\n\n\nFIGURE 3 Scenario\n\n\n2 direct reuse\n\n\nJOURNAL OF\n\n\n6 of 13 WILEY-\n\n\nADVANCED\n-MANUFACTURING\nAND PROCESSING\n\n\nOVERCASH ET AL.\n\n\nwith these processes. This assessment is based on the\nreduced global warming potential of the process publi-\nshed on the website of loop industries [10] reporting a 63%\nreduction. This would mean the production of PET pel-\nlets from the depolymerized/purified/repolymerized ter-\nephthalic acid (TPA) plus ethylene glycol is 37% of virgin\nPET. Manufacture of the first polymer product is again\n89 MJ nre/kg PET bottles. The recycle process begins\nwith grinding (2.5 MJ nre/kg) and depolymerizing the\n\n\nflake in a process that reacts the PET to the co-mono-\nmers, terephthalic acid, and ethylene glycol. These co-\nmonomers are then purified and repolymerized at some\nlocation back to PET as a pellet product. From the energy\nreduction as reported on the website, the Loop process is\n(89*0.37 = 33 MJ nre/kg PET bottles). Their PET product\nis then injection molded (10 MJ nre/kg PET bottles) to\ngive the second cycle of PET bottles, Figure 5. This gives\n135 MJ nre/2 kg PET bottles or 67 MJ nre/kg PET bottles\n\n\nReuse by grinding, pelletizing, and injection molding to bottle\n\n\n300\n\n\nCumulative energy (nre) from natural resources in the\n\n\n285\n\n\n270\n\n\n255\n\n\n240\n\n\nearth to PET bottles 1, 2 and 3, MJ\n\n\n225\n\n\nProduct A3,\n40 MJ/kg PET\nproduct\n\n\n210\n\n\nProduct A2,\n52 MJ/kg\nPET product\n\n\n195\n\n\nProduct A1,\n89 MJ/kg\nPET product\n\n\n180\n\n\n165\n\n\n150\n\n\n135\n\n\n120\n\n\n105\n\n\n90\n\n\n75\n\n\n60\n\n\n45\n\n\n30\n\n\n15\n\n\n0\n\n\n0 0123456 7 8 9 R1 R2 R3 R4 R1 R2 R3 R4\nSteps of supply chain and recycling technology from natural resources in the\nearth through to PET bottles 1, 2, and 3\n\n\nMechanical\n\n\nFIGURE 4\n\n\nrecycling\n\n\nReuse by grinding, depolymerizing to terephthalic acid and ethylene\nglycol, repolymerizing, and injection molding to bottle\n\n\n300\n\n\nCumulative energy (nre) from natural resources in the\n\n\n285\n270\n\n\nProduct A3,\n60 MJ/kg PET\nproduct\n\n\n255\n240\n\n\nearth to PET bottles 1, 2 and 3, MJ\n\n\n225\n210\n195\n\n\nProduct A1,\n89 MJ/kg PET\nproduct\n\n\n180\n165\n150\n135\n\n\n120\n\n\n105\n90\n\n\n75\n\n\nProduct A2,\n\n\n60\n\n\n67 MJ/kg PET\nproduct\n\n\n45\n30\n\n\n15\n\n\nReuse through\n\n\nFIGURE 5\n\n\n0\n\n\n8 9 R1 R2\n\n\n3\n\n\n0\n\n\n2\n\n\n5\n\n\n6\n\n\n7\n\n\nR3 R1 R2 R3\n\n\n1\n\n\n4\n\n\ndepolymerization and reforming\n\n\nSteps of supply chain and recycling technology from natural resources in the\nearth through to PET bottles 1, 2, and 3\n\n\nPET for bottle. PET,\n\n\npolyethylene terephthalate\n\n\nJOURNAL OF\n\n\n-MANUFACTURING- WILEY\n\n\nADVANCED\n\n\nOVERCASH ET AL.\n\n\n7 of 13\n\n\nAND PROCESSING\n\n\n(3.2 kg CO2eq/kg PET bottles). Continuing this circular-\nity, to the third bottle, we have a cumulative energy of\n180 MJ nre/3 kg PET bottles or 60 MJ nre/kg PET bottle\n(3.1 kg CO2eq/kg PET bottles). This technology in the\nchemical recycling continuum is an improvement\nof 150%.\n\n\nreaches a limit of just 70 MJ nre/kg PET bottles. With\nthe fuel value of crude oil and natural gas at 45 and\n53.5 MJ/kg, respectively, the use of PET as a fuel\n(22 MJ/kg PET) is much smaller because it is 39% oxy-\ngen, while crude oil is 1%-3% non-carbon and hydrogen,\nand natural gas is nearly zero. PET is just a less desir-\nable fuel energy source. Even polymers with higher fuel\nvalue like polyethylene or polypropylene, only offset\nmore of the category 1 energy leaving the category 2a\nenergy of the whole supply chain to make the polymer\nrequired for subsequent bottle cycles.\n\n\n3.5 | Scenario 5: Energy Recovery: Reuse\nby grinding and burning as a solid fuel,\nwith new PET bottles from virgin inputs\n\n\nAs before, manufacture of the first polymer product is\n89 MJ nre/kg PET bottles. This bottle is ground (2.5 MJ\nnre/kg PET) as waste polymer and used as a solid fuel to\nprovide heat, electricity or combined heat and power\n(CHP). The fuel value of PET is only about 22 MJ nre/kg\nPET bottle.[11]\nIn this case, the second product virgin supply\nchain category 1 energy is credited with the 22 MJ nre/kg\nPET bottles, regardless of where it is used because that is a\nfossil fuel \"credit.\u201d The cumulative energy for two PET bot-\ntle cycles is 89 MJ nre/kg PET bottles +2.5 MJ nre/kg\ngrinding plus (89-22) MJ nre/ kg PET bottles or 160 MJ\nnre/2 kg PET bottles or 80 MJ nre/kg PET bottles (2.8 kg\nCO2eq/kg PET bottles), Figure 6. Even if grinding can be\navoided this would shift this value to 89+ [89-22] =\n156 MJ nre/2 kg PET bottles or 78 MJ nre/kg PET bottles.\nContinuing this scenario to bottle three, we have 229 MJ\nnre/3 kg PET bottles or 76 MJ nre/kg PET bottles (3.1 kg\nCO2eq/kg PET bottles), a 120% improvement. Extending\nthis scenario to 100 cycles (10 000 PET bottles combusted),\n\n\n3.6 | Scenario 6: Varying recycled\ncontent of products\n\n\nUsing any recycling technology yields a material that can\nbe reincorporated into the same or similar product or\ninto different products, thus producing a product with a\npost-industrial or post-consumer recycled content. We\ndeveloped an equation that determines the environmen-\ntal benefit of producing successive product(s) while fac-\ntoring in the type of recycling technology, the polymer,\nand the amount of recycled content. It can also be used\nfor any other recycling technology, at any percent\nrecycled content, for any plastic or other material, and\nfor any life cycle metric, like energy, global warming\npotential, human health, and so forth.\nWith the environmental genome energy and other life\ncycle metric data on the primary product (product A) and\nthe incremental energy to produce the same product by\n\n\nReuse by grinding and burning as a fuel then reforming PET bottle\nwith conventional process\n\n\n300\n\n\nCumulative energy (nre) from natural resources in the\nearth to PET bottles 1, 2 and 3, MJ\n\n\n285\n\n\n270\n\n\n255\n240\n225\n210\n195\n\n\nProduct A1,\n89 MJ/kg PET\nproduct\n\n\n180\n\n\n165\n150\n135\n120\n105\n\n\nProduct A3,\n76 MJ/kg\nPET product\n\n\n90\n\n\nProduct A2,\n80 MJ/kg\nPET product\n\n\n75\n60\n\n\n45\n\n\n30\n\n\n15\n\n\n0\n\n\nR1 0 2\n\n\n0\n\n\nR1 0\nSteps of supply chain and recycling technology from natural resources in the\nearth through to PET bottles 1, 2, and 3\n\n\n2\n\n\n4\n\n\n6\n\n\n8\n\n\n2 4 6\n\n\n8\n\n\n4\n\n\n6 8\n\n\nFIGURE 6\n\n\nReuse as a\n\n\nsolid fuel\n\n\nJOURNAL OF\n\n\n8 of 13 WILEY-\n\n\nADVANCED\n-MANUFACTURING-\nAND PROCESSING\n\n\nOVERCASH ET AL.\n\n\nReuse of 50% of recovered PET bottle plus 50% from conventional\nprocess (50% recycled content PET bottle\n\n\nFIGURE 7 Reuse of 50% of\nrecovered PET bottle material\n\n\n(from mechanical process of\ngrinding, extrusion, pelletizing,\nand injection molding) plus 50%\nfrom conventional virgin PET\nprocess (ie, a 50% recycled\ncontent PET bottle). PET,\n\n\n300\n285\n270\n255\n240\n225\n210\n195\n\n\nCumulative energy (nre) from natural resources in the\n\n\nProduct A3,\n65 MJ/kg\nPET product\n\n\nearth to PET bottles 1, 2 and 3, MJ\n\n\npolyethylene terephthalate\n\n\n180\n\n\n165\n150\n\n\n135\n120\n105\n90\n75\n60\n\n\nProduct A2,\n\n\n71 MJ/kg PET\n\n\nproduct\n\n\n89 MJ/kg PET\nProduct A1,\n\n\n45\n\n\n30\n\n\nproduct\n\n\n15\n0\n\n\n0.5*R1\n\n\n0.5*4\n\n\n0.5*9\n\n\n0.5*6\n\n\n0.5*1\n\n\n0.5*3\n\n\n0.5*5\n\n\n0.5*7\n\n\n0.5*0\n\n\n0.5*2\n\n\n0.5*8\n\n\n2 4 6\n\n\n8\n\n\nSteps of supply chain and recycling technology from natural resources in the\nearth through to PET bottles 1, 2, and 3\n\n\nany recycling technology from waste materials we can\nnow determine the benefits of increasing recycle content\n(Equation 1).\n\n\ndownscaled or upscaled. The lifecycle metric of nre is used\nhere, but in this method, any life cycle impact assessment\nmetric can be calculated with Equation (1). Thus, with this\napproach of comparing technologies, it is neither difficult\nto add different products of PET nor to change the mate-\nrial being displaced by the recycled PET, hence it applies\nto open and closed loop systems.\nRecent publications of LCA-based analyses (1\nasserting that increasing recycled content does not neces-\nsarily confer environmental benefits have missed the\nimpact of category 1 energy. The process energy for incor-\nporating recycled content does not change since both vir-\ngin and secondary chemicals are processed under the\nsame conditions (ie, polymerization or injection mold-\ning). If recycled content has a positive environmental\nbenefit at any level, then a disadvantage of increased\nrecycled content would only occur if the product\nmanufacturing process somehow required substantially\ngreater energy as secondary sources are increased (virgin\nsources decrease). The authors know of no wide-spread\nincidence in which this is true.\n\n\nERC (MJ/kg product A) = PPE/n+ [PPE*(1 \u2212 f)\n+RTE* f]*[(n\u22121)/n]\n\n\n(1)\n\n\n[12]\n\n\nWhere ERC = energy of recycled content product\nafter the first virgin manufacturing, calculated as the\ncumulative energy needed for n products divided by n.\nPPE = virgin product energy, MJ/kg recycled product A.\nIn the Figures of this article, PPE is the energy segment\nillustrated with solid lines. RTE = recycling technology\nenergy to convert waste product into the recycled\nproduct A. In the Figures of this article, RTE is the\nenergy segment illustrated with dotted lines. f fraction\nof product A that is recycled content (0 to 1). n = number\nof successive products (n: = 1 is the virgin manufacturing\nof product A).\n[6]\nThis equation allows one to scale any recycling tech-\nnology and any polymer to explore the benefit of recycling\nand incorporating recycled material into products. It is\nmore detailed than Geyer, thus making it clear why\nrecycling some polymers is better, even if the recycling\ntechnologies employ the same amount of energy. It also\nincludes the effect of multiple loops. Note: by writing two\nequations for a \u201ctwo-product\u201d system and solving these\nsimultaneously, Equation (1) can be used for any open\nloop regardless of whether the recycled product is\n\n\n-\n\n\nAs an example, using mechanical recycling to pro-\nduce material to be added as recycled content, and a 50%\nrecycled content in successive PET bottles, we can see\nfrom Equation 1 the second bottle is 71.5 MJ nre/kg PET\nbottle. Figure 7 depicts the use of recycled PET content\nderived from the mechanical recycling technology,\nFigure 4. After the first product, the mechanical recycled\nPET is shown as 7.4 MJ nre/kg (the delta energy between\nproduct 1 and 2 in Figure 4, R1 to R4 times 50% to get MJ\n\n\nJOURNAL OF\n\n\n-MANUFACTURING- WILEY\n\n\nADVANCED\n\n\nOVERCASH ET AL.\n\n\n9 of 13\n\n\nAND PROCESSING\n\n\nEnergy of each product after (n-1) recycles, MJ/kg \u017a\n\n\n(A) 120\n\n\n120\n\n\n-A=direct reuse\n-B-mechanical\nC=depolymerization\n\n\nEnergy of each product after 2 recycles, MJ/kg PET\n\n\n-D=fuel\n\n\n100\n\n\n100\n\n\n- E=comparison\n\n\nD\n\n\nD\n\n\nE\n\n\n80\n60\n\n\n80\n\n\n0\n\n\nPET bottles\n\n\nbottles\n\n\n660\n\n\n\u0448\u043e\n\n\n60\n\n\nB\n\n\n40\n\n\n40\n\n\nA\n\n\n-A=direct reuse\n-B-mechanical\n\n\nB\n\n\n20\n\n\nC=depolymerization\n\n\n20\n\n\nD=fuel\n\n\n-- E-comparison\n\n\n0\n\n\nA\n\n\n0.2\n\n\n0.4\n\n\n0.6\n\n\n0.8\n\n\n1\n\n\n0\n\n\n0\n\n\nRecycle fraction\n\n\n20\n\n\n0\n\n\n5\n\n\n10\nn (number of uses)\n\n\n15\n\n\nFIGURE 8 Influence of number of recycling loops, A, and recycled content, B, on recycling technology energy for 1 kg PET bottles.\nPET, polyethylene terephthalate\n\n\nnre/0.5 kg PET). The other 50% of the PET is from the\nvirgin supply chain, also multiplied by 50% to get MJ\nnre/0.5 kg virgin PET. On the X-axis of Figure 7 the nota-\ntion is R1*0.5 and each of the nine stages of the virgin\nsupply chain are given as Z*0.5. These graphics are used\nto describe that this second bottles is half-recycled PET\nand half-virgin PET. As a result, the second bottle is an\naverage of 71 MJ nre/kg PET bottle (2.9 kg CO2eq/kg PET\nbottles). The third bottle is a cumulative energy of\n138 MJ nre/kg PET bottle which is an average of 65 MJ\nnre/kg PET bottle (2.8 kg CO2eq/kg PET bottles). If we\nincrease to 70% recycled content, the third bottle is an\naverage of 55 MJ nre/kg PET bottle (vs 65 MJ nre/kg PET\nbottle with 50% recycled content). These cases show that\nboth increasing the number of recycling loops and\nincreasing the recycled content give continued energy\nimprovement in making PET bottles.\nEquation (1) helps explore the environmental benefits\nof recycle content and the number of recycle loops,\nFigure 8. In Figure 8A we see that there is a consistent con-\ntinuum of energy benefit when each technology is used\nacross multiple recycle loops. However, if mechanical\nrecycling is used for one cycle (line E) and depolymeriza-\ntion is used for nine recycle lops, there is little difference in\nthe energy benefit of these two technologies. It is thus clear\nthat multiple loops are fundamentally better for energy\nimprovement, but the magnitude of improvement depends\non the recycling technology.\nRecycle technologies are located in relatively few and\ndispersed locations across the U.S., where the availability\nof waste PET may vary. This can lead to limits on the\nsupply of recycled content for one technology due to loca-\ntion vs another technology in a different location can\nhave higher availability of recycled materials. This\n\n\nvariation in recycled content can lead to localized shifts\nin the better environmental benefits when comparing\ntechnologies. In Figure 8B, mechanical recycling at 10%\nrecycled content has a higher energy/kg PET bottles than\ndepolymerization at 30% recycled content, despite the\nconsistent energy difference patterns of these two tech-\nnologies when compared at equal recycled contents\n(Figures 4 and 5). Therefore, it is critical to factor in the\nvarying product recycle content when evaluating the\nenergy benefit of recycling technologies.\nEquation (1) helps explore the environmental benefits\nof recycle content and the number of recycling loops,\nFigure 8. In Figure 8A we see that there is a consistent\ncontinuum of energy benefit when each technology is used\nacross multiple recycling loops. However, if mechanical\nrecycling is used for one cycle (line E) and depolymeriza-\ntion is used for nine recycling loops, there is little differ-\nence in the energy benefit of these two technologies. It is\nthus clear that multiple loops are fundamentally better for\nenergy improvement, but the magnitude of improvement\ndepends on the recycling technology.\nRecycling technologies are in relatively few and dis-\npersed locations across the U.S., where the availability of\nwaste PET may vary. This can lead to limits on the supply\nof recycled content for one technology due to location vs\nanother technology in a different location that can have\nhigher availability of recycled materials. This variation in\nrecycled content can affect the overall environmental bene-\nfits when comparing technologies. In Figure 8B, mechani-\ncal recycling at 10% recycled content has a higher energy/\nkg PET bottles than depolymerization at 30% recycled con-\ntent, despite the consistent energy difference patterns of\nthese two technologies when comparing two materials with\nequivalent amounts of recycled content (Figures 4 and 5). It\n\n\nJOURNAL OF\n\n\n10 of 13 WILEY-\n\n\nADVANCED\n-MANUFACTURING\nAND PROCESSING\n\n\nOVERCASH ET AL.\n\n\nfollowing this method. This is done by writing two equa-\ntions for a two product systems and solving these simulta-\nneously. Equation 1 can thus be used for any open loop\nregardless of whether the recycled product is downscaled or\nupscaled. Thus, this equation applies to and compares open\nand closed loops. However, to get a more basic comparison\nof these secondary technology issues the approach used\nmust also be transparent.\n\n\nis critical to factor in variations of recycled content of prod-\nucts (the varying product recycle content) when evaluating\nthe energy benefit of recycling technologies.\n\n\n4 | METHODS\n\n\nFive recycling scenarios representing the common catego-\nries of plastics recycling are examined. These are based on\nPET polymer in the form of a beverage bottle as the prod-\nuct. However, PET for the window washer fluid container\nin an automobile is also used as an example for one sce-\nnario as it is easier to illustrate this first recycling concept.\nThe functional unit is 1 kg of PET bottle product (~100\nsingle-serve 0.5 L water bottles). For all five scenarios, the\nfirst product is made from virgin inputs and the energy for\nthe first bottle is the same across all scenarios. The next bot-\ntle begins with one of the recycling technologies and con-\ntinues until a second bottle is made. The cumulative energy\nat the second bottle stage is the sum of the first and second\nbottles, which is two bottles for X total MJ, which then is\nX/2 per bottle. To learn from the progression involving the\nenergy in categories 1 and 2, the third product is made with\nthe same scenario. The cumulative energy at the end of the\nthird bottle is Y total MJ, which then is Y/3 per bottle. Fur-\nther, the energy used is given as the nre, which accounts\nfor the full energy needed in all the manufacturing and\nrecycling processes as described in Table 1. Most of the data\nare from the environmental genome initiative (www.\nenvironmentalgenome.org) and Environmental Clarity, Inc.\ndatabase [9] and described in Appendix 1. It is neither diffi-\ncult to add different products from the recycled PET nor to\nchange the material being displaced by the recycled PET by\n\n\n5 | DISCUSSION OF FUTURE\nANALYSIS GOALS\n\n\nBased on this preliminary study, improved explanation\nand energy use information of these five scenarios and\nvariants thereof will strengthen the credibility of\nrecycling decision-making. At present, all of these alter-\nnatives are better than landfilling and should be accepted\nby society as improvements. It must be emphasized that\nthese results attempt to quantify the environmental bene-\nfits of recycling on cumulative energy use and do not\nattempt to address other critically important aspects such\nas toxicity or economic impacts.\nThe supply of waste material varies across the coun-\ntry, so in one location more of plastics A may be available\nthan plastics B, thus leading to potentially different\namounts of recycled content of end products. From Equa-\ntion (1), this can change the relative benefit of one\nrecycling technology (with higher recycled content) com-\npared to another technology (with lower recycled con-\ntent). Thus, waste-limited cases cannot be easily\ncompared as this factor overrides the actual technology\nenergy improvement performance.\n\n\nTABLE 2\n\n\nComparative summary of MJ nre energy per kg PET bottles (about 100 single-serve 0.5 Liter water bottles) for various\n\n\nrecycling technologies\n\n\nPET bottle 1\n\n\nScenario\n\n\nPET bottle 2\n\n\nPET bottle 3\n\n\nDirect PET bottle reuse with only transport or simple process\nto next user\n\n\n89\n\n\n45\n\n\n30\n\n\nReuse by grinding, extruding, pelletizing, and injection\nmolding to bottle\n\n\n52\n\n\n36\n\n\n89\n\n\nReuse by grinding, depolymerizing to terephthalic acid and\nethylene glycol, repolymerizing to pellets, and injection\nmolding to bottle\n\n\n89\n\n\n60\n\n\n67\n\n\n50% recycle of PET from grinding, extruding, pelletizing, and\ninjection molding to bottle and 50% virgin PET\n\n\n65\n\n\n89\n\n\n71\n\n\nReuse by grinding and burning as a fuel then reforming PET\nbottle with the conventional virgin process\n\n\n89\n\n\n80\n\n\n76\n\n\nNon-recycle PET product with bottle to landfill and virgin\nsupply chain for successive bottles\n\n\n89\n\n\n89\n\n\n89\n\n\nAbbreviation: PET, polyethylene terephthalate.\n\n\nJOURNAL OF\n\n\nWILEY\n\n\nADVANCED\n\n\nOVERCASH ET AL.\n\n\n11 of 13\n\n\n-MANUFACTURING-\n\n\nAND PROCESSING\n\n\nrecycling technologies. The environmental genome (www.\nenvironmentalgenome.org) can facilitate these calculations\nas it contains the polymers and whole supply chains for\nsuch evaluations.\n\n\nAnother vital but currently missing body of informa-\ntion is the number of recycling loops a plastic can\nundergo before some factor forces it into lower material\nquality uses where the performance requirements are less\nstringent, or it is converted into recovered energy or liq-\nuid fuel sources. These cycle limits can further shift the\nenvironmental benefit of any technology as Table 2\nshows the energy profile improves with higher number of\nloops. This article illustrates loops one and two only. This\nis a degradation factor that could be studied by just\nlooping immediately to discover this limiting factor value.\nFor example, mechanical recycling may degrade the PET\npolymer and thus have fewer cycle loops as a bottle,\nwhile depolymerization to monomers and purification\nmay extend the life of the polymer substantially. If this\ndegradation limit to the number of recycling loops for\nPET recycling for these and other technologies can be\nestimated, then analysis herein can be modified by\ninserting a virgin product at this recycle limit and then\ncontinuing as shown in the current Figures. Additionally,\nif some recycling technologies do not need cleaned poly-\nmers, the benefit of reduced energy in the collection and\nprocessing stages can be added to those technologies.\n\n\nACKNOWLEDGMENTS\n\n\nElizabeth Ritch of GreenBlue has selflessly helped us\nwith the graphing challenges in this article and as an\nexternal evaluator.\n\n\nCONFLICT OF INTEREST\n\n\nThis engineering-based analysis was undertaken with no\nconflict of interests or direct funding support.\n\n\nORCID\n\n\nMichael R. Overcash https://orcid.org/0000-0002-6291-\n\n\n6159\n\n\nREFERENCES\n\n\n[1] Franklin Associates, Life cycle inventory of 100% postconsumer\nHDPE and PET recycled resin from post-consumer containers\nand packaging. Washington, DC: Plastics Div, American\nChemistry Council, p. 73, 2010\n[2] X. Xiao, W. Zmierczak, J. Shabtai, Energy Fuel 2010, 11(1), 4.\n[3] Leblanc, R., An overview of polypropylene recycling, The balance\nsmall business, https://www.thebalancesmb.com/an-overview-of-\npolypropylene-recycling-2877863 (accessed: September 2018).\n[4] Van Dusan, L and P. Van Dusan, Nylon 6 and 6,6, https://\noecotextiles.wordpress.com/2012/06/05/nylon-6-and-nylon-66/,\n(accessed: December 2018), 2012.\n[5] The Association of Plastics Recyclers, HDPE, http://www.\nplasticsrecycling.org/hdpe, (accessed: December 2018), 2017.\n[6] R. Geyer, B. Kuczenski, T. Zink, A. Henderson, J. Ind. Ecol.\n2015, 20(5), 1010.\n[7] M. Overcash, Green Chem. 2016, 18, 3600. https://doi.org/10.\n1039/c6gc00182c.\n[8] C. Ponder, B. Gregory, E. Griffing, Y. Li, M. Overcash, J Adv\nManuf Process 2019, 1, e10012. https://doi.org/10.1002/amp2.\n10012.\n\n\n6 | CONCLUSIONS\n\n\nTable 2 captures the continuum of recycling or circularity\nbenefits for these representative scenarios for PET and\nshows consistently that more recycling loops provide lower\nproduct impact. The data are assembled from the greatest\nbenefit expressed in energy per kg PET bottles (about\n100 single-serve 0.5 L water bottles), listed first, and the\nbase case of landfill listed last. From Appendix 1, there are\nranges in values of the process energies on the order of\n1-3 MJ nre/kg PET. This variation does not significantly\nchange the relative energy use for these technologies. All\ncases, except solid fuel, use injection molding of the bottle\nand these larger variations in Appendix 1 are thus the\nsame for all technologies examined herein and likewise do\nnot alter the comparisons. Thus, one can conclude all the\ncircularity scenarios lead to an improvement over land-\nfilling or littering of PET bottles. For some technologies\nlike direct reuse, the energy improves significantly with\nthe third and higher product recycling loops, while others\nsuch as solid fuel use show smaller incremental improve-\nment. This successive improvement demonstrates the long\nterm goal of these reuse or recycling scenarios. The data\nshown here describe the kind of analysis and level of\ntransparency that are required for each polymer or chemi-\ncal considered for recycling in order to quantitatively\ndefine the energy reduction benefit, and hence to form a\nscience-based comparative methodology for more plastics\n\n\n[9] Environmental Genome Initiative and Environmental Clarity, Inc.,\nwww.environmentalgenome.org and www.environmentalclarity.\ncom, respectively, Raleigh, NC and Reston, VA, respectively.\n[10] Loop Industries, https://www.loopindustries.com/en/sustain,\n(accessed: December 2018).\n\n\n[11] D. Tsiamis, M. Castaldi, Determining accurate heating values of\nnon-recycled plastics, Earth Engineering Center, City College,\nCity University of New York, New York, NY 2016, p. 27.\n[12] Leif, D., Study: Recyclable packaging not always greenest\noption, Plastics Recycling Update, https://resource-recycling.\ncom/recycling/2018/12/11/study-recyclable-packaging-not-always-\ngreenest-option/ (accessed: January 2018).\n[13] Polyretec, polymer granulating and recycling equipment,\nhttps://www.polytecpm.com/products/plastic-granulating-\nrecycling/, (accessed: July 2019).\n[14] Paradise Distribution and Recycling, 11-2020 Ellesmere Road,\nToronto, ON, http://www.paradiserecycling.com/, personal\ncommunication, 2019.\n\n\nJOURNAL OF\n\n\n12 of 13 WILEY-\n\n\nADVANCED\n-MANUFACTURING\nAND PROCESSING\n\n\nOVERCASH ET AL.\n\n\n[15] DuPont Teijin Films, Life cycle report on recycled PET films,\nEnvironmental Clarity, Inc. Reston VA, Information available\nupon request, 2011.\n[16] Belzer, D., A Comprehensive System of Energy Intensity Indi-\ncators for the U.S.: Methods, Data and Key Trends, Pacific\nNorthwest National Laboratory, PNNL-22267, Richland,\nWashington 99352, 203, 2014.\n\n\nSymposium on Electronics and the Environment, DOI:\nhttps://doi.org/10.1109/ISEE.2006.1650060, 2006.\n\n\n[20] Pelletizer Heuristic, Environmental Clarity, Inc, Reston, VA,\navailable on request, 5 p. 2017.\n[21] Extruder Heuristic, Environmental Clarity, Inc., Reston, VA,\navailable on request, 10 p. 2010.\n\n\n[17] Ecoinvent database, https://www.ecoinvent.org/, (accessed:\nJuly 2019).\n[18] Wang, M., The Greenhouse Gases, Regulated Emissions,\nand Energy Use in Transportation (GREET) Model Version\n1.5, U.S. Department of Energy's Office of Energy Effi-\nciency and Renewable Energy (http://greet.es.anl.gov/),\n\n\nHow to cite this article: Overcash MR,\nEwell JH, Griffing EM. Life cycle energy\ncomparison of different polymer recycling\nprocesses. J Adv Manuf Process. 2020;2:e10034.\nhttps://doi.org/10.1002/amp2.10034\n\n\n1999.\n\n\n[19] Thiriez, A. and T. Gutowski, An Environmental Analysis of\nInjection Molding, Proceedings of the 2006 IEEE International\n\n\nAPPENDIX 1\n\n\nReferences and data used in plastics recycling technology comparisons\n\n\nPolyethylene\n\n\nterephthalate\n\n\nValue judged as\nrepresentative\n\n\n(PET) recycling\n\n\ntechnologies\n\n\nSource\n\n\nValue\n\n\n790 kg crude oil*45 MJ/kg crude oil\n+30.2 kg ng\n\n\nCategory 1, fossil natural\nresources in PET\n\n\nVirgin supply chain energy\n\n\n37.2 MJ/kg PET\n\n\n*53.5 MJ/kg ng = 37 160/1000 kg\n\n\nCategory 2, supply chain energy in\nPET, see this article Figure 1\nTotal supply chain\nPolyretec [13]\n\n\n52.2 MJ/kg PET\n\n\n89.4 MJ/kg PET\n2.5 MJ nre/kg PET\n\n\n5.2 MJnre/kg\n\n\nGranulating or Grinding\n\n\nof Plastics\n\n\nPolyretec [13]\n\n\n2.5 MJnre/kg\n\n\n2.0 MJnre/kg\n\n\nParadise distribution and\n\n\nrecycling[14]\n\n\nDupont Teijin study[15]\nBelzer, US DOE 2014 single\ncompartment trucks [16]\nBelzer, US DOE 2014 dual\ncompartment trucks [16]\n\n\n1.7 MJnre/kg\n\n\n0.003 MJ nre/kg km\n\n\n0.0025 MJ nre/kg km\n\n\nTransport\n\n\n0.0055 MJ nre/kg km\n\n\nEcoinvent operation, roads,\n\n\n0.0018 MJ nre/kg km\n\n\nmaintenance, manufacture [17]\n\n\nWang, 1999 p. A-72 [18]\n\n\n0.0055 MJ nre/kg km\n\n\nThiriez, 2006[19]\n\n\n0.0019 MJ nre/kg km\n0.19 MJ nre/kg\n0.12-0.19 MJ nre/kg\n0.16 MJ nre/kg\n1.7 MJ nre/kg PET\n\n\nDupont Teijin study[15]\nPelletizer heuristic [20]\nThiriez, 2006[19]\n\n\nPelletizing\n\n\n0.18 MJ nre/kg PET\n\n\nDupont Teijin study [15]\nThiriez, 2006[19]\n\n\nExtrusion\n\n\n2.2 MJ nre/kg PET\n\n\n1.8-5 MJ nre/kg\n\n\n(Continues)\n\n\nJOURNAL OF\n\n\nWILEY\n\n\nADVANCED\n\n\nOVERCASH ET AL.\n\n\n13 of 13\n\n\n-MANUFACTURING-\nAND PROCESSING\n\n\nPolyethylene\n\n\nterephthalate\n(PET) recycling\ntechnologies\n\n\nValue judged as\nrepresentative\n\n\nSource\n\n\nValue\n\n\nExtruder heuristic [21]\n\n\n2.7 MJ nre/kg PET\n\n\n2.0 MJ/kg\n\n\nParadise distribution and\n\n\nrecycling [14]\nThiriez, 2006[19]\n\n\nInjection Molding\n\n\n3-8 MJ nre/kg\n\n\n10 MJ nre/kg PET\n\n\n4-70 MJ nre/kg\n\n\n2-15 MJ nre/kg\n\n\nEcoinvent [17]\n\n\n21 MJ nre/kg\n\n\nNote: Significance of bold is for numbers judged as representative where there are multiple values cited.\n"}, "expected_output": {"claims": [{"unit": "MJ nre/kg PET", "value": 2.38, "evidence": ["Manufacture of one polymer product is 89 MJ nre/kg\nproduct, then mechanically grinding PET into flake\n(2.5 MJ nre/kg PET), extruding and pelletizing (2.2 MJ\nnre/kg PET and 0.18 MJ nre/kg, respectively), and injec-\ntion molding (10 MJ nre/kg PET) into the second bottle,\nFigure 4 (data in Appendix 1). Based on two bottle cycles\n(104 MJ nre/2 kg PET bottles), this is 52 MJ nre/kg PET\nbottle cycle (2.3 kg CO2eq/kg PET bottles). The category\n1 energy is eliminated and replaced by the grinding,\nextrusion, pelletizing, and injection molding. For the\nthird bottle cycle, we get 119 MJ nre/3 kg PET bottle, or\n36 MJ nre/kg PET bottle (1.9 kg CO2eq/kg PET bottles).\nAt the third cycle, mechanical recycling offers a 250%\nimprovement over all virgin energy inputs.\n"]}]}, "metadata": {"product_category": "Paper and plastic products", "request_id": "req_009939bc270f9ee7"}}
{"id": "ac9ae3e27097bbd35d834efd", "input": {"query": "What are the energy consumption values in BTU/lb or MJ/kg for gas-fired reverberatory furnaces melting aluminum?", "source_url": "https://www.theschaefergroup.com/docs/Schaefer_SS7_EN.pdf", "document_text": "SIMPLE SOLUTIONS\nTHAT WORK!\n\n\nFURNACE FACTS, ROI'S\n& ENERGY USE NUMBERS\n\n\n2) It pays to not go cheap on\nthe hot face furnace linings.\nModern central melt furnaces\nhave 80-to-90% alumina\nnon-wetting hot face linings.\nThey are easily cleaned (build-\nup is easily removed), rugged\nand will not penetrate at the\nall important belly-band area\n(molten metal contact area).\na) Premium hot face linings\npay. We recommend higher\nalumina products containing\n\n\nThe\n\n\nSchaefer Group, Inc\n\n\nDAVID WHITE\n\n\nNational Sales Manager\nTHE SCHAEFER GROUP\n\n\nARTICLE TAKEAWAYS:\n\n\n1. Super insulate your furnace linings to reduce energy costs\n\n\na phos bonding agent. If you\ndo use the cheaper hot\nface linings, a product like\n70-to-85% alumina phos-\nbonded plastic refractory\nwill hold up better in a\nmelter than the same\nalumina content non-wetting\nlow cement castables.\n\n\n2. Why Sow pre-heat hearths are a wise investment\n\n\n3. Understanding \"hard energy\" use numbers for Gas-fired, Electric Radient-roof,\nCrucible and Tower Melters\n\n\nIn this article we will give you\nsome basic facts about melting\nand holding aluminum in\neveryday furnaces as well as a\nranking of ROI on improvements\nyou can make to those furnaces\nto increase efficiency and energy\nusage is a number of different\ntypes of furnaces.\n\n\n1. Buying the best furnace designs and\nmost cost effective materials.\n\n\na. Central melt furnaces are large\n- it is difficult to clean furnaces\nmanually that are larger than\n50-to-60,000 pounds hold\ncapacity. Mechanized cleaning\n(fork truck and hoe) does the\nbest job on larger high\nheadroom furnaces. Most\nlarge furnaces have single end\nclean-out doors that are\nnarrower than the interior\nfurnace width. This makes for\nhidden, right-angled corners\nthat are difficult to clean.\nUnacceptable oxide build-up\nleads to premature relines and\nimpaired efficiencies.\n\n\n2. Spend the money to super insulate\nthe furnace linings. New products,\nsuch as micro porous silica insulating\nmaterials will save a huge amount of\n\"fixed heat loss\" energy. If the\nlining is properly engineered, the all-\nimportant \"freeze plane\" will still\noccur in non-wetting lining materials.\nThis is a case of your being able to\n\"have your cake and eat it too.\"\nThese super insulating products\nnormally add about $18.00/sq. ft.\nto the cost of the furnace lining,\nbut they normally provide about a\n16-to-20 month ROI.\n\n\nRanking of ROI Expenditures\nThe 'Ranking of ROI'\nexpenditures for aluminum\nfurnaces, in other words, how to\n\n\n3. Sow pre-heat hearths are a wise\n\n\ninvestment. If 50% of the aluminum\nyou melt is new metal (typical of a\nfoundry with a 50/50 yield), and\n50% is scrap and returns melted in\nthe charge well, the metal pre-\nheated for about 30 minutes on the\nhearth and then pushed into the\nbath will save 12-15% of the normal\nenergy required to melt the metal if\nit were all cold charged into the bath.\n\n\nget the biggest bang for your\nbucks from quickest-to-slowest\ninvestment recovery.\n\n\n1) The solution is having better\naccess to the furnace interior\nwith full width double-end\ndoors. The floors should have\ngentle transition slopes from\ndoor opening hearths-to-the\nflat portion of the floor (no\ngreater than 35\") so that the\nfurnace can be easily cleaned\nfor \"sludge\" on the floor.\n\n\n6\n\n\nFoundry/Die Caster Engineers TOOLBOX\n\n\nb. Transfer pumps are also a good investment as they transfer metal to the\nladle much faster and are safer for the metal handlers. The new overflow\npumps available are very efficient and provide a less turbulent transfer into the\ntransfer ladle see movie clip below!\n\n\nWATCH THE VIDEO\n\n\n5. Pre-heated combustion air through a regenerative combustion system, added\nto the features mentioned above, will drive the energy consumption down to\n900-to-1050 BTU/pound of aluminum melted in a fully utilized melter\nBecause of the efficiencies of the first four items above, the added cost of\nthe regenerative combustion system takes about 8,400 hours of full\ncapacity operation per year of $3.00/MCF natural gas to yield a 60-month\nROI. Escalating energy costs can shorten the ROI drastically. These burners\nwork in pairs and as one burner is firing the other is exhausting the products\nof combustion into the bed of tabular aluminum balls which heat up to the\nexhaust temperature then the burners switch and the combustion air is drawn\nthrough that heated media to preheat the combustion air significantly.\n\n\nSEALECA\n\n\na. This method of preheating and\ncharging normally provides about\na 20-to-24 month ROI, based on\n5,200 hours of melting per year.\n4. Circulation of the molten metal\nwithin the furnace bath (from\nthe charge well-to-the-thermal\nhead chamber-and-back) has the\nadvantage of saving another\n9-to-12% of the energy that it\ntakes to melt the aluminum, reduces\nmelt loss through enhancing more\nrapid melting and reduces sludging\nby convectively maintaining a\nhomogenous bath temperature. In\nrecent years great strides have been\nmade in improving molten metal\npump efficiency and drastically\nreducing their need for maintenance.\n\n\nREGENERATIVE BURNERS \"CYCLE A\u275e\n\n\nGas\n\n\nHot POC\n\n\nMea\n\n\nMedia\n\n\nCold Air\n\n\nWarm POC\n\n\nREGENERATIVE BURNERS \"CYCLE B\"\n\n\na. Typically, circulation pumps,\nand the wells into which they\nare designed, have a 24-to-28\nmonth ROI.\n\n\nMedia\n\n\nMedia\n\n\nNOTE: If the first four items above are\nfurnished on a furnace, a fully utilized\ncentral melter will melt at about 1235\nBTU/pound in a SGI radiant roof-fired\nreverb furnace, and at about 1,590 BTU/\npound in a SGI high headroom wall-fired\nfurnace. This is all being accomplished\nin a wet-bath reverb furnace, which\nabsolutely provides the aluminum\nfoundry the lowest metal melt loss by\nseveral percentage points.\n\n\nContinued on page 8\n\n\n7\n\n\nSIMPLE SOLUTIONS\nTHAT WORK!\n\n\n6. Recuperators for combustion air\npre-heat offer the faster return\non your investment for pre-heating\ncombustion air. They come in various\nsizes and are easily retrofitted in any\nsize furnace for you to start saving\nenergy instantly. The BTU's required\nto heat the combustion air up to\n700\u00b0 F are saved immediately upon\ninstalling this heat exchanger.\nCustomers are realizing a 19-25%\nreduction in fuel usage with these\nheat exchangers. At today's gas\nprices ROI's are averaging\n20 months.\n\n\nHARD USE ENERGY NUMBERS\n\n\nElectric: Connect .31KW/# of metal\nmelted it uses about .25-.27KW/# of\nmetal melted about 48% efficient.\nD. Tower or Stack Melters:\n\n\nLet's talk some \"hard energy\" use\nnumbers.\n\n\nA.Gas-fired\n\n\nGenerally connect about\n1800BTU's/# of metal melted and\nuse about 1,000BTU's/# of metal\nmelted when the stack is kept full,\nwhich puts them (depending on their\nfixed heat loss) into the 74%\nefficiency range.\n\n\n1. A well designed and fully utilized\nradiant roof-fired melting furnace\nwill melt for 1,500-1600 BTU/lb.\n(34% effic.): 100% cold metal\ncharging.\n\n\n2. With the addition of the \"easy\"\nenergy recovery enhancements\nof the lining super insulation\npackage, sow pre-heat hearth\nand molten metal circulation =\n1,235BTU/lb. (41% effic.).\n\n\nENERGY VALUES OF THE MOST\nCOMMONLY USED ENERGY SOURCES:\n\n\n\u2022 Natural gas 1,050 BTU/CF Some\ncountries are less, some are more!\n\n\n\u2022 100,000 BTU/therm\n\n\n3. The more expensive energy\nenhancements start with:\n\n\n1,000,000 BTU/decatherm,\n\n\n\u2022\n\n\nor 1,000 CF\n\n\na. Recuperation, in conjunction\nwith 1&2 above, = 1,095\nBTU/lb. (50% effic.).\n\n\nElectric-3,412 BTU/KWH\n\n\n\u2022\n\n\n\u26ab #2 Fuel Oil - 138,000 BTU/U.S. gallon\n\n\nEnergy Diagram Recuperative Burner\n\n\nPropane - 92,000 BTU/\n\n\nFURNACE\n\n\n\u2022\n\n\nU.S. gallon in liquid\n\n\nSCHAEFER\nFURNACES\n\n\nSCHAEFER\nFURNACES\n\n\nBOTTOM LINE!\n\n\nMY FLUE GAS\nTEMPERATURE\n\n\nThe information in this article is\nmeant to provide you with ways\nof saving energy which even at\ntoday's prices is still one of your\nmost expensive costs to operate\na foundry or die cast facility.\n\n\nb. Regenerative burners, in\nconjunction with 1&2 above, =\n940 BTU/lb. (72% effic.).\nEnergy Diagram Regenerative Burner\n\n\n7. Well Covers should be placed on\nany open well that will be out of\nproduction for more than 30\nminutes. At higher temperatures\n1400\u00b0 F you lose approximately\n7800 BTU's/square foot/hr of\nsurface area off an open well with\nsome dross on the surface. Since\nthe average charge well on a large\nreverb is about 30 square feet, that\nis 234,000 BTU's/hr off that well.\n\n\nKnow what your present melting\nand holding energy uses are\nnow and meter them. A Peter\nDrucker's truism \"If you can't\nmeasure it, you can't manage it.\"\nThis is just a relevant today as it\nwas years ago.\n\n\nPURNACE\n\n\nTOTAL\n\n\nJISARTY - FULL PUT\n\n\nWF FLUE GAS\nTEMPERATURE\n\n\nOf course none of these items are free\nbut the cost vs return on investment\nmake most of these worthwhile\ninvestments for your furnaces.\n\n\nB.Electric Radiant-roof\n\n\n23-.24 kwh/# electric reverb\n784 btu/lb. 66.7%\n\n\n1. with molten metal circulation\n21-22 kwh/# 687btu/lb. 72.8%\n\n\n2. immersion element melter 18-19\nkwh/# with molten metal\ncirculation 655 btu/lb. 76.3%\n\n\nC. Crucible furnaces:\n\n\nGas: Connect 3,000BTU\"s/# of metal\nmelted uses about 2,300BTU's/#\nMelted 32% efficient\n\n\nContact:\n\n\nDAVID WHITE\n\n\ndavid.white@theschaefergroup.com\n\n\n8\n"}, "expected_output": {"claims": [{"unit": "BTU/lb", "value": 940, "evidence": ["b. Regenerative burners, in conjunction with 1&2 above, = 940 BTU/lb. (72% effic.).", "5. Pre-heated combustion air through a regenerative combustion system, added to the features mentioned above, will drive the energy consumption down to 900-to-1050 BTU/pound of aluminum melted in a fully utilized melter", "Regenerative burners, in conjunction with 1&2 above, = 940 BTU/lb. (72% effic.)."]}, {"unit": "BTU/lb", "value": 1095, "evidence": ["a. Recuperation, in conjunction with 1&2 above, = 1,095 BTU/lb. (50% effic.).", "Recuperation, in conjunction with 1&2 above, = 1,095 BTU/lb. (50% effic.)."]}, {"unit": "BTU/lb", "value": 1235, "evidence": ["2. With the addition of the \"easy\" energy recovery enhancements of the lining super insulation package, sow pre-heat hearth and molten metal circulation = 1,235BTU/lb. (41% effic.).", "NOTE: If the first four items above are furnished on a furnace, a fully utilized central melter will melt at about 1235 BTU/pound in a SGI radiant roof-fired reverb furnace", "With the addition of the \"easy\" energy recovery enhancements of the lining super insulation package, sow pre-heat hearth and molten metal circulation = 1,235BTU/lb. (41% effic.).", "If the first four items above are furnished on a furnace, a fully utilized central melter will melt at about 1235 BTU/pound in a SGI radiant roof-fired reverb furnace"]}, {"unit": "BTU/lb", "value": 1500, "evidence": ["1. A well designed and fully utilized radiant roof-fired melting furnace will melt for 1,500-1600 BTU/lb. (34% effic.): 100% cold metal charging.", "A well designed and fully utilized radiant roof-fired melting furnace will melt for 1,500-1600 BTU/lb. (34% effic.): 100% cold metal charging."]}, {"unit": "BTU/lb", "value": 1600, "evidence": ["1. A well designed and fully utilized radiant roof-fired melting furnace will melt for 1,500-1600 BTU/lb. (34% effic.): 100% cold metal charging.", "A well designed and fully utilized radiant roof-fired melting furnace will melt for 1,500-1600 BTU/lb. (34% effic.): 100% cold metal charging."]}]}, "metadata": {"product_category": "Metal, mineral, plastic & glass products", "request_id": "req_af09758b7180f107"}}
{"id": "679029e631fd1d3f033019c0", "input": {"query": "What is the energy consumption for scrap pretreatment operations (decoating, cleaning, drying) in secondary aluminum production? Provide values in Btu/lb or MJ/kg for thermal/natural gas energy.", "source_url": "https://www.aceee.org/files/proceedings/2005/data/papers/SS05_Panel01_Paper17.pdf", "document_text": "E2 and P2 Improvement Opportunities in Secondary Aluminum Processing:\nA Case Study\n\n\nMatthew D. Swanson, Robert A. Miller, Jonathan J. Aardsma, and Michael J. Chimack\nUniversity of Illinois at Chicago, Energy Resources Center\n\n\nABSTRACT\n\n\nThis paper will present the results of an energy assessment at a secondary aluminum plant\nin northern Illinois, specifically, the energy and non-energy benefits of using a new combustion\ntechnology for scrap decoating. This new technology, a variable floatation decoater (VFD), is\nbeing developed by the U.S. Department of Energy and a private engineering firm to replace the\nstandard rotary kilns currently used for scrap decoating.\nRotary kilns have traditionally been employed in the decoating process. However, a\nVFD removes organics of any type from scrap aluminum and steel using less energy. High\nvelocity gases mechanically strip off liquid organics in a low oxygen environment. The organic\nladen gases are then sent to an afterburner to vaporize the organics. The VFD meets the EPA's\nnew Clean Air Act standards for dryers, decoaters and delacquering kilns. Based on lab testing\nand empirical data, VFDs can reduce the energy used by the decoating process by 50% while\nreducing the decoating cycle by 80%. [AA2] In addition, VFDs produce the required quality of\naluminum turnings in a single pass, which eliminates the need for repeating the process, thereby\nincreasing energy savings potential.\n\n\nIntroduction\n\n\nThe Aluminum Industry\n\n\nThe United States' aluminum industry is the world's largest, generating sales of\napproximately $39.1 billion in products and exports. [AA1] The U.S. aluminum industry\noperates over 300 plants in 35 states, produces more than 23 billion pounds of metal annually\nand employs over 145,000 people with an annual payroll of about $5 billion. [AA1] The\naluminum industry is divided into two primary segments: primary aluminum processing and\nsecondary aluminum processing. Primary aluminum processing consists of domestic production\nof aluminum from ore material. In 1998 primary aluminum processing accounted for 63 percent\nof the U.S. production (15.5 billion pounds). [AA2] Although the primary aluminum sector of\nthe industry accounts for the majority of the total aluminum supply in the U.S., a vital\ncontributor is the secondary aluminum industry. The secondary aluminum industry uses\naluminum scrap as its principal feedstock. This scrap is metal recovered during industrial\nmachining and fabrication and consists of machine turnings, borings, clips, and skeletons from\nstamping plants. The scrap is re-melted and reformed. In 1998, the secondary aluminum\nindustry processed 4.3 million metric tons (9.4 billion pounds) of scrap and manufactured 3.4\nmillion metric tons (7.5 billion pounds) of aluminum, approximately 37 percent of the total U.S.\naluminum supply of 104 million metric tons (23 billion pounds). [AA2]\n\n\n\u00a9) 2005 ACEEE Summer Study on Energy Efficiency in Industry\n\n\n1-179\n\n\nThe contribution of the secondary aluminum industry, a third of total U.S. aluminum\nsupply, becomes even more significant when placed in historical perspective. In 1972, recycled\naluminum accounted for 19 percent of the nation's total aluminum supply. [AA2] Over the next\n27 years, production of recycled aluminum rose 242 percent to 3.4 million metric tons (7.5\nbillion pounds). Over the same period, the total U.S. aluminum supply increased just 91 percent.\n[AA2] These trends demonstrate that there have been concentrated efforts to increase the\namount of aluminum recovered. Furthermore, the secondary aluminum industry is an essential\npart of the U.S. focus on energy conservation, as recycling aluminum requires only five percent\nof the energy that primary-ore production requires. [AA2] This significant difference in energy\nusage shows the true impact secondary aluminum processing can have on the industry's energy\nconsumption as a whole. A transfer of one unit of primary production to secondary production\nwill result in an energy reduction of 95 percent. As more and more aluminum is processed from\nscrap as opposed to ore, the total energy consumed by the aluminum industry will fall.\nWhile the United States has long consumed the majority of the world's resources, the\nmarket for aluminum and other metals in China continues to experience growth as the country\ncontinues its rapid industrial expansion. China's economic growth reached a seven-year high at\n9.1 percent in 2003. [CD] China's State Council Development Research Center, the World\nBank, Asian Development Bank and Goldman Sachs Co. have all predicted that China's\neconomy will maintain a growth rate of over 6 percent until 2010. [CIIC]\nChina currently consumes over 20 percent of the world's aluminum and copper and is the\nworld's biggest user of steel and cement. Even allowing for a slowdown in growth rates, China\nis expected to account for 30 percent of global base metals demand by 2010. [CS] This\ncontinued economic expansion, coupled with China's increasing demand for base metals, is\nexpected to maintain aluminum's, and hence secondary aluminum's, long-term trend of just\nunder 3 percent annual growth for the foreseeable future. [NRC]\nFor these reasons, the secondary aluminum industry will continue to be a large consumer\nof energy for some time to come. The worldwide requirement for aluminum, coupled with\nnatural gas price trending, will increase the demand for new energy conservation and reduction\ntechnologies in the secondary aluminum industry.\nThe aluminum industry has been proactive in adopting new equipment and technologies\nin the past several years, making significant reductions in its energy usage. In 2002 alone, the\nindustry reduced its energy consumption by 22 percent. [AA2] The U.S. Department of Energy\n(DOE) has established several grant programs intended to increase the attractiveness of energy\nrelated projects. Aside from grants to manufacturing facilities, the DOE has provided millions of\ndollars in research grants. Several new technologies have been built and tested, such as a\nceramic coating for recuperation, cogeneration from furnace exhaust and innovative furnace\ncontrol mechanisms. This paper details one such innovation.\n\n\nProcessing Secondary Aluminum at the Plant\n\n\nAluminum scrap comes from a variety of sources. \"New\" scrap is generated by pre-\nconsumer sources, such as drilling and machining of aluminum castings, scrap from aluminum\nfabrication and manufacturing operations and aluminum bearing residual material (dross)\nskimmed off molten aluminum during smelting operations. \"Old\" aluminum scrap is material\n\n\n\u00a9) 2005 ACEEE Summer Study on Energy Efficiency in Industry\n\n\n1-180\n\n\nthat has been used by the consumer and discarded. Examples of old scrap include used\nappliances, aluminum foil, automobile and airplane parts, aluminum siding, and beverage cans.\nScrap pretreatment involves sorting and processing scrap to remove contaminants and to\nprepare the material for smelting. Sorting and processing separates the aluminum from other\n(ferrous) metals, dirt, oil, plastics, and paint. Pretreatment cleaning processes are based on\nmechanical and pyrometallurgical (drying) techniques. Once a facility is ready to process scrap\nmetal, it is first transported to a shredder, where it is reduced to a manageable size. The material\nleaving the shredder is called \u201cturnings.\" After the shredder the ferrous metal is separated using\nan iron separator. The resulting aluminum scrap is then transported to a dryer.\nBefore aluminum scrap can be put into the furnace it must be decoated and visually\ninspected to ensure that no organics or oils remain on the surface. Typically, this cleaning is\naccomplished by using a rotary kiln dryer. A typical rotary kiln dryer uses heat to separate\naluminum from contaminates and other materials. The turnings are heated to high temperature\n(300\u00b0F-450\u00b0F) to vaporize or carbonize the organic contaminates, but not high enough to melt\nthe aluminum (1,220\u00b0F). Once the turnings are dried, they fall into another bin which is\ntransported to the charging well of the furnace, where they are added to the pool of molten\naluminum. If the quality of the scrap resulting from the drying process is not adequate, the\ndrying process may be repeated several times. Besides this uncertainty, kilns have additional\nshortcomings: 1) inability to process solid organics such as rubber and plastics; 2) high levels of\noil and paint; 3) a tendency to formulate dust as a result of direct flame impingement on the scrap\naluminum; 4) a tendency to oxidize aluminum scrap, which increases the amount of dross in the\nsmelter; and 5) a tendency to emit volatile organic compounds (VOCs).\nAfter scrap pretreatment in the rotary dryer, smelting and refining are performed.\nSmelting and refining in secondary aluminum recovery takes place primarily in reverberatory\nfurnaces. These furnaces are refractory brick lined and constructed with curved ceilings. A\ntypical reverberatory furnace has an enclosed melt area where the flame heat source operates\ndirectly above the molten aluminum. The furnace charging well is connected to the melt area by\nchannels through which molten aluminum is impelled from the melt area into the charging well.\nAluminum flows back into the melt section of the furnace by force gravity.\nMost secondary aluminum recovery facilities use batch processing in smelting and\nrefining operations. It is common for one large melting reverberatory furnace to support the flow\nrequirements for two or more holding furnaces. The melting furnace is used to melt the scrap\nand remove impurities or entrained gases. The molten aluminum is then gravity fed into a\nholding furnace. Holding furnaces are better suited for final alloying and for making any\nadditional adjustments necessary to ensure that the aluminum meets product specifications.\nMolten aluminum is then poured into molds or is used as feedstock for continuous casters.\n\n\nFacility Description\n\n\nThe facility described in this paper performs secondary smelting and refining of non-\nferrous and ferrous metals. The company produces approximately 18,000 tons of ferrous and\n4,000 tons of non-ferrous metal monthly. This facility is a secondary aluminum recycler,\nrecovering aluminum from scrap metal to make die cast alloy ingots and sows for the metal\ncasting industry. Their aluminum recycling strategy differs only slightly from that of a typical\nsecondary smelter, in that only one reverberatory furnace is used to support the entire facility's\n\n\n\u00a9) 2005 ACEEE Summer Study on Energy Efficiency in Industry\n\n\n1-181\n\n\noutput. This facility is one of the primary providers of aluminum in the Chicago metropolitan\narea and continues to make efforts to reduce its energy consumption.\nThe energy consumption at this facility is comprised of electricity and natural gas. The\ntotal electricity consumption at this facility was 7,000,000 kWh/yr, with an average demand of\n1,513 kW/month during the base year. The total natural gas consumption at this facility was\n246,171 MMBtu/yr. Natural gas at this facility is consumed by two systems, the furnace and the\nrotary kiln dryer. The breakdown of the natural gas consumption at this facility is shown in\nTable 1 below.\n\n\nTable 1. Natural Gas Breakdown\n\n\nFurnace Dryer Total\n172,319 73,852 246,171\n70% 30% 100%\n\n\nAs with all secondary aluminum smelters the main energy users are the drying process\nand the reverberatory furnace. This facility uses a conventional rotary kiln dryer to clean its\nscrap. The dryer consists of one 7 MMBtu/hr burner on each end of the kiln, as well as two 7\nMMBtu/hr burners, which are used in the afterburner. The afterburner is used to incinerate the\norganics and oils, which are stripped from the aluminum scrap. The exhaust of the afterburner at\nthis facility is cooled and then routed into a baghouse where any aluminum particles are removed\nbefore the air is exhausted into the atmosphere. In a conventional rotary kiln dryer, scrap builds\nup at the bottom of the chamber. To control the amount of buildup, a 50 hp motor is used to\nrotate the kiln at 5 RPM. The rotation provides evenly cleaned scrap, as well as a control\nmechanism for the scrap feed rate. The kiln is also pitched to ensure a consistent direction of\nscrap flow. A picture of the rotary kiln dryer used at this facility is shown in Figure 1 below.\n\n\nFigure 1. Rotary Kiln Dryer\n\n\n\u00a9) 2005 ACEEE Summer Study on Energy Efficiency in Industry\n\n\n1-182\n\n\nPlant management states that the drying process is a bottleneck for the facility. In some\ncases the drying process must be repeated four times before the quality of the scrap is\nsatisfactory for melting. This bottleneck limits the amount of aluminum that can be produced in\na day, which in turn limits the facility's profit potential. In addition to the productivity and\nthroughput losses, each additional pass through the dryer consumes more energy. Eliminating\nthe need to repeat the process will result in a natural gas usage reduction of up to 75 percent by\nthe kiln. In addition, removing the bottleneck will also enable the facility to increase aluminum\nthroughput each day.\n\n\nThe Evolution of Drying Technologies\n\n\nThe limitations of the conventional rotary kiln dryer have sparked several projects related\nto improvements of the design. In 1996, a grant was given to an engineering research company\nthrough the DOE to begin researching design improvements for rotary dryers. The first\nimprovement realized was heat generated by the afterburner could be recirculated back into the\nkiln. This improvement would result in a significant reduction in natural gas usage by\nrecovering the waste heat and preheating incoming air. This new rotary decoater was called the\nIDEXT\nTM \nor indirect fired system. The only difference between this system and the conventional\ndryers was the heat recovery, leaving the same rotary element in place. Because the rotary\nelement has not been eliminated, the IDEX\u2122\nexperiences similar productivity drawbacks to that\nof a conventional kiln. A diagram of the IDEX dryer is shown in Figure 2 below.\nTM\n\n\n\u0648\n\n\nTM\n\n\nFigure 2. The IDEX\n\n\nKiln\n\n\nScrap Entry\n\n\nAir Lock\nSystem\n\n\nVariable-\nSpeed\nFan\n\n\nTo\nExhaust\nBaghouse\n\n\nProduct Flow\nGas Flow\n\n\nAfter-\nburner\n\n\nLocating Spiders\n\n\nRotary Drum\n\n\nHot Gas\n\n\n\u2193\n\n\nAccess\n\n\n3 Door\n\n\nAir Clean\nLock Scrap\nSystem\n\n\nIntegral Return\nGas Duct\n\n\nIDEX Aluminum Scrap Decoater\n\n\nSource: Industrial Technologies Program. 2000.\n\n\n\u00a9) 2005 ACEEE Summer Study on Energy Efficiency in Industry\n\n\n1-183\n\n\nAs depicted in Figure 2, hot gasses generated by the afterburner can be routed through\ntwo outlets. One outlet is routed to the baghouse, while the other outlet is routed back into the\nkiln. Hot gasses entering the kiln are sent through a scooping tube and ejected onto the opposite\nwall. A curved shield directs the gasses back through the kiln in the opposite direction of the\nscrap flow. This increases the amount of heat transfer as counterflow heat transfer designs are\nsometimes more effective due to the increased temperature difference. A variable speed fan\ncontrols the amount of gasses returned into the kiln. As the diagram shows, there are no burners\nin the kiln itself. Instead, indirect hot gasses generated by the afterburner are used to clean the\nscrap. This causes the IDEX\u2122\nto consume approximately half the natural gas that a\nconventional kiln uses because of the heat recovery used.\nAlthough the realization of the potential for heat recovery was a large step in the\nevolution of drying technology, many engineers thought the design could be further improved. It\nwas thought if the design of the dryer allowed the scrap to be cleaned without touching the\nsurface of the chamber, it would be cleaner the first time through, thus reducing the energy\nconsumption. Therefore, the DOE and a private engineering firm developed and tested a new\ntechnology for decoating aluminum scrap. This technology, the variable floatation decoater,\nremoves organics of any type from scrap aluminum and steel without the need of a rotary\nelement. High velocity gases mechanically strip off liquid organics, while high gas temperatures\nvaporize the remaining organics in a low oxygen environment. The organic laden gases are sent\nto an afterburner to destroy the organics; the heat finally produced is used to drive the process.\nThe VFD meets the Environmental Protection Agency's new Clean Air Act standards for dryers,\ndecoaters and delacquering kilns. The main difference between the VFD and the IDEX\nis that\nthe need for a rotating kiln has been eliminated. The rotating kiln has been replaced with a\nvertical cone shaped decoating chamber and a holding tank. Scrap is fed through the top of the\ncone using a bucket loader. Upon entering the cone the scrap is exposed to high temperature,\nhigh velocity gasses. The gas flowrate is strictly regulated so that the aluminum scrap floats in\nthe gasses. As the scrap floats down through the cone and into the holding tank, it is cleaned.\nThis innovation allows scrap to be decoated evenly, quickly and thoroughly. A conceptual VFD\nis shown in Figure 3.\nThe benefits of the vertical floatation decoater include increases in decoating\neffectiveness, speed and energy efficiency. It has a small footprint, minimal moving parts and is\nless expensive to install than a standard rotary kiln.\n\n\n-TM\n\n\nAnalysis\n\n\nEnergy Usage of the VFD\n\n\nThe energy usage of the VFD is significantly less than that of a conventional kiln. The\nincorporation of heat recovery eliminates two burners. The vertical decoating element eliminates\nthe need for a large motor while reducing the cycle time. The effects of these changes on the\nenergy consumption of the VFD are discussed in detail below.\nThe VFD does not have a rotating core, therefore, there is no electric motor for the\nmechanical rotation. The scrap is evenly decoated in the cone. Although the energy\nconsumption of a large motor is significant, the majority of the energy consumption is in the\nform of natural gas. In a conventional rotary kiln, as well as the VFD, natural gas is used in two\n\n\n\u00a9) 2005 ACEEE Summer Study on Energy Efficiency in Industry\n\n\n1-184\n\n\nplaces: to accomplish decoating and to incinerate organics and oils from the aluminum. The\nmain difference is a conventional rotary dryer uses direct heat to accomplish the drying, while\nthe VFD uses waste gasses to drive the decoating process. The VFD is similar to the IDEX\u2122 in\nthis manner. Aside from the direct reduction in natural gas consumption, there are additional\nreductions due to indirect benefits of introducing hot metal directly into the furnace.\n\n\nTM\n\n\nFigure 3. VFD Diagram\n\n\nCone\n\n\nHolding\nTank\n\n\nOptional\nConveyer\nAssembly\n\n\nMaterial\nShredder\n\n\nSource: Energy Research Company. 2003.\n\n\nThe unique design of the cone enables scrap to be cleaned properly the first time through\nthe machine. This eliminates the repetition normally associated with conventional dryers, saving\nthe natural gas required to repeat the cleaning process. Properly decoating scrap the first time\ncan save as much as a 75 percent in natural gas usage per pound. In addition, hot and dry scrap\nfrom the VFD at temperatures of 500\u00b0F to 700\u00b0F can be introduced directly into melting furnace.\nBy contrast, conventional dryers, which often require material to be reworked, necessitate a\nreserve of pre-dried scrap for the melting furnace to maintain production. Introducing cold\naluminum scrap into the furnace has a negative impact on energy use not realized by introducing\nthe hot turnings.\nWhen analyzing the energy usage of the VFD it is best to talk in terms of specific energy\nusage in units of Btu/lbm of scrap. Energy Research Company, located in New York,\nconstructed a scale model of the VFD in an effort to analyze its natural gas consumption. The\nmodel was capable of 1,000 lbm/hr of scrap. Several test runs were preformed using different\nscrap rates and operating conditions. Figure 4 shows the results of the tests; there is an\nexponential relationship between the scrap rate and the specific energy use.\n\n\n\u00a9) 2005 ACEEE Summer Study on Energy Efficiency in Industry\n\n\n1-185\n\n\nFigure 4. Test Data for the Vertical Flotation Decoater\n\n\nSpecific Energy Use (Btu/lbm)\n\n\n7,000\n\n\n6,000\n\n\n5,000\n\n\n4,000\n\n\n3.000\n\n\n2,000\n\n\n1,000\n\n\n2,000 4,000 6,000 8,000 10,000 12,000\nScrap Feed Rate (Ibm/hour)\n\n\n14,000\n\n\nSource: Industrial Technologies Program. 2000.\n\n\nProductivity Improvements of the VFD\n\n\nThe VFD offers productivity benefits in addition to reduced energy consumption,\nincluding reduced decoating time and increased throughput. While this is due to a number of\ndesign changes, the vertical cone is the only fundamental difference between the VFD and its\nolder counterpart. Due to the vertical decoating chamber, the gasses within the cone move with\nhigher velocities than a standard kiln. Measurements taken by Energy Research Company on the\nVFD model show that an increase of up to six times the flowrate is possible by using a vertical\nchamber. The specific flowrate of the gasses in the cone is designed to maintain a floating bed\nof scrap depending on the scrap feed rate. Therefore, increased velocity is required in order for\nthe VFD to operate properly. This design innovation enables scrap to be decoated over a shorter\ntime period as compared to conventional rotary kilns. The increase in the velocity of the gasses\nresults in an increase of the heat transfer coefficient. Due to the increased heat transfer\ncoefficient, the time required to decoat scrap can be reduced.\nFigure 5 illustrates that as the gas velocity increases the heat transfer coefficient\nincreases, thereby raising the overall energy efficiency of the VFD. Typical velocities in the\nVFD range from 40 ft/sec to 80 ft/sec. As the graph shows, the relationship between the gas\nvelocity and the heat transfer coefficient in this range is linear. The increased heat transfer\ncoefficient enables the VFD to decoat scrap quickly and efficiently as compared to a\nconventional kiln resulting in increased productivity and lower operating costs.\nFigure 6 illustrates an exponential relationship between the gas velocity and the relative\ndecoating time. The relative decoating time refers to percent time as compared to a conventional\nkiln. The graph shows that as the gas velocity is increased, the amount of time saved by using\nthe VFD as opposed to a conventional system increases. However, the curve approaches 20\npercent time asymptotically, denoting a limit on the time savings. Because the speeds at which\ngasses flow through the cone of the VFD are dependent on the scrap rate, the most efficient\noperation of the VFD would be at higher loads. At high scrap rates the specific energy usage is\n\n\n\u00a9) 2005 ACEEE Summer Study on Energy Efficiency in Industry\n\n\n1-186\n\n\nsmall, the gas velocity is high and the scrap is decoated in less time. Therefore, a VFD\nreplacement is most attractive in the scrap rate range of 6,000 lbm/hr and higher.\nAs described earlier, conventional kilns represent a bottleneck in the aluminum recycling\nprocess. The reduction in decoating time would enable facilities to eliminate the bottleneck\nassociated with conventional kilns while increasing the amount of aluminum produced each day.\nIn addition to the decreased cycle time of the VFD, scrap from the VFD is cleaned more\nthoroughly the first time through. Built up residue on the interior of a conventional kiln has a\ntendency to accumulate on the scrap as it is fed through the chamber. The higher velocity gasses\nin a VFD do not allow buildup on the interior walls or the scrap, enabling the scrap to be cleaned\nmore thoroughly the first time through. Figure 7 shows decoated scrap after one pass through a\nconventional rotary kiln. As the picture shows, this scrap contains residual organics and cutting\noils. It is likely that the decoating process would have to be repeated for this sample of scrap.\nFigure 8 shows the decoated scrap after one pass through a VFD. This scrap does not have any\nresidual oils or organics and is significantly cleaner than the previous picture. The scrap in this\npicture can be directly fed into the melting furnace eliminating the bottleneck.\nIt is important to note that there is a small, but noticeable, increase in the size of the fan\nused by the VFD to accommodate the higher gas velocities. This small increase is insignificant\nwhen compared to the value of reduced decoating time. The improvement in the quality of the\nscrap, as well as the reduced cycle time, results in increased profit potential.\n\n\nFigure 5. Heat Transfer Coefficient\n\n\nVFD Decoater\n\n\nHeat Transfer Coeff.\n\n\n\u0f0d\n\n\n(Btu/Ibm ft^2 f)\n\n\n80\n60\n862\n40\n20\n\n\nConventional\n\n\nVFD\n\n\n30 40 50 60 70 80 90 100\n\n\n10\n\n\n20\n\n\nGas Velocity (fps)\n\n\nSource: Energy Research Company. 2003.\n\n\nFigure 6. Relative Decoating Time\n\n\nVFD Decoater\n\n\n1.2\n1.0\n0.8\n0.5\n0.4\n0.2\n\n\nConventional\n\n\nVFD Decoater\n\n\nT\n\n\nT\n30\n\n\nT\n\n\nT\n\n\nT\n40 50 60 70 80 90 100\n\n\n10 20\n\n\nGas Velocity (fps)\n\n\nSource: Energy Research Company. 2003.\n\n\n\u00a9) 2005 ACEEE Summer Study on Energy Efficiency in Industry\n\n\n1-187\n\n\nFigure 7. Scrap from a Conventional Decoater\n\n\nSource: Energy Research Company. 2003.\n\n\nFigure 8. Scrap from the VFD Decoater\n\n\nSource: Energy Research Company. 2003.\n\n\nResults\n\n\nEnergy Savings\n\n\nWhen comparing the energy consumption of a conventional rotary kiln to that of a VFD\ndecoater, several things must be considered. The existence of heat recovery and variable\nfrequency fan elements make it difficult to analytically predict the energy consumption of the\nVFD, whereas the energy consumption of a conventional kiln can be calculated using a simple\nanalysis of energy per pound of scrap.\nIn order to calculate the amount of natural gas saved by replacing the current rotary kiln\nwith the VFD system, it is necessary to know what the current energy usage is. Specific energy\nusage, or energy usage per pound, will be used for energy savings calculations, as this is the best\n\n\n\u00a9) 2005 ACEEE Summer Study on Energy Efficiency in Industry\n\n\n1-188\n\n\nway to express each of the system's effective energy usage. In addition, the specific energy\nusage of the VFD has already been measured and graphed in terms of specific energy usage. On\naverage, the scrap rate through the kiln for the facility is 6,000 lbsm/hr. The annual operating\nhours were determined to be 6,336 hr/yr after consulting plant management.\nNatural gas\nmetering data collected on site revealed that the natural gas usage of the dryer was 30% of the\nnatural gas usage of the building, which equates to 74,000 MMBtu/yr. The specific energy use\nof the kiln, SEUC, can be found using the following equation:\n\n\nGUK\n(C\u2081\u00d7SRxhk)\n\n\nSEUC=\n\n\nwhere\n\n\nannual natural gas usage of the kiln, 74,000 MMBtu/yr\naverage scrap rate kiln, 6,000 lbsm/hr\n\n\nGUK\n\n\nSR\n\n\n|| || ||\n=\n\n\nannual operating hours of the kiln, 6,336 hr/yr\nconversion constant, 1\u00d7 106 MMBtu/Btu\n\n\nhk\nC\u2081\n\n\nThe specific energy use of the kiln, SEUC, is then:\n\n\n74,000\n(1\u00d710\u200d6\u00d76,000\u00d76,336)\n\n\nSEUC=\n\n\nSEUC 1,950 Btu/lbm\n\n\nAccording to Figure 4, at a scrap feed rate of 6,000 lbsm/hr the specific energy usage of\nthe VFD would be 1,000 Btu/lbsm. Comparing this specific energy usage to the specific energy\nusage of the current system, this facility would realize a savings of 950 Btu/lbsm. The total\nannual natural gas savings, GS, and the total annual natural gas cost savings, GCS, can be\ncalculated using the following equations:\n\n\nGS=(SEU-SEU\u2081)\u00d7SR\u00d7h\u00d7C\u2081\n\n\nK\n\n\nand\n\n\nGCS=GS GC\n\n\nwhere\n\n\nspecific energy usage of the VFD, 1,000 Btu/lbm\n\n\nSEUP =\n\n\nThe total annual natural gas savings, GS, and the total annual natural gas cost savings, GCS, are\nthen:\n\n\nGS=(1,950-1,000)\u00d76,000\u00d76,336\u00d71\u00d7106\n\n\nGS=36,000 MMBtu/yr\n\n\nGCS 36,000 MMBtu/yr $4.43/MMBtu/yr\nGCS=$160,000/yr\n\n\nThese calculations show a reduction of 50% in the natural gas usage of the drying\nprocess. This represents a considerable impact on the process' energy consumption. Using this\nfacility as a benchmark, a 50% reduction in the natural gas used for drying by facilities in the\nUnited States represents potential for a significant decrease in not only energy consumption, but\nalso in associated volatile emissions. It should be noted that both the VFD and the conventional\n\n\n\u00a9) 2005 ACEEE Summer Study on Energy Efficiency in Industry\n\n\n1-189\n\n\nrotary kiln energy consumption are both once through systems. Since, in many cases, the\nconventional system requires the metal to pass through up to four times before the quality of\nscrap is achieved, the energy consumption per pound of scrap for the conventional system is\nhigher.\n\n\nProductivity Savings\n\n\nIn addition to the energy savings resulting from the VFD's installation, this facility would\nalso realize a reduction in cycle time. The success of the VFD to reduce the time required to\ndecoat is based on its high heat transfer coefficient, which increases the rate and effectiveness of\ndecoating. With the type of scrap and feed rate at this facility, the velocity would be 60 ft/sec,\naccording to the VFD's designer. As a result of the vertical decoating chamber, the VFD at this\nfacility should be capable of decoating the same amount of scrap in 80% less time than a\nconventional decoater, based on the graph shown in Figure 6. The details of the impact of such a\ntime decrease are beyond the scope of this report. However, by analyzing the amount of time\nspent drying scrap, 6,336 hr/yr, at a scrap rate of 6,000 lbsm/hr, it is estimated that this facility\ndecoats a total of 38 million pounds of scrap per year. By installing the VFD, this facility would\nbe able to produce 38 millions pounds of scrap in 1,267 hours, corresponding to an 80%\nreduction in cycle time. Conversely, the facility could produce 152 million more pounds of\nscrap per year at 6,336 hr/yr. This dramatic output increase would enable the facility to increase\nits production and its profit potential. Furthermore, if the facility is not capable of producing this\nlevel of output for any reason, the VFD would be operating less time throughout the year, which\nwould result in further cost savings due to less operating time.\n\n\nConclusion\n\n\nOver the past several years the aluminum industry has been actively pursuing energy\nconservation improvements. In 2002 alone, the industry reduced its energy consumption by 22\npercent. [AA2] The VFD is a primary example of the effort put forth by the aluminum industry\nto formulate and implement new energy related innovations. The VFD is a solution to many of\nthe drawbacks of a conventional kiln. Its rapid decoating time, and ability to produce clean\nproduct the first time through, would enable facilities to eliminate their bottlenecks as well as\nincrease their output and profit margin. Its small footprint makes installation easy and\naffordable. The VFD has no moving parts, which reduces the need for maintenance and\ncleaning. It has the capability of reducing the natural gas usage by half, which would result in\nlower operating costs and decreased VOC emissions. These improvements to the VFD make it\nan ideal replacement for the conventional rotary kiln. At the plant studied, the VFD has the\npotential to reduce annual natural gas consumption of the drying process by 36,000 MMBtu, a\ndecrease of 50%. Additionally, VOC emissions by the plant would be reduced and the amount\nof energy used to melt the scrap decreased. With its improved design, the VFD has the potential\nto save over 17 trillion Btus of natural gas in the U.S. alone. [EERE] The increasing volatility of\nthe natural gas market along with the increasing demand for aluminum makes energy and\nproductivity related advances crucial considerations for the industry. The capability of the VFD\nto reduce operating costs, decrease cycle time, and increase profit margins will make the VFD a\nviable option for scrap decoaters in the United States and worldwide.\n\n\n\u00a9) 2005 ACEEE Summer Study on Energy Efficiency in Industry\n\n\n1-190\n\n\nReferences\n\n\n[AA1] The Aluminum Association, Inc. 2004. \u201cEnvironment and Climate Change: Conservation,\nPreservation, and Recycling.\" Available online: www.aluminum.org/Content/\n\n\nNavigationMenu/The_Industry/Government_Policy/Climate_Change/\nClimate Change.htm. Washington, D.C.\n\n\n[AA2] The Aluminum Association, Inc. 2004. \u201cIn-Depth Information/Recycling Process.\"\nAvailable\nonline:\nwww.aluminum.org/Content/NavigationMenu/\n\n\nWashington,\n\n\nThe_Industry/Recycling/In-depth_information/In-depth_information.htm.\n\n\nD.C.\n\n\n[CD] China Daily Website. 2005. \u201cGDP Surge Epitomizes Sound Growth Of Economy.\"\nAvailable online: www.chinadaily.com.cn/en/doc/2004-01/21/content_300663.htm.\n\n\n[CIIC] China Internet Information Center. 2005. \u201cChinese and Foreign Institutions Predict\nChina's 2000-10 Economic Growth.\" Available online: www.china.org.cn/baodao/\nenglish/newsandreport/2002june1/11-5.htm.\n\n\n[CS] Credit Suisse Asset Management. 2005. Asset Management Market Update, January 2005.\nAvailable online: lu.csam.com/dss/fundsamdi2/mu_0501b_eng.pdf.\n\n\n[EERE] Energy Efficiency and Renewable Energy. 2005. \u201cInnovative Vertical Floatation Melter\n(VFM) and Scrap Dryer.\" Aluminum Project Fact Sheet. Available online:\nwww.eere.energy.gov/industry/metalcasting/pdfs/floatation.pdf. Washington, D.C.\n\n\n[ERCO] Energy Resource Company. 2005. \u201cVFM Decoater.\u201d Available online: www.er-\nco.com/rd_decoat.htm. New York, N.Y.\n\n\n[ITP] Industrial Technologies Program. 2000. \u201cIndirect-Fired Kiln Conserves Scrap Aluminum\nand Cuts Costs.\" Energy Matters. Energy Matters Newsletter. November/December\n2000.\nAvailable online: www.oit.doe.gov/bestpractices/energymatters/\n\n\nnov2000sup_kiln.shtml. Washington, D.C.\n\n\n[NRC] National Resource Center. 2005. Available online: http://www.nrcan.gc.ca/\nmms/pdf/nfo/nfo00/alum-e.pdf. Ottawa, ON, Canada.\n\n\n1-191\n\n\n\u00a9) 2005 ACEEE Summer Study on Energy Efficiency in Industry\n"}, "expected_output": {"claims": [{"unit": "Btu/lbm", "value": 1000, "evidence": ["specific energy usage of the VFD, 1,000 Btu/lbm", "According to Figure 4, at a scrap feed rate of 6,000 lbsm/hr the specific energy usage of the VFD would be 1,000 Btu/lbsm. Comparing this specific energy usage to the specific energy usage of the current system, this facility would realize a savings of 950 Btu/lbsm.", "According to Figure 4, at a scrap feed rate of 6,000 lbsm/hr the specific energy usage of the VFD would be 1,000 Btu/lbsm."]}, {"unit": "Btu/lbm", "value": 1950, "evidence": ["SEUC 1,950 Btu/lbm"]}]}, "metadata": {"product_category": "Metal, mineral, plastic & glass products", "request_id": "req_16a05dd3ed365544"}}
{"id": "7a5ca608ecc24a0763b676fb", "input": {"query": "What is the thermal energy consumption range for peeled tomato processing in kWh per ton or MJ per kg?", "source_url": "https://www.accelwater.eu/files/articles/sp/AW_SP_05.pdf", "document_text": "Journal of Cleaner Production 425 (2023) 138996\n\n\nJournal of\n\n\nContents lists available at ScienceDirect\n\n\nCleaner\nProduction\n\n\nJournal of Cleaner Production\n\n\nELSEVIER\n\n\njournal homepage: www.elsevier.com/locate/jclepro\n\n\nReview\n\n\nEnhancing resource efficiency and sustainability in tomato processing: A\ncomprehensive review\n\n\nCheck for\nupdates\n\n\na,b\n\n\na, b\nGiovanna Ferrari Gianpiero Pataro\n\n\na, b,\n\n\na,b\n\n\n*\n\n\nEmad Abdurrahman\n\n\nElham Eslami\n\n\n\u0648\n\n\n,\n\n\n,\n\n\na\n\n\nDepartment of Industrial Engineering, University of Salerno, Via Giovanni Paolo II, 132, 84084, Fisciano, SA, Italy\nb ProdAl Scarl - University of Salerno, Via Giovanni Paolo II, 132, 84084, Fisciano, SA, Italy\n\n\nARTICLE INFO\n\n\nABSTRACT\n\n\nHandling Editor: Panos Seferlis\n\n\nThe increasing global demand for water and energy resources, coupled with the scarcity of freshwater and fossil\nfuels, highlights the urgent need for efficient resource utilization and sustainable practices across industries.\nIndustrial tomato processing, a prominent segment within the food processing industry, consumes substantial\namounts of water and energy which are interconnected each other through various processing stages.\nA systematic approach characterizing the water and energy flows and their link in tomato processing helps to\nunderstand how these resources are used in tomato processing and what opportunities exist for improving ef-\nficiency. This enable decision makers to implement tailored strategies for water and energy conservation, and\nwaste management enabling to enhance both efficiency and sustainability in tomato processing.\nThis review provides a comprehensive description of the processing lines involved in tomato processing, with a\nspecific focus on the key steps impacting water and energy consumption as well as waste generation. Further-\nmore, it proposes a quantitative methodological approach based on water-energy nexus (WEN) assessment,\nwhich establishes baselines and identifies opportunities for improving resource efficiency. The review also ex-\nplores a range of conventional and novel measures and technologies for water conservation, energy recovery, and\nefficiency across the various stages of tomato processing. It delves into their advantages and limitations, offering\ninsights into their applicability within the industry. By examining these approaches, the review aims to provide\nvaluable guidance for stakeholders in the tomato processing industry seeking to optimize resource utilization,\nreduce environmental impact, and improve overall sustainability.\n\n\nKeywords:\n\n\nTomato processing industry\nWater-Energy Nexus (WEN)\n\n\nAdvanced technology\nEnergy efficiency\nWater savings\n\n\nSustainability\n\n\nvegetable processing (Mekonnen and Gerbens-Leenes, 2020; Peterson\net al., 2022). In these sectors, water of potable quality is commonly\nemployed as an ingredient, for cleaning, heating, cooling, trans-\nportation, and other essential processes (Maxime et al., 2006). Unfor-\ntunately, while significant strides have been taken to enhance water use\nefficiency in agriculture using modern technologies like the Internet of\nThings (IoT), drones, and satellites, as well as innovative methods such\nas smart farming (Abdul Rajak, 2022), there is still limited effort from\nthe food and beverages industry to reduce freshwater consumption\nduring the processing of raw materials. Moreover, around 70% of the\nfreshwater used being discharged as effluent containing high levels of\nbiological oxygen demand (BOD) and chemical oxygen demand (COD)\n(Meneses et al., 2019; \u00d6lmez, 2013). Hence, the management of water\nresources within the food industry remains less than optimal (Meneses\net al., 2019). This is despite wastewater treatment facilities progres-\nsively integrating state-of-the-art technologies to meet increasingly\n\n\n1. Introduction\n\n\nThe growing need for water and energy, coupled with the limited\navailability of freshwater and fossil fuels, the alarming climate fluctu-\nations, and environmental concerns, urgently demand for efficient\nresource utilization and the adoption of sustainable and optimized in-\ndustrial practices.\nIndustries of the food and beverage sector are among the most\nenergy-intensive industries that use huge amounts of fresh water for\nvarious processes (Islam and Karim, 2019). According to the United\nNations, globally about 72% of water resources are used for agriculture\nand irrigation, 16% is consumed by municipalities, and 12% goes to-\nward industrial uses (UN-Water, 2021), with 56% of it being consumed\nby the food and beverages industry (Bhatt et al., 2022). Among them, the\nmost water-intensive sectors include soft drinks and bottled water, dairy\nproducts, brewing, wine and spirits, as well as meat and fruits and\n\n\nCorresponding author. Department of Industrial Engineering, University of Salerno, Via Giovanni Paolo II, 132, 84084, Fisciano, SA, Italy.\nE-mail address: gpataro@unisa.it (G. Pataro).\n\n\nhttps://doi.org/10.1016/j.jclepro.2023.138996\n\n\nReceived 29 June 2023; Received in revised form 19 September 2023; Accepted 23 September 2023\n\n\nAvailable online 25 September 2023\n\n\n0959-6526/\u00a9 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-\nnc-nd/4.0/).\n\n\nE. Eslami et al.\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\nstringent legal discharge constraints as well as to enhance reclamation\nefficiency (Borzooei et al., 2020).\nIn terms of energy requirements, the global food sector consumes\napproximately 200 EJ per year (FAO, 2017; Mead, 2017), with pro-\ncessing and distribution activities contributing to about 45% of this total\n(FAO, 2011; Sims et al., 2015). It is also worth noticing that electricity\nconsumption, accounts for one-third of the overall energy consumption\nin the sector. These substantial energy demands significantly impact\nproduction costs in food manufacturing and contributes to air pollution\nand greenhouse gas emissions (GHGES) (FAO, 2017). However, it's\nworth noting that there exists a noteworthy potential for energy savings\nin the domain of food production (Panepinto et al., 2014).\nTherefore, there is an urgent need to rationalize the use of these\nresources, as well as redesign and optimize existing food processing\nplants by implementing tailored strategies for water and energy con-\nservation, waste management, and the utilization of conventional or\nadvanced technological solutions and renewable energy sources. Such\nmeasures aim to significantly improve the efficiency and sustainability\nof the food manufacturing industry (Grinberga-Zalite and Zvirbule,\n2022; Ringler et al., 2016).\nWithin this context, it's essential to highlight the interconnected\nnature of water and energy involved in food processing. Energy (thermal\nend electrical) is required to transport, heat, and cool water. Further,\nwater in the form of steam can be harnessed to produce energy through\nturbines (Am\u00f3n et al., 2017). These relationships are termed the water\nenergy nexus (WEN). In recent years there is growing awareness that\nunderstanding the WEN across many industrial sectors is important for\ncharacterizing water and energy use through the different stages of\nprocessing. Moreover, it facilitates the identification of specific pro-\ncessing areas where water conservation and energy efficiency efforts can\nhave the greatest impact (Am\u00f3n et al., 2017; Peterson et al., 2022).\nThe WEN holds particular significance in the context of food pro-\ncessing, a sector known for its substantial consumption of both water\nand energy resources (Am\u00f3n et al., 2017). In this context, with a global\nproduction exceeding 40 million metric tons annually, the processing of\ntomatoes represents a significant segment within the food processing\nindustry and an intriguing case study. The United States is the leading\nproducer, followed by China and Italy (De Meo et al., 2022). Tomato\nprocessing involves a multi-stage process to produce peeled tomatoes\nand tomato concentrate, which contributes substantially to water con-\nsumption and thermal and electrical energy expenses. The extent of\nthese impacts depends on factors such as the final product, technological\nsolutions, and processing practices (Latini et al., 2017). Furthermore,\ntomato processing generates two primary wastes that require valoriza-\ntion and reutilization. Wastewater is produced during the raw product\nfluming and washing stages, with an estimated production ranging from\n1.5 to 7.5 m\u00b3 per ton of processed tomatoes (Behzadian et al., 2015;\nLatini et al., 2017). Additionally, tomato pomace, consisting of skins and\nseeds, is generated during juice extraction, constituting approximately\n2-5% of the total weight of processed fruits (Eslami et al., 2023; Pataro\net al., 2020). Currently, it is utilized in low-value applications such as\nanimal feed or compost or is sent directly to landfills (Rossini et al.,\n2013; Strati and Oreopoulou, 2014). However, this by-product contains\nvaluable components such as natural carotenoids with antioxidant\nproperties, as well as oil, pectin, cutin, and proteins. Exploring methods\nto recover these components bring significant economic and environ-\nmental benefits (Eslami et al., 2023; Pataro et al., 2020). Additionally,\nresidual biomass could be used as a renewable source to obtain energy,\nin order to reduce both greenhouse gas emissions and reliance on fossil\nfuels (Grinberga-Zalite and Zvirbule, 2022; Panepinto et al., 2014).\nTo address these challenges, it is crucial to adopt a holistic approach\nthat optimizes water and energy utilization while efficiently managing\nwaste in the tomato processing industry. This entails adopting state-of-\nthe-art monitoring systems as well as the adoption of conventional\nmeasures and cutting-edge technologies to minimize the environmental\nimpact and achieve the highest level of economic and environmental\n\n\nsustainability in the industry.\n\n\nA comprehensive review of current literature, which was conducted\npredominantly through the Scopus and Science Direct databases, along\nwith the retrieval of open access project reports, unveiled a collection of\npublications focusing on the efficient use of resource and sustainability\nwithin the food industry and, specifically, in tomato processing sector.\nThese previous works, using different methodological approaches such\nas WEN assessment, Life Cyle Assessment (LCA) and Current Value\nSteam Mapping (CVSM), underscore the pivotal significance of quanti-\nfying water and energy flows within tomato processing facilities. Such\nquantification forms the foundation for propelling improvements in ef-\nficiency and sustainability. Notably, these previous works have high-\nlighted substantial opportunities for savings electrical energy, peak\ndemand, natural gas consumption, and water usage within tomato\nprocessing facilities (Am\u00f3n and Simmons, 2017; Trueblood et al., 2013).\nMoreover, these investigations were primarily addressed at exam-\nining specific tomato processing lines of varying sizes, tailored to the\nproduction of particular products like peeled or paste. Their main focus\nrevolved around thermal and electric energy-related aspects, striving to\nuncover opportunities for augmenting efficiency within the domain of\nindustrial tomato processing (Am\u00f3n et al., 2013; Am\u00f3n and Simmons,\n2017; Trueblood et al., 2013). The emphasis on elucidating the WEN\nwas comparatively limited in these studies even though the use of water\nand energy are inherently linked and thus important for overall process\nefficiency (Am\u00f3n et al., 2013, 2017). Additional explorations have\ndelved into the environmental impact of tomato production (Brodt et al.,\n2013; Folinas et al., 2017; Garofalo et al., 2017; Manfredi and Vignali,\n2014), and the application of innovative technologies aimed at\nimproving process efficiency and product quality (Arnal et al., 2018;\nVidyarthi et al., 2019).\nThis review work is the first attempt to gather, standardize, and\ncritically analyse data achieved from different research groups in\ndifferent processing plant and employing different methodological ap-\nproaches. The primary goal is to equip readers, especially decision-\nmakers, with a valuable instrument that facilitates the implementation\nof tailor-made strategies enabling to enhance both efficiency and sus-\ntainability in tomato processing.\nSpecifically, this review provides a comprehensive description of the\nprocessing lines involved in tomato processing, with a specific focus on\nthe key steps impacting water and energy (thermal and electrical) con-\nsumption as well as waste generation. Furthermore, it also addresses a\nmethodological approach based on the water-energy nexus (WEN)\nassessment for setting up the baselines of water and energy consumption\nand identifying opportunities for improving the efficiency of resource\nusage. Finally, the review also explores a range of conventional and\nnovel measures and technologies for water conservation, energy re-\ncovery, and efficiency across the various stages of tomato processing. It\ndelves into their advantages and limitations, offering insights into their\napplicability within the industry. By examining these approaches, the\nreview aims to provide valuable guidance for stakeholders in the tomato\nprocessing industry seeking to optimize resource utilization, reduce\nenvironmental impact, and improve overall sustainability.\n\n\n2. Tomato processing\n\n\nTomato processing facilities operate continuously during a specific\nperiod, usually spanning from late July to early October, with processing\nseasons typically lasting around 90-100 days (equivalent to approxi-\nmately 2,300 h per year) (Trueblood et al., 2013). The majority of\nprocessed tomatoes are utilized in the production of peeled tomatoes\n(whole, diced, and sliced) as well as tomato concentrates, such as puree\nand tomato paste. Tomato puree has a natural total soluble solids con-\ntent ranging from 6 to 9\u00b0 Brix, while tomato paste ranges between 22 and\n36\u00b0 Brix. In addition to fresh tomatoes, the production of these\ntomato-based products involves various materials, including packaging\ncontainers, fresh water, natural gas, and electricity (Behzadian et al.,\n\n\n2\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\nE. Eslami et al.\n\n\n2015).\n\n\nOverall, the sorting process results in the removal of up to 5% of the\nincoming raw materials (Reyes-de-corcuera et al., 2014), which are\ncollected on a reject conveyor and stored for disposal.\nThe cleaned and sorted tomatoes then undergo thermal treatments,\nthe specifics of which depend on the desired final product.\nFor puree and paste production, suitable tomatoes are sent to\ncrushing machines, which convert them into coarse pulp. The crushed\ntomatoes are subsequently subjected to steam heating within a heat\nexchanger, raising their temperatures to a range of 65-75 \u00b0C for Cold\nBreak (CB) treatment or 85-95 \u00b0C for Hot Break (HB) treatment. The\ntemperature choice depends on the desired consistency of the finished\nproduct and aims to partially or totally inactivate pectolytic enzymes\n(Latini et al., 2017). The heated tomato pulp is subsequently pumped to\na series of refiners that extract the juice (~5\u00b0Brix) with a yield of\napproximately 95%, removing skins and seeds (Giagnacovo et al.,\n2016). The extracted juice is conveyed to a large holding tank, which\nsupplies the evaporation step. In this step, a significant amount of water\nis removed using steam heating, resulting in the formation of tomato\npuree (6-9\u00b0Brix) or tomato paste at different concentrations (>18\u00b0 Brix),\nnamely double (28\u00b0 Brix) or triple concentrate paste (36\u00b0Brix) (Latini\net al., 2017). The concentration step occurs under vacuum conditions\nand at low temperatures, typically ranging from 50 to 85 \u00b0C (Latini et al.,\n2017). This stage is one of the most energy-intensive in the entire pro-\nduction line, with the main operating cost attributed to the steam\ngenerated by a boiler (Giagnacovo et al., 2016; Meneses et al., 2019).\nThe tomato concentrate is then sent to the sterilization/packaging stage.\n\n\nFig. 1 illustrates the typical steps involved in tomato processing lines\nfor the production of either tomato puree/paste or peeled tomatoes,\nstarting from the reception of raw materials and up to the storage of the\nfinal products. It should be noticed that in this schematics, \u201cin-container\nprocessing\" is considered instead of \"aseptic processing\", which involves\nthe cooking, sterilization, and cooling stages prior to packaging (True-\nblood et al., 2013). Furthermore, key stages that consume significant\namounts of water and energy and generate substantial waste are high-\nlighted with kaizen burst icons, indicating areas that require\nimprovement.\nThe processing of tomatoes, regardless of the final products, begins\nwith the arrival of raw tomatoes in trucks at the plant's offloading area.\nFrom there, they are transported to a hydraulic flume where they un-\ndergo washing and sorting before being taken to the processing line. The\nwashing process takes place within the flume network, where a\ncontinuous supply of fresh and recirculated water is used to move and\nwash the tomatoes, removing foreign materials such as leaves, branches,\nsoil, and stones, which can make up to 3-5% (w/w) of raw tomatoes\n(Eslami et al., 2023). This stage is highly water-intensive, requiring\napproximately 3-5 m\u00b3 of water per hour for every 1 m\u00b3 of tomatoes\nprocessed (Latini et al., 2017). Consequently, a significant amount of\nwastewater is generated, which is then pumped to the wastewater\ntreatment unit. After washing, the tomatoes go through a grading and\nsorting station, where manual and automated sorting processes remove\ndefective fruits and unwanted materials, including green tomatoes.\n\n\nPreliminary Stage\n\n\nSolid wet waste (soil, stones,\nbranches, leaves)\n& Wastewater\n\n\nGreen and damaged tomato\n\n\nHigh Water\nDemanding\n\n\nWashing\n\n\nUnloading\n\n\nManual and Optical Sorting\n\n\nTomato Puree and Paste Processing Stage\n\n\nVery High\nThermal Energy\n\n\nHigh Thermal\n\n\nTomato Waste (seeds and peels);\n\n\nEnergy\n\n\nHot/Cold break\n\n\nChoppng\n\n\nEvaporation\n\n\nJuice Extraction\n\n\nPeeled Tomato Processing Stage\nThermophysical Peeling\n\n\nHigh\nThermal\nEnergy\n\n\nTomato Waste (peels)\n\n\nDicing\n(for peeled diced tomato)\n\n\nSteam Blanching\n\n\nVaccum Cooling\n\n\nPinch Roller\n\n\nPackaging/Sterilization Stage\n\n\nHigh\nThermal\nEnergy\n\n\nWater\nDemanding\n\n\nPackaging:\n\n\nWarehouses/\nTransporting\n\n\n-Peeled tomato with juice (8 \"Brix) in cans\n-Tomato puree and paste (8-36 \"Brix) in jars/bottles\n\n\nPalleting\n\n\nSterilization\n\n\nFig. 1. Schematic of tomato puree/paste and peeled tomato production lines.\n\n\n3\n\n\nE. Eslami et al.\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\nDepending on the method chosen, cooking, sterilization, and cooling\nstages can occur either before (aseptic in-line sterilization) or after\n(in-container sterilization) the packaging process in glass bottles and\nplastic bags or jars containers (Trueblood et al., 2013). During\nin-container sterilization, containers filled with tomato concentrate are\nsealed and heated in tunnel spray sterilizer with hot water or steam\nbefore being cooled to room temperature with water spray. Alterna-\ntively, for aseptic in-line sterilization, the tomato paste, or puree un-\ndergoes cooking and sterilization through direct steam injection or\ntubular heat exchangers using overheated water. The sterilized tomato is\nrapidly cooled in tube-in-tube cooling systems before aseptic packaging.\nIn the production of peeled (whole, diced, and sliced) tomatoes, the\nwashed and sorted fruits are routed to the peeling operation, where the\ntomato peel is typically removed using chemical or steam methods\n(Arnal et al., 2018; Kohli et al., 2021; Rock et al., 2012) Peeling typically\noccurs via chemical or thermal methods, which are very water and\nenergy-demanding, and waste-generating and whose performance\nsignificantly impacts the overall process efficiency and quality of the end\nproduct (Rock et al., 2012; Zhou et al., 2022). Following peeling, the\npeeled tomatoes undergo manual and optical sorting to eliminate fruits\nthat do not meet commercial standards in terms of size, color, or the\npresence of black spots or scars on the surface.\nA portion of the whole peeled tomatoes may also be sent to dicers to\nproduce diced or sliced tomatoes. Subsequently, the peeled tomatoes\n(whole, diced, sliced) are filled into tinplate cans and jars of various sizes\n(ranging from 0.5 to 3 kg). The containers then pass through a filler\nwhere tomato juice or a very thin pur\u00e9e is added before removing the air\nto create a vacuum and mechanically or thermally sealing the package\n(Trueblood et al., 2013). The ratio of peeled tomatoes to pur\u00e9e is\napproximately 60:40 (w/w). The sealed packages are then conveyed to\nthe in-container sterilization unit, where the cans are heated by im-\nmersion in a hot water bath before being cooled in water.\nThe exact sterilization temperatures and durations depend on the\nproduct's pH and the package's geometry.\nFinally, containers of canned or aseptically sealed tomato products,\nas well as canned peeled tomatoes, undergo cleaning processes using hot\nwater, steam, or blasts of pressurized air (Trueblood et al., 2013). They\nare then placed in an automatic palletizer for labeling, packaging, and\nsubsequent storage in ambient temperature warehouses until they are\nready to be delivered to clients upon request (Manfredi and Vignali,\n2014).\n\n\nusing an integrated WEN assessment approach. Data gathered from\ninstalled sensors and monitoring systems, simulation software, thermal\nproperties, and interviews with plant operators and technicians were\nutilized for this purpose (AccelWater, 2020). A similar approach has\nbeen also applied in assessing the WEN at an industrial tomato paste\nprocessing plant in California, USA, resulting in the development of a\nmap of water and associated energy use at each processing step (Am\u00f3n\net al., 2017). Furthermore, it was utilized to appraise the potential for\nrecuperating waste heat from condensate and utilizing that energy for\nprocess heating, thereby reducing the use of steam and, as a result,\ndecreasing the consumption of natural gas in boilers (Am\u00f3n et al., 2015).\nNotably, this methodology has also demonstrated its effectiveness in\nvarious other food processing sectors, such as small breweries, where a\nsystematic approach to analyzing water and energy flows has identified\nopportunities for enhancing efficiency by reducing waste (Peterson\net al., 2022).\nIn general, a WEN assessment involves the development of a WEN\nmap that specifically considers unit operations in which water streams\nare directly involved in transforming tomato fruits into bulk concentrate\nor peeled tomatoes while taking into account the different ways in which\nelectrical and thermal energy are embedded in the process water (Am\u00f3n\net al., 2017). These specific unit operations, where water and energy\ninteract during tomato processing, are referred to as WEN points. Fig. 2\nprovides a general schematic of the WEN map for a tomato processing\nfacility, in which certain WEN points are grouped as general operations\nwhere energy is embedded in water during processing. At these points,\nmeasurement or estimation of water and energy demands is necessary to\nquantify the WEN. A detailed description of the WEN points is also re-\nported in Table 1.\nAs shown in Fig. 2, all fresh water used for tomato processing\ngenerally originated from on-site wells. Groundwater may undergo pu-\nrification in mechanical separators to remove grit before being used in\nvarious processes (Am\u00f3n et al., 2017).\nThe treated well water is mainly pumped to pre-processing units to\nunload, wash, sort, and convey tomatoes as they enter the facility. The\nflume water is then processed by electric rotary separators to remove\nsolid waste (e.g., leaves, branches, soil, and stones) and partially recir-\nculated to the washing channel. Wastewater leaving the washing phase\nis sent to wastewater treatment.\nA portion of the treated well water is allocated to supply vacuum\npumps and hose systems for cleaning flume debris separators and facility\nsurfaces (Am\u00f3n et al., 2017). Another portion is delivered to both the\nsingle pass cooling section of the sterilization units and sprayed into\nevaporator condensers to promote condensation and maintain vacuum\n(Am\u00f3n et al., 2017). The spent water from these units is typically sent to\ncooling towers to dissipate waste heat before being recycled to other\nprocessing units. Excess water may be directly sent to wastewater\ntreatment.\n\n\n3. The tomato industrial processing Water-Energy Nexus (WEN)\n3.1. Water-Energy Nexus assessment framework\n\n\nThe food processing industry typically uses substantial quantities of\nwater and energy, which are often linked to each other, given that en-\nergy is required to transport, heat, and cool water, and water in the form\nof steam can be used to generate thermal energy (Am\u00f3n et al., 2017; Liu\net al., 2019). This interdependence is defined as water-energy nexus\n(WEN) (Hamidov and Helming, 2020). Concerning the tomato pro-\ncessing industry, it typically uses great volumes of water for tasks such as\nunloading, sorting, transportation, and heating of tomatoes. Thermal\nand electrical energy is imparted to this water during each processing\nstep primarily by pumps, fans, and boilers to form the tomato processing\nWEN (Am\u00f3n et al., 2017).\nTo gain a comprehensive understanding of the water and energy\nusage throughout the tomato processing facility, a WEN assessment is\nessential. This evaluation provides a quantitative foundation that holds\nutmost importance for the industry's pursuit of enhancing resource ef-\nficiency (Am\u00f3n et al., 2017; Peterson et al., 2022). WEN assessment\nshould systematically account for water consumption and the energy\nrequired to process water at each stage of industrial tomato processing.\nIn the frame of the European project AccelWater (Project ID: 958266), a\nreal scenario of an Italian tomato processing industry was evaluated\n\n\nFor certain applications, treated well water may undergo further\npurification through reverse osmosis (RO). The resulting permeate is\ntypically deaerated and utilized as boiler feed water (Am\u00f3n et al., 2017).\nSteam is employed in different thermal units such as hot/cold break,\nevaporators, peelers, and sterilizers (Fig. 1). Indirect steam heating is\nused in hot/cold break units, as well as rotary coil and shell-and-tube\nheat exchangers in sterilization and evaporation units. Most of the\nsteam condensate from these units is recovered and recycled as boiler\nfeed water. However, condensate from steam supplied to thermal units\nrelying on direct steam heating, such as steam injection systems for paste\nsterilization and tomato peelers, along with tomato water condensate,\nnamely the water vapor removed from tomato juice in the evaporators to\nform tomato concentrate, cannot be recycled due to their impurities.\nThese condensate streams are typically directed to cooling towers to\ndissipate waste heat before being recycled in other units, such as the\npreliminary washing phase. Excess condensate is sent directly to\nwastewater processing (Am\u00f3n et al., 2017).\n\n\nWastewater primarily consists of process water from flumes, and to a\n\n\n4\n\n\nE. Eslami et al.\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\nBoiler\n\n\nWell Water\n\n\nThermal Unit\n\n\nRO Permeate\n\n\nSteam\n\n\nWater Supply\n\n\nWater Treatment\n\n\nSteam Generation\n\n\nTreated\nWater\n\n\nTreated Water\n\n\nSteam Condensate\n\n\n(from direct steam heating\n\n\nSteam Condensate\n\n\n(from indirect steam heating\n\n\nCondensate\nTomato evaporate\n\n\nsingle-pass cooling\n\n\nSpent water from\n\n\nCooled water\n\n\nProcess Water\n\n\nRO Retentate\n\n\nBlow-down\n\n\nScreen Rinse\nWater\n\n\nOverflow\n\n\nPre-processing\n\n\nWash\n\n\nFacility Cleaning Water\n\n\nWaste Water Treatment and Discharge\n\n\nWater Cooling\n\n\nFresh water\nRecycled water\nSteam\nCondensate\nWastewater\n\n\nAerator\n\n\nPump\n\n\nSeparator\n\n\nFan/blower\n\n\nFig. 2. Simplified schematic of the Water-Energy Nexus (WEN) map and the primary WEN points in a tomato processing facility. Equipment involved in embedding\neither electrical or thermal energy in water is depicted within boxes outlined by dashed and dotted lines, respectively. The dashed and dotted arrows represent energy\ninputs, with gold dashed lines indicating electrical energy and red dotted lines representing thermal energy. Arrow width serves as a qualitative indicator of the\nmagnitude of water/steam mass being transferred. The abbreviation \"RO\" corresponds to reverse osmosis. (For interpretation of the references to color in this figure\nlegend, the reader is referred to the Web version of this article.)\n\n\nlike innovative electromagnetic or Clamp-on Doppler or Transit-time\nUltrasonic flow meters (DiGiacomo, 2011; Hauptmann et al., 2002;\nPeterson et al., 2022; Xu et al., 2019) along with monitoring systems at\nWEN points where such installations are feasible.\nMoreover, certain water flow rates, such as those associated with the\nmass of evaporated tomato condensate, can be calculated using facility\nmetrics such as throughput and the solids content of tomato juice\n(~5\u00b0Brix) and tomato concentrate (8-36\u00b0 Brix) (Am\u00f3n et al., 2017). In\ncases where direct flow measurement is not possible for certain streams,\nbut sufficient data is available for related streams within the same\noperation, process simulation tools can be utilized to solve water mass\nbalances and estimate flow rates accurately.\n\n\nlesser extent, blowdown water from the boiler system, cooling tower\noverflow, retentate from the reverse osmosis system, and water used for\nfacility cleaning (Am\u00f3n et al., 2017). Flume water is usually pumped to a\nsedimentation pond to remove solids and then transferred to aerated\nlagoons to facilitate the aerobic microbial degradation of organic mat-\nter. It is subsequently routed to a sump collector along with wastewater\nfrom the steam system and cooling tower before being discharged into\nthe municipal sewer or used for various purposes such as aquifer\nrecharge, irrigation, and truck washing, in accordance with local regu-\nlations (Meneses et al., 2019).\n\n\n3.2. Water, electrical, and thermal energy use assessment\n\n\n3.2.2. Thermal energy use assessment\n\n\nThe data collected from the WEN assessment of water, thermal, and\nelectrical energy usage is essential for identifying inefficiencies within\nunit operations and determining the processing operations that consume\nthe most resources. This data serves as a baseline for identifying op-\nportunities to improve resource efficiency by adjusting water loads on\nequipment (Peterson et al., 2022). It also helps in developing specific\nstrategies for conserving and recovering water and energy, as well as\nmanaging waste effectively throughout the various stages of tomato\nprocessing (Am\u00f3n et al., 2015).\n\n\nThe steam generation system used in the tomato processing facility\ntypically consists of fire tube boilers fuelled by natural gas. These boilers\nproduce steam at a gauge pressure ranging from 10 to 30 bar. Standard\nfire tube boilers, without economizers, generally achieve an 80% con-\nversion efficiency rate from input to output energy (Trueblood et al.,\n2013). The steam generated by these boilers is then directed to collectors\nlocated near different thermal units. These collectors are equipped with\npressure-reducing valves to ensure that the steam is delivered at the\nrequired pressure for specific operations.\nBy examining utility provider records for the amount of natural gas\nconsumed throughout the processing season and considering its ther-\nmophysical properties, it is possible to determine the steam generation\nrate and the total thermal energy associated with steam production. This\nevaluation involves solving mass and energy balances at the boiler,\nusing input data such as the boiler's capacity, operating conditions, ef-\nficiency, heat loss, natural gas supply conditions, boiler makeup water\nusage and temperature, flow rate and temperature of condensate recy-\ncled as boiler feed water from indirect heating operations, annual\n\n\n3.2.1. Water use assessment\n\n\nThe assessment of water usage during tomato processing, both at the\nfacility level and within each unit operation, can start from the known\nseasonal quantity of water pumped from on-site wells. Additionally,\nwater usage specifications provided by equipment manufacturers or\nmeasured flow rate data from pumps supplying water to specific oper-\nations can be utilized to estimate the water demand for those particular\noperations. To improve data precision and facilitate real-time moni-\ntoring, it is advantageous to incorporate water flow rate meter sensors\n\n\n5\n\n\nE. Eslami et al.\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\nTable 1 (continued)\n\n\nTable 1\n\n\nOverview of the WEN points (processes and equipment) of the tomato processing\nindustry where energy (electrical and/or thermal) is embedded with water\nduring the process.\n\n\nWEN point\n\n\nEquipment\n\n\nSource of\nenergy\n\n\nElectric/\nThermal\n\n\nDescription\n\n\nenergy\n\n\nWEN point\n\n\nEquipment\n\n\nElectric/\nThermal\nenergy\nElectric\n\n\nSource of\nenergy\n\n\nDescription\n\n\nsolid waste\n\n\nand\ndischarge\n\n\nremoval.\n\n\nPumping to collect\nand discharge\n\n\nPumps\n\n\nPumping of fresh\nwater from on-site\nwells\nRemoval of grit\nfrom groundwater.\nReverse osmosis to\nsoften water for\npump sealing and\n\n\nWater supply\n\n\nElectricity\n\n\nPumps\n\n\nwastewater.\nBlowers\n\n\nAeration of\nwastewater\nlagoons.\n\n\nWater\n\n\nElectricity\n\n\nMechanical\n\n\nElectric\n\n\nseparators\nMembrane\n\n\ntreatment\n\n\nseparators\n\n\noperating hours, and steam pressure (Am\u00f3n et al., 2017).\n\n\nboiler makeup.\n\n\nLikewise, the steam usage in each relevant thermal unit can be\nestimated by solving local mass and energy balances. This estimation\ntakes into account input data like local steam pressure, heat transfer\ncoefficient, heat transfer area, flow rate, and inlet and outlet tempera-\ntures of the processed product. Alternatively, for a more precise evalu-\nation of boiler steam generation and steam usage at each relevant\nprocessing step, the installation of appropriate steam flow meters such\nas orifice, vortex, and in-line ultrasonic flow meters (Murakawa et al.,\n2021; Steven and Hall, 2009; Zhoua et al., 2018) and thermal energy\nmeters can be implemented (Am\u00f3n et al., 2017; Peterson et al., 2022).\n\n\nPumps\nFlume\n\n\nElectric\n\n\nNon-thermal\n\n\nUnloading,\nconveying,\n\n\nElectricity\n\n\nsystems\n\n\nprocesses\n\n\nwashing, and\n\n\nsorting of tomatoes\nin flume systems.\nRemoval of solids\nfrom flume water\nand subsequent\n\n\nWater\n\n\npumps\n\n\nrecirculation.\n\n\nSealing of pump\nshafts.\n\n\nSorting\nmachine\n\n\nMechanical\n\n\nseparators\nFire-tube\n\n\n3.2.3. Electricity use assessment\n\n\nPumping and\ndeaeration of\nboiler make-up\n\n\nElectricity\nFuel\n(Natural\ngas)\n\n\nSteam\n\n\nElectric\n\n\nThe assessment of electrical energy usage in WEN primarily involves\nthe pumps responsible for distributing and recirculating water within\nand between units. It also includes other equipment motors, such as fans\nused for air supply to boiler furnaces and water evaporation in the\ncooling tower, mechanical separators for removing solids from flume\nwater, and blowers used in lagoons for aerobic wastewater treatment\n(Am\u00f3n et al., 2017). However, the electrical power consumption of\nadditional machinery and equipment, like belt conveyors, packing units,\npinch peelers, choppers, juice, and product circulation pumps, etc., is\nnot considered in the WEN analysis since they are not directly involved\nwith process water (Am\u00f3n et al., 2017).\nThe obtain information about the electrical motors' characteristics,\nsuch as voltage, amperage, and power, one can use one can refer to\nequipment nameplates, and manufacturers' data sheets, or measure\ndirectly using power meters and data loggers (Peterson et al., 2022).\nFurthermore, it is advisable to determine a coefficient of usage for each\nequipment motor, ranging from 0 to 1, which represents the actual\nfraction of time the motors operate during the processing season. This\ncan be achieved through a comprehensive review of operational records,\nincluding data logged by sensors, and by conducting interviews with\nfacility personnel responsible for operating specific equipment (Am\u00f3n\net al., 2017).\nUsing these data, the seasonal energy usage of the equipment\n(measured in kWh) can be calculated as the product of the power\ndelivered to the equipment motors (in kW), the coefficient of usage, and\nthe number of operating hours.\n\n\nBoilers\n\n\ngeneration\n\n\nand\n\n\nthermal\n\n\nwater.\n\n\nSupplying of air to\n\n\nPumps\n\n\nboiler furnaces.\n\n\nBlower of\n\n\nBlower\n\n\ncombustion air.\nSteam generation\n\n\nby boilers.\n\n\nCold/Hot\nbreak\n\n\nThermal\nand\nElectric\n\n\nThermal unit\n\n\nUse of steam to\n\n\nSteam\nElectricity\n\n\nheat products for\n\n\nenzyme\ninactivation,\npeeling, tomato\n\n\nwater evaporation,\n\n\nand sterilization\nprocesses.\n\n\nEvaporators\nPeelers\n\n\nUse of water for: a)\ncondensation of\nevaporated tomato\n\n\nwater and\n\n\nmaintaining the\nvacuum in the\n\n\nevaporators and b)\nproducts cooling\n\n\nafter the\n\n\nsterilization stage.\n\n\nCookers/\nSterilizers\n\n\nPumping of\n\n\nCoolers\n\n\ncondensates and\nexhaust water.\n\n\n3.2.4. Water and energy usage in the tomato processing industry\n\n\nPumps\nPumps\n\n\nTo enhance efficiency in tomato processing, the initial step involves\nidentifying the operations that have the highest water and energy de-\nmands. Several authors have employed various methodological ap-\nproaches, such as WEN assessment, Life Cycle Assessment (LCA), and\nCurrent Value Stream Mapping (CVSM), to estimate water and energy\nusage data in the most significant processing steps of tomato facilities.\nTable 2 presents a comprehensive overview of the findings, highlighting\nthe key processing steps that consume water, electrical and thermal\nenergy in the production of concentrate (puree/paste) and/or peeled\ntomatoes.\n\n\nWater cooling\n\n\nPumping of water\nand circulation of\nair in cooling\ntowers to promote\nwater evaporation\nPumping water to\nrinse facility\n\n\nElectric\n\n\nElectricity\n\n\nFans\n\n\nFacility\ncleaning\n\n\nElectric\n\n\nElectricity\n\n\nPumps\n\n\nsurfaces and\nequipment\nScreening of\n\n\nElectric\n\n\nWastewater\ntreatment\n\n\nSolid\n\n\nElectricity\n\n\nwastewater for\n\n\nseparators\n\n\nIn general, comparing data from different processing plants and\n\n\n6\n\n\nE. Eslami et al.\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\nTable 2\n\n\nSummary of water, electric, and thermal energy consumption in the main processing steps of industrial tomato facility for the production of peeled and/or tomato\nconcentrate.\n\n\nWater\n\n\nProcessing (thermal)\n\n\nProcessing (non-thermal)\n\n\nTomato Product\n\n\nPre-processing\n\n\nReferences\n\n\nAm\u00f3n et al. (2017)\n\n\nPaste (29\u00b0 Brix)/diced tomato\n\n\n67%\n\n\n24%\n\n\n9%\n\n\nManfredi and Vignali (2014)\nArnal et al. (2018)\n\n\nPuree (8\u00b0 Brix)\nPeeled tomato\n\n\n88%\n\n\n12%\n\n\nN/A\n\n\n20%\n\n\n80%\n\n\nN/A\n\n\nThermal energy\n\n\nSteam Peeling\n\n\nTomato Product\n\n\nEvaporation\n\n\nSterilization\n\n\nCB/HB\n\n\nReferences\n\n\nPaste (36\u00b0 Brix)\nPaste (30\u00b0 Brix)\nPuree (8\u00b0 Brix)\nPeeled tomato\nPeeled tomato\n\n\n76.2%\n63.4%\n44.2%\n\n\nGiagnacovo et al. (2016)\nFolinas et al. (2017)\nManfredi and Vignali (2014)\nGarofalo et al. (2017)\nArnal et al. (2018)\n\n\n15.2%\n30.9%\n\n\n8.6%\n\n\nN/A\n\n\n5.7%\n18.0%\n32%\n\n\nN/A\n\n\n37.8%\n\n\nN/A\n19%\n\n\n49%\n\n\nN/A\n\n\n39%\n\n\n61%\n\n\nN/A\n\n\nN/A\n\n\nElectrical energy\n\n\nProcessing (thermal)\n\n\nProcessing (non-thermal)\n\n\nOther usages\n\n\nPackaging\n\n\nTomato Product\n\n\nReferences\n\n\nPre-processing\n\n\nGiagnacovo et al. (2016)\nFolinas et al. (2017)\n\n\nPaste (36\u00b0 Brix)\nPaste (30\u00b0 Brix)\n\n\n7.9%\n5.2%\n\n\n45.2%\n\n\n31.0%\n\n\n6.9%\n5.8%\n\n\n9.0%\nN/A\n\n\n9.8%\n\n\n79.2%\n\n\nTrueblood et al. (2013)\nManfredi and Vignali (2014)\n\n\nPaste (24-39\u00b0 Brix)/diced tomatoes\n\n\n6%\n\n\n4%\n\n\n45%\n\n\n33%\n\n\n12%\n25.3%\n\n\n(47.4%) a\n\n\nPuree (8\u00b0 Brix)\n\n\n4.5%\n\n\n(47.4%)\n16%\n(47.5%)a\n\n\n22.8%\n\n\nAm\u00f3n et al. (2017)\n\n\nPaste (29\u00b0 Brix)/diced tomato\n\n\n29%\n\n\n30%\n\n\n19%\n22.8%\n49%\n\n\n7%\n\n\nPeeled tomato\nPeeled tomato\n\n\nGarofalo et al. (2017)\n\n\n(47.5%) a\n30%\n\n\n25.3%\n\n\n4.5%\n\n\nArnal et al. (2018)\n\n\n21%\n\n\nN/A\n\n\nN/A\n\n\nPre-processing: unloading, washing, and sorting.\n\n\nProcessing (thermal): steam peeling, evaporation, CB/HB, sterilization, and boilers.\n\n\nProcessing (non-thermal): chopping, optical sorting, juice extraction, holding, refinement, filtration, pump sealing, cooling tower, and facility cleaning.\nPackaging: filling and closing, labeling, and palletizing.\n\n\nOther usage: lighting, water treatment, and auxiliary process.\nN/A means Not Available.\na It includes both thermal and non-thermal processing data.\n\n\nusing different methodologies is challenging. However, based on the\nresults presented in Table 2, it can be observed that the quantity and\ndistribution of water, thermal and electrical energy usage in the tomato\nprocessing facility primarily depend on the type of final product (peeled\nor concentrate tomato). After the initial washing and sorting stages, the\nprocessing lines for these products differ significantly, as illustrated in\nFig. 1.\nSpecifically, the use of water is unevenly distributed across the\nprocessing units. The majority of the water is pumped to the pre-\nprocessing steps, such as unloading, washing, sorting, and conveying\ntomatoes into the facility. The remaining portion of the total water is\nprimarily used as steam in thermal processes and for cooling after\nsterilization treatment.\n\n\nevaporators, tube-in-tube heat exchangers), enabling the recovery and\nreuse of approximately 95% of the condensate in the boilers (Trueblood\net al., 2013). Among thermal operations, evaporation and CB/HB are the\nmost energy-intensive stages during tomato concentrate production,\nwhile steam peeling and sterilization consume the largest amount of\nthermal energy in peeled tomato production. For instance, in the study\nby Giagnacovo et al. (2016), the energy-intensive stages of a triple\nconcentrate tomato paste processing plant were identified. Evaporation,\nCB/HB, and sterilization accounted for 76.2%, 15.2%, and 8.6% of the\ntotal thermal energy, respectively. Similarly, Folinas et al. (2017) found\nthat in the production of canned double-concentrate tomato paste, the\nmajority of steam consumption occurred during evaporation (63.4%),\nCB/HB (30.9%), and sterilization (5.7%). The slight variation in distri-\nbution observed in the results achieved compared to Giagnacovo et al.\n(2016), can be likely attributed to the lower concentration of solids in\nthe tomato paste product.\nRegarding the production of peeled tomatoes, Garofalo et al. (2017)\nevaluated the distribution of thermal energy in the production line of\ncanned peeled tomatoes mixed with tomato sauce using the LCA\nmethodology. Their results indicated that 49% of the total thermal en-\nergy was consumed during the evaporation stage for sauce production,\nfollowed by 32.4% in the sterilization stage and 18.6% in the peeling\nstage. Using the same methodology, Arnal et al. (2018) assessed the\nthermal energy consumption in the production of peeled tomatoes,\nexcluding energy requirements for tomato sauce production. They found\nthat thermal energy was primarily used in two main steps: steam peeling\n(61%) and sterilization (39%).\nWhile the major energy requirements in large-scale tomato pro-\ncessing plants are thermal, electricity consumption also plays a signifi-\ncant role (Latini et al., 2017). Generally, electrical energy is more evenly\ndistributed throughout the production line compared to water and\nthermal energy. However, certain thermal and non-thermal processes\nconsume more electrical energy than others (Table 2). For example,\nwhen Giagnacovo et al. (2016) evaluated the electricity distribution in a\n\n\nFor instance, Am\u00f3n et al. (2017) conducted a study on water con-\nsumption in a tomato facility using the WEN approach. The facility\nprocessed approximately 90% of the tomatoes into the paste and the\nremaining into diced tomatoes. According to their findings, around 8.3\nmetric tonnes of water were used per metric tonne of product. Out of this\nwater, the majority (67%) was directed to flumes for unloading,\nwashing, sorting, and conveying tomatoes, while 24% was used in the\nsteam utilization system. The remaining 9% of the water was allocated\nto pump sealing, boiler make-up water, and facility cleaning. Similar\nresults were reported by Manfredi and Vignali (2014), who assessed\nwater usage in the processing phases of a tomato puree production line\nusing the LCA methodology. They found that the most water-consuming\nstage was unloading and washing (88%), followed by evaporation, juice\npasteurization, and bottle pasteurization (12%).\nOn the other hand, the tomato processing industry extensively uti-\nlizes steam, primarily in various thermal processing stages such as\nevaporation, sterilization, CB/HB, and peeling. The distribution of steam\ndepends on factors such as raw material characteristics, the type and\nquantity of the end products, equipment type, and operational condi-\ntions. Approximately half of the total steam generated is directed to\nclosed-system, indirect heating operations (e.g., cold or hot break and\n\n\n7\n\n\nE. Eslami et al.\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\ninformation and data gathered during energy audits conducted in Italian\ntomato facilities of similar capacity as part of the EU \"AccelWater (ID:\n958266)\" project. The results, presented in Table 3, highlight the\naverage KPIs for water, thermal, and electrical energy consumption per\nton of final products in triple tomato paste, tomato puree, and peeled\ntomato production lines.\nIt is evident that triple tomato paste processing consumes more en-\nergy and water compared to tomato puree and peeled tomato produc-\ntion. This can be primarily attributed to the high water and energy\nintensity of the thermal processes involved in triple tomato paste pro-\nduction, particularly the evaporation step used to concentrate tomato\njuice from approximately 5\u00b0 Brix to 36-40\u00b0Brix. LCA studies have esti-\nmated the thermal and electrical energy footprints of processing tomato\npaste and diced tomatoes. For instance, Brodt et al. (2013) found that\ntomato paste processing required more energy per unit mass of final\nproducts compared to diced tomatoes (approximately 8 and 2 MJ/kg,\nrespectively). This difference is mainly due to the energy-intensive\nevaporation step in paste production (Karakaya and \u00d6zilgen, 2011),\nwhich requires significant energy due to the high specific heat capacity\nand latent heat of vaporization of water (Am\u00f3n and Simmons, 2017).\nThese findings strongly emphasize the need to identify the main\nwater and energy-consuming stages during tomato processing to estab-\nlish local average KPIs for highly demanding water and energy unit\noperations. Improving efficiency in these stages can lead to significant\nbenefits (Am\u00f3n and Simmons, 2017; Latini et al., 2017).\n\n\ntriple concentrate tomato paste processing plant, they found that the\nevaporation stage accounted for approximately one-third (34%) of the\ntotal electrical energy, followed by juice extraction (16%) and chopping\n(15%) steps. Trueblood et al. (2013) examined electricity consumption\nin different stages of a tomato paste/puree processing plant and iden-\ntified the cooling tower, evaporation, and HB as the most\nelectricity-demanding stages, consuming 17%, 13%, and 13% of the\ntotal electrical energy, respectively. This was primarily attributed to the\nrecirculation of paste/puree in the evaporators and product cooling.\nOther significant consumers included steam boiler combustion blowers\n(7%), boiler feedwater pumps (7%), facility lighting (2%), and air\ncompressors (5%). Regarding the electricity usage distribution in peeled\ntomato processing, Garofalo et al. (2017) found that the in-container\nprocessing stage consumed the highest amount of energy, accounting\nfor 67% of the total electricity usage. The remaining electricity was\ndistributed to a lesser extent between thermal units (23%) and pre-\nliminary stages (10%). Arnal et al. (2018) also investigated the electrical\nenergy consumption in the production of peeled tomatoes and reported\nthat the canning stage consumed the most electricity (49%), followed by\nwashing (21%), sterilization (21%), and peeling (9%). These findings\nare consistent with the results reported by Garofalo et al. (2017). Am\u00f3n\net al. (2017) conducted a systematic study on the distribution of elec-\ntricity usage in a processing line producing paste and diced tomatoes.\nThey found that approximately 53% of the overall electricity used at the\nfacility was consumed in processing water (WEN points), amounting to\n4.4 million kWh. The remaining electrical energy was utilized in\nnon-WEN points for activities such as facility lighting, climate control,\ncompressed air generation, juice extraction, pumping tomato juice and\npaste, and aseptic packing. Pumping operations accounted for the ma-\njority of electrical WEN usage (approximately 81%), while the remain-\ning energy was allocated to power fans, separators, and aerators. Among\nthe non-pumping electrical demands, cooling tower fans required the\nhighest energy consumption at approximately 12.5% of the total elec-\ntrical WEN for the season, followed by boiler furnace fans (5.7%), and to\na lesser extent, aerators, and separators (0.8%).\n\n\n4. Opportunities for water conservation and energy efficiency in\nthe tomato processing industry\n\n\nAs described in the previous sections, tomato processing facilities\nconsist of inherently water and energy-intensive processes and are\nextremely production-oriented, with tomato processors that typically do\nnot have time to optimize the performance of their equipment during the\nshort harvest season for tomatoes (Giagnacovo et al., 2016; Trueblood\net al., 2013). However, there are numerous opportunities for water\nconservation and energy efficiency, which are crucial for enhancing the\nprofitability of tomato processors in the global market and promoting\nthe environmental sustainability of tomato processing (Trueblood et al.,\n2013).\n\n\n3.3. Key Performance Indicators (KPIs) for water and energy\nconsumption in the tomato processing industry\n\n\nThis section provides an overview of recommended conventional and\nunconventional practices and technologies that can be employed to\nachieve water conservation, energy recovery, and efficiency improve-\nments across various stages of tomato processing facilities. These ap-\nproaches encompass strategies such as maintaining and enhancing the\nefficiency of existing systems, implementing water recycling and waste\n\n\nThe assessment of water and energy usage in industrial tomato\nprocessing enables the identification of key processes necessary for\nestablishing baselines for water and energy consumption. These base-\nlines are crucial for conducting benchmarking analyses and developing\nrelevant Key Performance Indicators (KPIs) (Latini et al., 2017; Peterson\net al., 2022).\nIn the specific subsector of tomato processing, average KPIs can be\nsimply derived from the water, gas, and electricity bills, normalized by\nthe total production per tomato season (Giagnacovo et al., 2016; Latini\net al., 2017). Furthermore, to elucidate the intricate relationship be-\ntween energy and water across diverse process zones (WEN points), a\nlocal water-energy intensity KPI can be calculated. This involves\ndividing the energy consumption by the specific amount of water that\ntraverses a given WEN point. This metric contributes to contextualizing\neach WEN point within the overall process, as it offers insight into how\nenergy is being embedded into the water flowing through a given WEN\npoint (Peterson et al., 2022). It must be underlined that the water-energy\nintensity KPI cannot be calculated for non-WEN processes because\nalthough these processes consume either water or energy, energy is not\nbeing embedded into the water (Peterson et al., 2022).\nThese indicators facilitate the comparison of performance between\ndifferent tomato processing lines or analogous lines within separate fa-\ncilities. Moreover, their computation assumes pivotal importance in\nquantifying efficiency enhancements over time within the same pro-\ncessing plant.\nIn this review, average KPIs were determined through a compre-\nhensive literature review and, when necessary, estimated based on\n\n\nTable 3\n\n\nKPIs in triple tomato concentrate and peeled tomato production lines.\n\n\nAverage KPIs values\n\n\nKPI name\n\n\nReferences\n\n\nPeeled\n\n\nTomato\n\n\nTomato\npuree (8\n\u00b0Brix)\n\n\ntomato\n\n\npaste\n(36-40\nBrix)\n\n\nThermal energy\nconsumption per\nton of tomato\nproducts (kWh/\nton)\nElectrical energy\nconsumption per\nton of tomato\nproducts (kWh/\nton)\nWater consumption\nper ton of tomato\nproducts (m\u00b3/\nton)\n\n\n(Am\u00f3n et al., 2017;\nArnal et al., 2018;\nGiagnacovo et al.,\n2016; Latini et al.,\n2017)\n(Am\u00f3n et al., 2017;\nArnal et al., 2018;\nGiagnacovo et al.,\n2016; Latini et al.,\n2017)\n(Am\u00f3n et al., 2017;\nArnal et al., 2018;\nBehzadian et al.,\n2015)\n\n\n355\n\n\n710\n\n\n2340\n\n\n36\n\n\n43\n\n\n103\n\n\n4.6\n\n\n2.0\n\n\n8.3\n\n\n8\n\n\nE. Eslami et al.\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\nlead to energy savings, and vice versa. The potential benefits resulting\nfrom the implementation of these measures can serve as a motivation for\ncompany management to explore opportunities for water and energy\nconservation. However, it is crucial to conduct engineering studies to\nassess the technical and economic feasibility of each measure,\ncomparing their costs with the potential cost savings and estimating the\nexpected payback period (Am\u00f3n et al., 2013; Trueblood et al., 2013) (see\nTable 5).\n\n\nheat recovery, and adopting innovative processing unit operations and\nwaste management practices.\n\n\n4.1. Conventional water conservation and energy efficiency measures\n\n\nTable 4 provides a comprehensive summary of various conventional\nmeasures that can potentially be implemented in tomato processing\nfacilities. These measures are evaluated based on their relative impact in\nterms of water, thermal, and electrical energy savings, as well as the\nreduction in wastewater generation and associated discharge costs. It is\nimportant to note that, in some cases, due to the inherent link between\nwater and energy, implementing water conservation measures can also\n\n\n4.1.1. Water conservation measures\n\n\nTomato processing facilities are known to consume substantial\namounts of water, which is typically pumped from aquifers, used\n\n\nTable 4\n\n\nSummary of typical water and energy conservation measures in tomato processing facilities, including their impact on freshwater, thermal and electrical energy\nsavings, and wastewater generation.\n\n\nMeasure description\n\n\nImpact after implementation\n\n\nReferences\n\n\nRecourse\n\n\nElectricity\n\n\nDischarge cost\nReduced wastewater\ngeneration\nReduced wastewater\ngeneration\nReduced wastewater\ngeneration\nReduced wastewater\n\n\nFreshwater\n\n\nFuel\n\n\nRepairing water leaks\n\n\nReduced consumption\n\n\nReduced pumping well\n\n\nTrueblood et al. (2013)\n\n\nWater\n\n\nwater\n\n\nTrueblood et al. (2013)\n\n\nReduced consumption\n\n\nReduced pumping well\n\n\nPreventing overflow of cooling\ntower water\nReusing flume water in former\nstages\n\n\nwater\n\n\nReduced fresh makeup\nwater in the flume\nReduced fresh makeup\nwater in the flume\nReduced boiler makeup\nwater and blow down loss\n\n\nTrueblood et al. (2013)\n\n\nReduced pumping well\n\n\nwater\n\n\nTrueblood et al. (2013)\n\n\nReusing single-pass cooling\n\n\nReduced pumping well\n\n\ngeneration\n\n\nwater\n\n\nwater\n\n\nRecycling steam condensate\nfrom indirect heat exchangers\n\n\nReduced pumping well\nwater and cooling tower\nfans use\nReduced pumping of well\nwater and wastewater,\nand cooling tower fan use\n\n\nReduced fuel\n\n\nReduced wastewater\ngeneration\n\n\n(Behzadian et al., 2015;\nTrueblood et al., 2013)\n\n\nuse\n\n\nRecycling of tomato water\ncondensate\n\n\nReduced usage of fresh\nmakeup water in the flume,\nseal water pump floor,\nwashing\nReduced boiler makeup\nwater and blow down loss\n\n\nReduced fuel\n\n\nReduced wastewater\ngeneration\n\n\n(Am\u00f3n et al., 2013;\nTrueblood et al., 2013)\n\n\nuse\n\n\nRepairing steam leaks\n\n\nNatural gas\n(Steam)\n\n\nReduced use of the RO\nsystem\nReduced pumping and\nblower use\n\n\nReduced fuel\n\n\nTrueblood et al. (2013)\n\n\nReduced wastewater\ngeneration\n\n\nuse\n\n\nReducing the operating\npressure of the boilers\nReturning condensate from\nthermal units\nInstalling economizers,\nblowdown heat exchangers,\nand improving combustion\nefficiency\nControlling fouling on heat\nexchangers\n\n\nTrueblood et al. (2013)\n\n\nReduced fuel\n\n\nuse\n\n\nReduced consumption\n\n\nReduced fuel\nuse\nReduced fuel\n\n\nReduced pumping and\ncooling tower fan use\n\n\n(Am\u00f3n et al., 2017;\nTrueblood et al., 2013)\n(Am\u00f3n et al., 2017;\nTrueblood et al., 2013)\n\n\nReduced wastewater\ngeneration\n\n\nuse\n\n\nReduced boiler makeup\nwater and blow down loss\n\n\nReduced pumping and\nblower use\n\n\nReduced fuel\n\n\nReduced wastewater\ngeneration, and\nproduct wastage\n\n\nBalasubramanian and\nPuri (2009)\n\n\nuse\n\n\nReduced pumping and\nblower use\n\n\nTrueblood et al. (2013)\n\n\nReduced boiler makeup\nwater and blow down loss\n\n\nInsulation of equipment,\ncondensate tanks, steam, and\ncondensate pipelines\nWaste heat recovery from\ncondensate effluent (tomato\nwater condensate)\nInstalling mechanical vapor\nrecompression (MVR) systems\nor additional evaporation\nstages\nAssessing pumping efficiency\nRepair and replace pumps to\nimprove energy efficiency\nInstalling VFDs on pumps\n\n\nReduced fuel\n\n\nuse\n\n\nReduced boiler makeup\nwater and blow down loss\n\n\nReduced pumping of well\nwater and wastewater,\nand cooling tower fan use\nReduced pumping of well\nwater and wastewater,\nand cooling tower fan use\n\n\n(Am\u00f3n et al., 2013;\nAm\u00f3n et al., 2015)\n\n\nReduced wastewater\ngeneration\n\n\nReduced fuel\n\n\nuse\n\n\nReduced boiler makeup\nwater and blow down loss\n\n\nReduced wastewater\ngeneration\n\n\nLatini et al. (2017)\n\n\nReduced fuel\n\n\nuse\n\n\nReduced consumption\n\n\n(Am\u00f3n et al., 2017;\nAm\u00f3n and Simmons,\n2017)\n\n\nElectricity\n\n\nReduced consumption and\npeak demand\n\n\nTrueblood et al. (2013)\n\n\nRepairing air leaks\n\n\nReduced consumption\nReduced consumption\n\n\nTrueblood et al. (2013)\nTrueblood et al. (2013)\n\n\nSubstituting compressed air\n\n\nwith blower air\n\n\nInstalling VFDs on blowers and\nfans\n\n\nTrueblood et al. (2013)\n\n\nReduced consumption and\n\n\npeak demand\n\n\nReduced consumption and\n\n\nTrueblood et al. (2013)\n\n\nInstalling high-efficiency\n\n\npeak demand\n\n\nlighting and motion sensors\n\n\nReduced consumption\n\n\nReduced consumption\n\n\nLatini et al. (2017)\n\n\nReduced wastewater\ngeneration waste of\nfresh tomatoes\n\n\nReduced\nconsumption\n\n\nAll\n\n\nKeeping input/output\n\n\nbalancing and operating at the\n\n\nhighest capacity\n\n\nVFDs is the abbreviation of Variable Frequency Drives.\n\n\n9\n\n\nE. Eslami et al.\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\nTable 5\n\n\nAdvantages and disadvantages of tomato peeling by advanced technologies.\n\n\nMethod\n\n\nReferences\n\n\nPros\n\n\nCons\n\n\nFast heating and shallow\npenetration depth\n\n\n\u2022 Non uniform heating\n\n\n(Li et al., 2014a, 2014b; Vidyarthi et al., 2019)\n\n\nFlow Gas Divider\n\n\nHigh investment cost\n\n\n\u2022\n\n\n\u2022 Training of workers\n\n\n\u2022 No chemicals\n\n\nFlow Meter\n\n\nPressure\n\n\nOnline\nControl\n\n\n\u2022 No heating medium (water,\nsteam)\n\n\n\u2022 High peelability\n\n\nWatts Meter\n\n\n\u2022 Lower peeling loss\n\n\n\u26ab Firmer product texture\nHigh quality product\n\n\nInfrared peeling\n\n\nInfrared peeling\n\n\n\u2022\n\n\n\u2022 Less environmental impact\n\n\nScale-up\n\n\nHigh peelability\n\n\n(Gao et al., 2018; Kohli et al., 2021; Rock et al., 2010, 2012)\n\n\n\u2022\n\n\n\u2022 Reactor design complexity\n\u2022 Univen peeling in the up-\nscaled units\n\n\n\u2022 Reduced peeling loss\n\u2022 Reduced peeling time\n\n\nTransducer\n\n\nGenerator\n\n\nAmplifier\n\n\n\u2022 Reduced lye concentration\n\u2022 High-quality product\n\u2022 Increased lycopene content\n\u2022 Less environmental impact\n\n\nDisposal of waste effluent\n\n\nProbes\n\n\n\u2022\n\n\n\u2022 High investment cost\n\u2022 Training of workers\n\n\nUltrasound-assisted peeling\nUltrasound-assisted peeling\n\n\n\u2022 Fast heating\n\n\n(Gavahian and Sastry, 2020; Pataro et al., 2014; Rock et al.,\n2012; Wongsa-Ngasri and Sastry, 2016a, 2016b)\n\n\n\u2022 Electrode corrosion\n\n\nComplexity in design and\n\n\n\u2022 Reduced peeling time\n\n\n\u2022\n\n\nAccelerated lye diffusion\n\u2022 Reduced lye concentration\n\n\nprocess control\n\u2022 Need of process\noptimization\n\n\n\u2022\n\n\nElectric Current\n\n\nHigh peelability\n\n\n\u2022\n\n\nPower Supply Data Logger/PC\n\n\nDisposal of waste effluent\n\n\n\u2022 Reduced peeling loss\n\n\n\u2022\n\n\nHigh-quality product\n\n\nHigh investment cost\n\n\n\u2022\n\n\n\u2022\n\n\nOhmic heating-assisted lye peeling\n\n\n\u2022 Less environmental impact\n\n\n\u2022 Training of workers\n\n\nOhmic heating-assisted lye peeling\n\n\n\u2022 Long term reliability of\nPEF generator\n\n\n\u2022 Mild processing conditions\n\n\n(Arnal et al., 2018; Giancaterino and Jaeger, 2023; Pataro and\nFerrari, 2020)\n\n\n\u2022 Easy integration in e\nprocessing plant\n\n\n\u2022 Electrode corrosion\n\u2022 High investment cost\n\u2022 Training of workers\n\n\n\u2022 High peelability\n\n\n\u2022 Reduced peeling loss\n\u2022 High-quality product\n\n\n\u2022 Reduced water and energy\nconsumption\n\u2022 Less environmental impact\n\n\nPulsed Electric Field (PEF)-assisted steam\npeeling\n\n\nPulsed Electric Field (PEF)-assisted steam\npeeling\n\n\ninternally, and then discharged as treated wastewater into sewer sys-\ntems or for land application (Trueblood et al., 2013). As highlighted in\nTable 4, there are various opportunities for water conservation at\ndifferent stages of tomato processing.\nFor instance, simple and cost-effective measures like repairing water\nleaks from valves, hoses, and storage tanks, as well as installing a level\ncontrol system for cooling tower makeup water pumps to prevent\noverflow can significantly reduce total freshwater consumption (True-\nblood et al., 2013). The adoption of water conservation measures in\nclosed-loop systems such as the recovery and filtration of flume water\nfrom the final stage and its reuse in earlier stages can also contribute to\nsubstantial reductions in freshwater usage. Additionally, water used to\ncool down the temperature of tomato products after sterilization can be\nredirected to the cooling tower or flumes to offset the need for fresh\nmakeup water (Trueblood et al., 2013). Implementing all the\nabove-recommended measures could potentially reduce freshwater\nusage by 16% in the industry (Trueblood et al., 2013).\nThe return and reuse of condensate in boilers, particularly from in-\ndirect heat exchangers, can save freshwater and reduce boiler makeup\nwater treatment costs, blowdown losses, and fuel consumption due to\nthermal energy recovery (Behzadian et al., 2015; Trueblood et al.,\n2013).\nThe large volume of tomato water condensate generated during paste\nproduction has the potential not only for waste heat recovery, as will be\ndiscussed later, but also for water recovery and reuse in applications like\ncooling towers, flumes, seal water pumps, and floor washing (Am\u00f3n\net al., 2013; Trueblood et al., 2013). A WEN assessment estimated that a\nfacility with a capacity of 7000 tons of tomatoes per day could theo-\nretically produce 129,232,774 gallons of tomato water per season, of\n\n\nwhich around 70 million gallons could be technically recovered. This\nrecovery had the potential to reduce electricity consumption for well\nwater and wastewater pumping systems, as well as cooling tower fans,\nby 442,600 kWh and generate over 40,000 MMBtu of energy (Am\u00f3n\net al., 2013). This is because each cubic meter of recovered tomato water\ncorresponds to one less cubic meter pumped from wells, cooled in\ncooling towers, or discharged as wastewater. However, engineering\nstudies are necessary to assess the technical and economic feasibility of\nrecycling and utilizing tomato water in new applications.\n\n\n4.1.2. Thermal energy conservation and efficiency improvement measures\nTomato processing facilities rely heavily on thermal energy for\nvarious direct and indirect heat exchange processes. Therefore, there are\nseveral opportunities outlined in Table 4 to improve energy efficiency,\nconserve energy, and recover waste heat from boilers and thermal\nprocessing units.\nTypically, boilers produce steam at much higher pressures (>10 bar)\nthan required by the thermal processes in a tomato facility. Therefore,\nsimple measures like adjusting boiler pressure set points can reduce\nnatural gas consumption (Trueblood et al., 2013). Another significant\nsaving in natural gas can be achieved through conventional waste heat\nrecovery methods for boilers. These methods include returning\ncondensate from indirect heat exchangers, installing economizers and\nblowdown heat exchangers to pre-heat feed water, and improving\ncombustion efficiencies (Am\u00f3n et al., 2017; Trueblood et al., 2013).\nFurthermore, controlling fouling on heat exchanger surfaces is\ncrucial for enhancing energy efficiency and reducing the need for\nfrequent cleaning, which can cause process interruptions. The use of\nlow-friction, food-grade coatings specifically designed for heat\n\n\n10\n\n\nE. Eslami et al.\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\nexchangers can effectively minimize fouling and its negative impact\n(Balasubramanian and Puri, 2009).\nAddressing steam leaks is a highly cost-effective method for\nachieving substantial energy conservation. This is because the water lost\nthrough a steam leak necessitates the introduction of new treated water,\nleading to additional electrical energy usage in the RO (Reverse\nOsmosis) system and increased fuel consumption in the boiler (Peterson\net al., 2022; Trueblood et al., 2013).\nFurthermore, given the extensive use of steam in tomato processing,\napplying insulation materials such as fiberglass blankets to uncovered\nsurfaces of equipment, product, and condensate tanks, as well as steam\nand condensate pipelines can result in additional natural gas savings for\nsteam boilers (Trueblood et al., 2013).\nWaste heat can be also recovered from various process effluents\nother than those recycled to boilers from indirect heat exchangers.\nHowever, the feasibility of heat recovery from these streams depends on\nfactors such as temperature, quantity, purity, and availability of waste\nheat-containing streams, as well as the associated recovery costs (Am\u00f3n\nand Simmons, 2017). In tomato processing, a significant source of waste\nheat is tomato water condensate, which exits the evaporator at tem-\nperatures ranging from 55 to 85 \u00b0C (Meneses et al., 2019). Typically, this\nlow-grade waste heat is dissipated in cooling towers before being dis-\ncharged (Am\u00f3n et al., 2015). However, due to the relatively clean nature\nof this effluent, it can be considered for reuse in other parts of the pro-\ncessing facility, such as in flumes, once appropriately cooled (Am\u00f3n\net al., 2013; Am\u00f3n et al., 2015). Researchers have explored heat re-\ncovery from tomato water condensate, by pre-heating crushed tomatoes\nentering the hot break stage, achieving substantial energy savings, ac-\ncounting for approximately 3.7% of the total seasonal energy usage. The\nmajority of the savings (over 95%) resulted from reduced natural gas\nusage at the boiler, while the remaining portion came from a reduced\nload on cooling towers, groundwater pumps, and wastewater processes\n(Am\u00f3n et al., 2015). It is worth noting that the technical and economic\nfeasibility of this measure should consider the costs associated with\nusing an additional heat exchanger upstream of the one used for the\nsteam-heated hot break, making it a capital-intensive measure.\nInstalling mechanical vapor recompression (MVR) systems or addi-\ntional evaporation stages on the evaporator is another opportunity for\nwaste heat recovery and reducing natural gas consumption (Latini et al.,\n2017). The MVR evaporator is the most efficient and capital-intensive,\nwhich recompresses steam from the evaporated tomato paste and re-\ndirects it to earlier stages in the evaporator (Trueblood et al., 2013). The\nsteam economy of MVR systems can reach up to 20 units of water\nevaporated from tomatoes for every unit of steam input. Implementing\nadditional effects in the evaporator, usually 2 to 5 effects in a\nmultiple-effect evaporator design, is also a capital-intensive measure. In\nthis design, each effect operates at a lower pressure than the previous\nstage, allowing the evaporated water from tomatoes to serve as a ther-\nmal energy source for the next effect (Latini et al., 2017; Trueblood\net al., 2013). These approaches can lead to significant energy savings,\nwith the ideal steam economies for a multiple n-effects evaporator\nranging from one unit of steam boiler evaporating n units of water from\ntomatoes (Trueblood et al., 2013).\n\n\npumping systems conducted at an industrial tomato processing facility\nrevealed an overall efficiency of 53.6%, lower than the expected effi-\nciency of well-functioning centrifugal pumps, which should be at least\n65% (Am\u00f3n et al., 2013) These findings highlight the importance of\nassessing pump efficiency for individual facilities and taking steps such\nas repairing and replacing pumps to improve energy efficiency. Addi-\ntionally, many pumps are oversized and throttled or bypassed to control\nflow and pressure. Installing Variable Frequency Drives (VFDs) and\npressure or level sensors on these pumps can yield significant electrical\nenergy savings (up to about 80%) and reduce peak demand (Trueblood\net al., 2013).\n\n\nElectricity also powers compressors that are used to provide com-\npressed air for various operations in tomato processing, including\ndriving diaphragm pumps, controlling valves, operating pneumatic\ntools, and handling aspects of packaging and labeling (Am\u00f3n and Sim-\nmons, 2017). Assessments of several industrial tomato processing fa-\ncilities have identified energy efficiency opportunities in compressed air\nsystems (Am\u00f3n et al., 2013). For example, implementing cost-effective\nmeasures such as reducing compressed air pressure to the minimum\nrequired and establishing a regular repair program for air leaks can\nenhance efficiency and significantly reduce compressor energy con-\nsumption up to 10% (Trueblood et al., 2013). Furthermore, replacing\ncompressed air with blower air in specific applications, such as package\nflattening, drying, or mechanical conveyance, where high-pressure\nblower air is a viable and efficient alternative, can further reduce en-\nergy usage (Trueblood et al., 2013).\nElectricity also powers motors in boiler furnaces blowers and cooling\ntower fans. Similar to pumps, installing VFDs on blowers and fans can\nyield substantial energy savings. Studies have demonstrated that\nimplementing this measure can save 33%-44% of combustion blower\nenergy consumption, and 42%-63% of cooling tower fan energy con-\nsumption (Trueblood et al., 2013).\nFinally, replacing inefficient lamps with energy-efficient lighting and\nusing motion and daylight sensors in unoccupied areas contribute to\nenergy savings and peak demand reduction (Trueblood et al., 2013).\n\n\n4.1.4. Other practices to save water and energy and reduce waste\ngeneration in tomato processing industry\n\n\nTo reduce water and energy consumption, as well as waste genera-\ntion in a medium to medium-large tomato processing plant handling\nhundreds of tons of fresh tomatoes daily, it is crucial to maintain a\ncontinuous operation of the processing lines and avoid operating below\nthe maximum capacity or intermittently (Latini et al., 2017). Processing\nequipment, in fact, operates most efficiently when it can run continu-\nously with minimal starts and stops (Brodt et al., 2013). For this reason,\neffective management of fruit harvesting and delivery is essential to\nensure a consistent and uninterrupted supply of fresh tomatoes at\nmaximum capacity throughout the processing season (Latini et al.,\n2017).\nFurthermore, it is important to minimize unplanned manufacturing\nprocess stops caused by events such as motor failures, material issues,\noperator shortages, or unscheduled maintenance (Giagnacovo et al.,\n2016). Every time the tomato processing line is shut down, machines\nneed to be thoroughly cleaned, resulting in the loss of several working\nhours, significant water and energy consumption, and waste of fresh\ntomatoes waiting in trucks outside the facility at temperatures that can\nexceed 30 \u00b0C, or tomatoes at various stages of processing, particularly in\nthe evaporators.\n\n\n4.1.3. Electrical energy conservation measures\n\n\nTomato processors are large consumers of electrical energy with a\nvery high electrical peak demand concentrated in a short period. How-\never, there are numerous opportunities to save energy throughout the\ntomato processing stages.\nThe primary use of electrical energy in tomato processing is for\npowering pumps, which rely on electric motors to convey products and\ntransport water (Am\u00f3n et al., 2017). Therefore, improving pumping\nefficiency is crucial for reducing electricity consumption in tomato\nprocessing. Factors such as flow rate, head, and the condition of\npumping systems can significantly impact efficiency (Am\u00f3n et al., 2017;\nAm\u00f3n and Simmons, 2017). A comprehensive assessment of water\n\n\n4.2. Unconventional water conservation and energy efficiency\ntechnologies\n\n\nThe tomato processing industry is currently focused on reducing\nwater usage, improving energy efficiency, and preserving the quality\nand health benefits of fresh tomatoes. In addition to conventional\nmeasures and technologies, advanced thermal and non-thermal\n\n\n11\n\n\nE. Eslami et al.\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\ntechnologies are being explored as sustainable and innovative alterna-\ntives for tomato processing. These technologies include high-pressure\nprocessing (HPP), pulsed electric field (PEF), infrared radiation (IR),\nohmic heating (OH), and ultrasound (US), among others. They have\ngained attention from researchers and food processors as they offer\npromising solutions for saving energy in evaporation, enzyme and mi-\ncrobial inactivation processes, and peeling operations while maintaining\ntomato quality and health properties. The following sections will pro-\nvide examples of how these novel technologies can be applied at\ndifferent stages of tomato processing.\n\n\nprocessing can improve product quality (Am\u00f3n and Simmons, 2017).\nTraditional thermal processes can lead to the loss of essential antioxi-\ndants like lycopene and \u1e9e-carotene (Seybold et al., 2004). To preserve\nthese crucial nutrients, mild and energy-efficient processes can be in-\ntegrated at specific stages of tomato processing. For example, tomato\npur\u00e9e processed with HPP exhibited higher retention of anti-radical\npower, ascorbic acid, and total carotenoids compared to thermally\nprocessed one. HPP-treated tomato pur\u00e9e also had higher levels of ca-\nrotenoids compared to unprocessed tomatoes (Patras et al., 2009).\nSimilarly, using PEF for processing tomato juice resulted in enhanced\navailability of certain nutrients, such as carotenoids, in the final prod-\nucts (Odriozola-Serrano et al., 2009).\nFurther research is necessary to scale up and accurately assess the\nwater and energy-saving potential of specific emerging technologies in\ntomato processing. Additionally, a careful optimization of process pa-\nrameters and equipment design is necessary in order to ensure the\ndesired degree of microbial or enzymes with the minimum expenditure\nof energy without overprocessing the food product. The final objective is\nto minimize or replace traditional heating methods while also ensuring\nthat these emerging technologies do not compromise, and ideally\nenhance, the quality of the final products compared to conventional\nthermal methods.\n\n\n4.2.1. Innovative technologies for enzyme and microbial inactivation\n\n\nThermal processes used in tomato processing, such as CB/HB and\nsterilization, consume a significant amount of water and energy and can\nhave a negative impact on product quality. As a result, there has been a\ngrowing interest in the past two decades to explore advanced technol-\nogies that offer water and energy savings and improved product quality\ncompared to traditional thermal processes (Pereira and Vicente, 2010).\nHPP, PEF, US, and OH are among the technologies that have shown\ngreat promise as mild and energy-efficient alternatives for producing\nsafe and high-quality tomato products (Pereira and Vicente, 2010;\nRathnakumar et al., 2023). For example, HPP utilizes intense hydro-\nstatic pressures (100-1000 MPa) to denature proteins and induce mi-\ncrobial death. Studies have demonstrated successful sterilization of\ntomato pur\u00e9e using HPP at 700 MPa and 20 \u00b0C, resulting in a reduction\nof viable microorganisms to undetectable levels (Krebbers et al., 2003).\nPEF involves subjecting a food product placed in contact with two\nconductive electrodes to a series of short (1-10 \u00b5s) electric pulses of high\nintensity (10-40 kV/cm) and energy input (50-150 kJ/kg), which re-\nsults in the permeabilization of the cell membrane by electroporation, as\nwell as disruption of intramolecular protein interactions, leading to\nmicrobial and enzyme inactivation (Raso et al., 2016; Shams et al.,\n2023). The technique has been employed to inhibit pectin methyl-\nesterase extracted from tomatoes, achieving a 93.8% reduction in\nenzyme activity (Giner et al., 2000). Subsequently, commercial-scale\ndemonstrations have showcased the potential of replacing traditional\nthermal hot break processes with PEF treatment (Jayathunge et al.,\n2019).\nUS treatment has also shown significant potential for microbial and\nenzyme inactivation in foods (Lauteri et al., 2023). By applying pressure\nwaves (16-100 kHz) to the food material, cavitation and turbulence are\ngenerated, which disrupt microorganisms and enzymes (Rathnakumar\net al., 2023). Many studies have demonstrated that ultrasound pro-\ncessing can effectively reduce pectin-degrading enzyme activity in to-\nmato juice, comparable to or even exceeding thermal hot break methods\n(Terefe et al., 2009; Wu et al., 2008). Ultrasonic treatment has also\nachieved a 5-log reduction in viable yeast in tomato juice (Adekunte\net al., 2010). These studies highlight the potential to decrease or elim-\ninate the need for heating during hot or cold break processes, as well as\nsterilization.\nOH is an alternative to traditional indirect thermal methods for\nevaporating, blanching, and sterilizing food products (Guida et al.,\n2013; Pataro et al., 2011). It involves passing alternating electrical\ncurrent (50 Hz - 100 kHz) through food placed between two electrodes\ngenerating internal heat due to the food's electrical resistance (Junqua\net al., 2021). OH offers rapid and uniform heating of materials, including\nviscous and particulate foods, and it ensures efficient energy transfer\n(Pereira et al., 2016). It also prevents fouling of heat exchanger surfaces,\nincreasing energy efficiency (Pereira and Vicente, 2010). Tomato paste,\nwith its high conductivity, is well-suited for ohmic heating (Darvishi\net al., 2012) allowing for effective moisture removal (Torkian Boldaji\net al., 2015), and enzyme inactivation (Yildiz and Baysal, 2006). Studies\nhave shown that OH can also inactivate harmful microorganisms in to-\nmato juice (Lee et al., 2012; Somavat et al., 2013; Yildiz and Baysal,\n2006).\n\n\n4.2.2. Innovative methods for tomato peeling\n\n\nPeeling is a crucial process in food processing to efficiently produce\nhigh-quality products (Kohli et al., 2021). The performance of peeling\nmethods is assessed based on factors like peelability, peeling loss, ease of\npeeling, and product quality (Li et al., 2014a; Pan et al., 2009). Concerns\nregarding water and energy consumption as well as environmental\nimpact are also important considerations (Arnal et al., 2018).\nChemical and steam peeling methods have been widely used in the\ntomato processing industry. Chemical peeling involves immersing to-\nmatoes in a hot caustic solution (usually sodium hydroxide, 8%-25%, at\ntemperatures of 85\u2013100 \u00b0C for 15\u201360 s), which effectively removes the\nskin, but poses challenges such as high water and energy consumption\nand disposal of peeling effluent (Arnal et al., 2018; Pan et al., 2009; Rock\net al., 2012). Steam peeling weakens the tomato skin using pressurized\nsteam (50-200 kPa, for 10\u201360 s), but it may result in inferior peelability,\nhigher peeling loss, and reduced firmness compared to chemical peeling,\nwhile it is also water and energy-intensive (Arnal et al., 2018; Rock\net al., 2012).\n\n\nTo address these issues, sustainable and non-chemical peeling al-\nternatives using innovative technologies like IR heating, OH, US, and\nPEF, have been developed (Andreou et al., 2020; Gao et al., 2018;\nGavahian and Sastry, 2020; Giancaterino and Jaeger, 2023; Kohli et al.,\n2021; Li et al., 2014a, 2014b; Rock et al., 2012; Vidyarthi et al., 2019).\nHowever, their industrial implementation has been limited so far due to\nhigh investment costs and low processing capacities, among others.\n\n\n4.2.2.1. Infrared (IR) peeling. IR peeling is an innovative and sustain-\nable dry-peeling method that eliminates the need for chemicals and\nheating mediums like water or steam in the peeling process. It effectively\nreduces product loss and maintains product quality (Li and Pan, 2014a,\n2014b; Pan et al., 2009; Vidyarthi, 2017; Vidyarthi et al., 2019). By\nrapidly heating the surface of tomatoes using electric, ceramic or the\nmore advanced catalytic IR generator, physical and biochemical changes\noccur in the peel, facilitating easy detachment (Li et al., 2014a; Qu et al.,\n2022; Vidyarthi et al., 2019). IR radiation has a shallow penetration\ndepth, resulting in minimal alterations to the texture and nutrient con-\ntent of the inner part of the fruit (Vidyarthi et al., 2019). Tests conducted\nboth at the bench scale and pilot scale have shown that tomatoes peeled\nusing IR technology have greater firmness and lower peeling loss\ncompared to lye peeling methods while consuming less energy (Li et al.,\n2014a, 2014b; Vidyarthi et al., 2019).\n\n\nHowever, achieving optimal peeling performance with IR technology\n\n\nIncorporating advanced energy-efficient technologies in tomato\n\n\n12\n\n\nE. Eslami et al.\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\ndepends on crucial parameters like tomato surface temperature and\nheating rate (Vidyarthi et al., 2019). Uniform heating can be a challenge,\nand careful optimization of process parameters and equipment design is\nnecessary (Pan et al., 2009). The initial investment cost for IR peeling\nequipment is high, but the long-term benefits may justify it.\n\n\n4.2.2.4. Pulse electric field (PEF)-assisted steam peeling. PEF-assisted\nsteam peeling offers a promising and efficient technological solution for\ntomato peeling, enhancing the ease of peel removal, while saving energy\n(Arnal et al., 2018; Giancaterino and Jaeger, 2023). By applying a\nmoderate electric field intensity (E < 5 kV/cm) and a relatively low\nenergy input (WT < 5 kJ/kg), structural modifications occur within the\ntomato's matrix, reducing the surface resistance of the skin and pro-\nmoting detachment from the flesh (Andreou et al., 2020; Arnal et al.,\n2018; Giancaterino and Jaeger, 2023; Koch et al., 2022). Moreover, the\napplication of an external electric field induces electroporation effects\n(Pataro et al., 2018), enhancing water mass transfer and increasing\nwater availability under the tomato skin compared to untreated to-\nmatoes (Arnal et al., 2018; Pataro et al., 2018). During subsequent steam\nheating, the greater pressure difference across the tomato skin, caused\nby vaporization, facilitates the formation of cracks, which aids in me-\nchanical peel removal using pinch roller systems (Arnal et al., 2018).\nThis results in reduced peeling loss, high-quality products, and reduced\nwater and steam usage compared to traditional steam peeling methods\n(Andreou et al., 2020; Arnal et al., 2018; Giancaterino and Jaeger,\n2023). The successful implementation of PEF-assisted steam peeling in\nexisting industrial plants has demonstrated its feasibility and positive\nenvironmental impact (Arnal et al., 2018; Pataro et al., 2018). In the\ncontext of the EU project \u201cFieldFood\" (635632-FieldFOOD-H2020),\nspecific industrial tests were conducted. These tests demonstrated that\nutilizing a relatively low-intensity pulsed electric field (PEF)\npre-treatment at values of 0.45 kV/cm and 0.40 kJ/kg on tomato fruits\nbefore steam peeling led to a notable reduction of up to 20% in the total\nsteam required during the thermo-physical peeling process. Addition-\nally, a LCA study revealed that integrating PEF technology prior to steam\npeeling resulted in significant enhancements across all measured envi-\nronmental indicators, with improvements ranging from 17% to 20%,\nthus suggesting that PEF is an environmentally friendly technology\n(Arnal et al., 2018).\nHowever, further research is required at an industrial scale to vali-\ndate energy savings and address technological challenges, including the\nlong-term reliability of PEF generators and electrodes, as well as high\ninitial investments (Pataro and Ferrari, 2020) before the widespread\nadoption and exploitation of PEF technology can be realized.\nIn conclusion, upscaling as well as optimization of process parame-\nters and refining equipment design for novel technologies applied to\nmicrobial/enzyme inactivation and the peeling of fruits and vegetables\nis a pivotal stride. This aims to secure the intended process outcomes\n(microbial/enzyme control and high peelability) with the minimum\nexpenditure of water and energy and avoiding excessive alteration of the\nfood product, thus reducing reliance on traditional heating and chemical\napproaches. All of these achievements go into direction to improve\nsustainability and foster cleaner production.\n\n\n4.2.2.2. Ultrasound (US)-assisted peeling. US-assisted peeling utilizes\nhigh-intensity sound waves (20-100 kHz) to generate a cavitation effect,\nleading to the degradation of the tomato skin and structural carbohy-\ndrates. This weakens the skin, resulting in the separation of the epicarp\nfrom the pericarp. The cavitation also generates free radicals aiding in\nthe chemical breakdown of carbohydrates and facilitating the peeling\nprocess (Rock et al., 2012).\nInitial studies have demonstrated that using power US in hot water\nyields better peeling performance compared to conventional lye peeling.\nHigher temperatures combined with ultrasound provide better peeling\nscores and lower losses. Applying US directly in the lye solution further\nminimizes peeling losses while reducing the lye concentration (Rock\net al., 2012). A cascade approach combining hot lye and US has been\nalso found to reduce the concentration and processing time of hot lye\nwhile increasing the yield and lycopene content of peeled tomatoes (Gao\net al., 2018).\nHowever, there are challenges associated with implementing\nultrasound-assisted peeling on a larger scale, such as insufficient power\nintensity, reactor design complexities, and the potential for uneven\npeeling.\nOverall, US-assisted peeling holds promise for enhancing tomato\nprocessing, but further research is required to address these limitations\nand optimize the technique (Gao et al., 2018; Kohli et al., 2021; Rock\net al., 2012).\n\n\n4.2.2.3. Ohmic heating (OH)-assisted lye peeling. The process of peeling\ntomatoes using OH involves immersing them in an electroconductive\nsolution containing sodium hydroxide or sodium chloride. By passing an\nalternating electrical current through the solution, a combination of\nthermal, chemical, and physical mechanisms, along with electrical ef-\nfects, effectively removes the tomato skin (Rock et al., 2012; Wong-\nsa-Ngasri and Sastry, 2016a, 2016b).\nThe current flow causes the solution to heat up, leading to the\ndegradation of the waxy cuticle and the disruption of hemicellulosic and\npectic substances, making the skin less rigid. This, along with increased\ntemperature and water vaporization (Gavahian and Sastry, 2020; Kohli\net al., 2021), facilitates the splitting of the tomato skin and the separa-\ntion of the outer layer from the inner part. As a result, a high peeling\nscore of 4.5-5 out of 5 can be achieved using ohmic heating with a\nrelatively low concentration (0.01-0.03% w/v) of NaCl. It's worth\nnoting that preheating the solution above 40 \u00b0C can shorten the peeling\ntime (Wongsa-Ngasri and Sastry, 2015). This use of OH for tomato\npeeling has the potential to reduce environmental challenges associated\nwith lye peeling methods by utilizing a low-concentration NaCl peeling\nmedium.\n\n\n5. Conclusion and remarks\n\n\nThe food and beverage industrial sector plays a significant role in\nenergy consumption and global water footprints, leading to substantial\nenvironmental impact. Among these industries, tomato processing\nstands out as one of the most resource-intensive, consuming large\namounts of water and energy (both thermal and electrical) while\ngenerating substantial solid and liquid wastes.\nThis review paper introduces a systematic approach based on Water-\nEnergy Nexus (WEN) analysis, which serves to pinpoint the production\nprocess stages with the highest water and energy demands, thereby\nhighlighting areas of inefficiency. This framework empowers decision-\nmakers to implement customized strategies aimed at enhancing both\nefficiency and sustainability in tomato processing. By applying this\napproach, numerous key opportunities for efficiency enhancements\nhave been identified.\n\n\nCombining OH with lye peeling at lower concentrations (0.5-1% w/\nv) than conventional methods can yield high-quality peeled products,\nminimize peeling losses, and accelerate the peeling process (Sawant\net al., 2018; Wongsa-Ngasri and Sastry, 2016b). This is because in\nlye-ohmic peeling the diffusion of lye or NaOH is accelerated by elec-\ntroporation, resulting in faster depolymerization of substances in the\nskin and separation of the peel (Gupta and Sastry, 2018; Rock et al.,\n2012; Wongsa-Ngasri and Sastry, 2015).\nHowever, implementing OH on an industrial scale requires further\nresearch on engineering design, electrode corrosion, economic factors,\nand scaling up the process (Gavahian and Sastry, 2020; Pataro et al.,\n2014). Additionally, the safe disposal of used salt solutions is an\nimportant consideration for the application of this emerging technology\n(Kohli et al., 2021).\n\n\nThe majority of water usage is concentrated in the initial pre-\nprocessing steps, primarily during tomato washing and conveying into\n\n\n13\n\n\nE. Eslami et al.\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\nproject (ID: 958266).\n\n\nthe facility. Additionally, a significant portion of the total water is\nconsumed as steam in thermal processes and cooling operations. It is\nimperative to focus on implementing water conservation measures,\nespecially in a closed-loop system during the initial stages, as well as\nexploring the potential for reusing steam condensate.\nSteam boilers within a tomato processing facility stand out as the\nmost energy-intensive equipment by a considerable margin. Conse-\nquently, any comprehensive energy efficiency audit should prioritize the\nassessment of these boilers. The recovery of waste heat from various\nprocess effluents, through both direct and indirect heat exchange pro-\ncesses, should be a central strategy to achieve substantial energy sav-\nings. The evaporation step, in particular, offers a unique opportunity for\nwaste heat and water recovery. While measures like installing me-\nchanical vapor recompression (MVR) systems or additional evaporation\nstages are capital-intensive, they can lead to significant energy savings.\nElectrical energy is more evenly distributed across the production\nline compared to water and thermal energy. The majority of this elec-\ntrical energy is allocated to pump operations, with the remainder used\nfor powering compressors, fans, separators, and aerators. Therefore,\nenhancing pumping efficiency is crucial for reducing electricity con-\nsumption in tomato processing.\nIn addition to conventional measures and technologies, integrating\nadvanced thermal and non-thermal technologies like Pulsed Electric\nField (PEF), High-Pressure Processing Homogenization (HPPH), Ultra-\nsound (US), Infrared (IR), and Ohmic Heating (OH) shows promise so-\nlutions in saving water, improving energy efficiency, and reducing\nenvironmental impact, all while preserving or enhancing the quality of\nfinal products compared to traditional thermal methods.\nFuture research should focus not only on the technical feasibility but\nalso on the economic viability of implementing conventional and un-\nconventional practices and technologies. This is especially important for\nadvanced technologies, as their integration into tomato processing lines\nrequires a thorough assessment of their water and energy-saving po-\ntential at an industrial scale. Moreover, there are several technological\nchallenges that must be addressed to enable their implementation on a\nlarger scale. These challenges encompass upscaling as well as the need to\nimprove equipment design, optimize process parameters, and manage\nthe significant initial investment costs. Additionally, the seasonal nature\nof tomato production could be a further obstacle to their spread.\nThe approach discussed in this review, specifically tailored to the\ntomato processing industry, can serve as a model for addressing similar\nchallenges in other water and energy-intensive sectors within the food\nindustry. Nevertheless, regardless of the specific food sector, imple-\nmenting alternative practices and technologies necessitates a rigorous\ncomparison with initial baseline levels of water and energy consumption\nat the key processing stages to quantify improvements effectively. In this\nregard, the accurate collection of data regarding water and\nenergy flow\nis of utmost importance. Achieving this may involve the installation of\nsensors and monitoring systems, as well as the utilization of process\nsimulation tools to precisely solve water mass balances and estimate\nflow rates. These measures can significantly assist facilities in enhancing\ntheir resource consumption efficiency.\n\n\nDeclaration of competing interest\n\n\nThe authors declare that they have no known competing financial\ninterests or personal relationships that could have appeared to influence\nthe work reported in this paper.\n\n\nData availability\n\n\nData will be made available on request.\n\n\nAcknowledgement\n\n\nThis work was supported by the research project Accelerating Water\nCircularity in Food and Beverage Industrial Areas around Europe\n(AccelWater, ID: 958266), Horizon 2020 (call H2020-LCCI-2020-\nEASME-singlestage).\n\n\nReferences\n\n\nAccelWater, 2020. Accelerating Water Circularity in Food and Beverage Industrial Areas\nAround Europe. https://www.accelwater.eu/. accessed 6.25.23.\nAdekunte, A.O., Tiwari, B.K., Cullen, P.J., Scannell, A.G.M., O'Donnell, C.P., 2010. Effect\nof sonication on colour, ascorbic acid and yeast inactivation in tomato juice. Food\nChem. 122, 500-507. https://doi.org/10.1016/j.foodchem.2010.01.026.\nAm\u00f3n, R., Simmons, C.W., 2017. Energy use and efficiency opportunities in industrial\ntomato processing. Acta Hortic. 1159, 143\u2013149. https://doi.org/10.17660/\nActaHortic.2017.1159.22.\nAm\u00f3n, R., Maulhardt, M., Wong, T., Kazama, D., 2013. Water Energy Nexus at a\nCalifornia Tomato Processing Facility: Pumping System Assessment. Contract No.\n400-10-013. California Energy Commission), Sacramento, CA, USA.\nAm\u00f3n, R., Maulhardt, M., Wong, T., Kazama, D., Simmons, C.W., 2015. Waste heat and\nwater recovery opportunities in California tomato paste processing. Appl. Therm.\nEng. 78, 525\u2013532. https://doi.org/10.1016/j.applthermaleng.2014.11.081.\nAm\u00f3n, R., Wong, T., Kazama, D., Maulhardt, M., Maulhardt, T., Simmons, C.W., 2017.\nAssessment of the industrial tomato processing water energy nexus: a case study at a\nprocessing facility. J. Ind. Ecol. 1\u201312. https://doi.org/10.1111/jiec.12600, 00.\nAndreou, V., Dimopoulos, G., Dermesonlouoglou, E., Taoukis, P., 2020. Application of\npulsed electric fields to improve product yield and waste valorization in industrial\ntomato processing. J. Food Eng. 270 https://doi.org/10.1016/j.\njfoodeng.2019.109778.\nArnal, \u00c1.J., Royo, P., Pataro, G., Ferrari, G., Ferreira, V.J., L\u00f3pez-Sabir\u00f3n, A.M.,\nFerreira, G.A., 2018. Implementation of PEF treatment at real-scale tomatoes\nprocessing considering LCA methodology as an innovation strategy in the agri-food\nsector. Sustain. Times 10, 1-16. https://doi.org/10.3390/su10040979.\nBalasubramanian, S., Puri, V.M., 2009. Thermal energy savings in pilot-scale plate heat\nexchanger system during product processing using modified surfaces. J. Food Eng.\n91, 608\u2013611. https://doi.org/10.1016/j.jfoodeng.2008.10.014.\nBehzadian, K., Farmani, R., Butler, D., 2015. Water Feasibility Project for Local Nexus\nNetwork of Food, Energy and Water. Centre for Water Systems, University of Exeter.\nBhatt, C., Kumar, N., Bashir, A.K., Alazab, M., 2022. Implementing data analytics and\narchitectures for next generation wireless communications. IGI Global. https://doi.\norg/10.4018/978-1-7998-6988-7.\nBorzooei, S., Campo, G., Cerutti, A., Meucci, L., Panepinto, D., Ravina, M., Riggio, V.,\nRuffino, B., Scibilia, G., Zanetti, M., 2020. Feasibility analysis for reduction of carbon\nfootprint in a wastewater treatment plant. J. Clean. Prod. 271 https://doi.org/\n10.1016/j.jclepro.2020.122526.\nBrodt, S., Kramer, K.J., Kendall, A., Feenstra, G., 2013. Comparing environmental\nimpacts of regional and national-scale food supply chains: a case study of processed\ntomatoes. Food Pol. 42, 106\u2013114. https://doi.org/10.1016/j.foodpol.2013.07.004.\nDarvishi, H., Hosainpour, A., Nargesi, F., 2012. Ohmic heating behaviour and electrical\nconductivity of tomato paste. J. Nutr. Food Sci. https://doi.org/10.4172/2155-\n9600.1000167, 02.\nDe Meo, E., Nardone, G., Bimbo, F., Carlucci, D.A., 2022. Hedonic analysis of processed\ntomato prices using Italian regional markets data. Foods 11, 816. https://doi.org/\n10.3390/foods11060816.\nDiGiacomo, R., 2011. Measuring flow: understanding how flowmeters work and the pros\nand cons of each type of device can help in selecting the right one. Chem. Eng. 118\n(6), 30-35.\nEslami, E., Carpentieri, S., Pataro, G., Ferrari, G., 2023. A comprehensive overview of\ntomato processing by-product valorization by conventional methods versus\nemerging technologies. Foods 12. https://doi.org/10.3390/foods12010166.\nFAO, 2011. Energy-smart food for people and climate. In: Issue Paper, vol. 66. Food\nAgric. Organ. United Nations.\nFAO, 2017. The Future of Food and Agriculture: Trends and Challenges. https://www.\nfao.org/3/16881E/16881e.pdf. accessed 6.24.23.\nFolinas, D., Karayannakidis, P., Aidonis, D., 2017. Making the canned tomato paste\nproduction green. In: The 8th International Conference on Information and\nCommunication Technologies in Agriculture. Food and Environment, pp. 759-771.\n\n\nCRediT authorship contribution statement\n\n\nElham Eslami: Conceptualization, Investigation, Data curation,\nFigure design, Writing - original draft. Emad Abdurrahman: Investi-\ngation, Data curation, Figure design, Writing \u00b7 original draft. Gian-\npiero. Pataro: Conceptualization, Supervision, Writing\nreview &\nediting. Giovanna. Ferrari: Supervision, Conceptualization, Writing -\nreview and editing, Funding acquisition, Project administration,\nResources.\n\n\nFunding\n\n\nThis work research has been financially supported by AccelWater\n\n\n14\n\n\nE. Eslami et al.\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\nGao, R., Ye, F., Lu, Z., Wang, J., Li Shen, X., Zhao, G., 2018. A novel two-step ultrasound\npost-assisted lye peeling regime for tomatoes: reducing pollution while improving\nproduct yield and quality. Ultrason. Sonochem. 45, 267\u2013278. https://doi.org/\n10.1016/j.ultsonch.2018.03.021.\nGarofalo, P., D'Andrea, L., Tomaiuolo, M., Venezia, A., Castrignan\u00f2, A., 2017.\nEnvironmental sustainability of agri-food supply chains in Italy: the case of the\nwhole-peeled tomato production under life cycle assessment methodology. J. Food\nEng. 200, 1-12. https://doi.org/10.1016/j.jfoodeng.2016.12.007.\nGavahian, M., Sastry, S.K., 2020. Ohmic-assisted peeling of fruits: understanding the\nmechanisms involved, effective parameters, and prospective applications in the food\nindustry. Trends Food Sci. Technol. 106, 345\u2013354. https://doi.org/10.1016/j.\ntifs.2020.10.027.\nGiagnacovo, G., Latini, A., Alberto Campiotti, C., Viola, C., Scoccianti, M., 2016. D.3.6\nExtended Value Stream Maps of NACE 10.3: Fruit Juices and Purees and Tomato\nConcentrates.\nGiancaterino, M., Jaeger, H., 2023. Impact of pulsed electric fields (PEF) treatment on\nthe peeling ability of tomatoes and kiwi fruits. Front. Food Sci. Technol. 3, 1152111\nhttps://doi.org/10.3389/frfst.2023.1152111.\nGiner, J., Gimeno, V., Espachs, A., Elez, P., Barbosa-C\u00e1novas, G.V., Mart\u00edn, O., 2000.\nInhibition of tomato (Licopersicon esculentum mill.) pectin methylesterase by pulsed\nelectric fields. Innov. Food Sci. Emerg. Technol. 1, 57\u201367. https://doi.org/10.1016/\nS1466-8564(00)00003-5.\nGrinberga-Zalite, G., Zvirbule, A., 2022. Analysis of waste minimization challenges to\nEuropean Food Production Enterprises. Emerg. Sci. J. 6 (3), 530-543. https://doi.\norg/10.28991/ESJ-2022-06-03-08.\nGuida, V., Ferrari, G., Pataro, G., Chambery, A., Di Maro, A., Parente, A., 2013. The\neffects of ohmic and conventional blanching on the nutritional, bioactive compounds\nand quality parameters of artichoke heads. Lwt 53, 569\u2013579. https://doi.org/\n10.1016/j.lwt.2013.04.006.\nGupta, S., Sastry, S.K., 2018. Ohmic heating assisted lye peeling of pears. J. Food Sci. 83,\n1292-1298. https://doi.org/10.1111/1750-3841.14116.\n\n\nManfredi, M., Vignali, G., 2014. Life cycle assessment of a packaged tomato puree: a\ncomparison of environmental impacts produced by different life cycle phases.\nJ. Clean. Prod. 73, 275-284. https://doi.org/10.1016/j.jclepro.2013.10.010.\nMaxime, D., Marcotte, M., Arcand, Y., 2006. Development of eco-efficiency indicators for\nthe Canadian food and beverage industry. J. Clean. Prod. 14, 636\u2013648. https://doi.\norg/10.1016/j.jclepro.2005.07.015.\nMead, I., 2017. International Energy Outlook 2017 Overview. US Energy Information\nAdministration.\nMekonnen, M.M., Gerbens-Leenes, W., 2020. The water footprint of global food\nporduction. Water 12, 2696\u20132780. https://doi.org/10.3390/w12102696.\nMeneses, Y.E., Martinez, B., Hu, X., 2019. Water Reconditioning in the Food Industry,\nSustainable Water and Wastewater Processing. Elsevier Inc. https://doi.org/\n10.1016/B978-0-12-816170-8.00010-7.\nMurakawa, H., Ichimura, S., Sugimoto, K., Asano, H., Umezawa, S., Taira, H., Ibi, M.,\nAkahane, H., 2021. Measurement of steam flow rates using a clamp-on ultrasonic\nflowmeter with various wetness fractions. Flow Meas. Instrum. 80, 101997 https://\ndoi.org/10.1016/j.flowmeasinst.2021.101997.\nOdriozola-Serrano, I., Soliva-Fortuny, R., Hern\u00e1ndez-Jover, T., Mart\u00edn-Belloso, O., 2009.\nCarotenoid and phenolic profile of tomato juices processed by high intensity pulsed\nelectric fields compared with conventional thermal treatments. Food Chem. 112,\n258-266. https://doi.org/10.1016/j.foodchem.2008.05.087.\n\u00d6lmez, H., 2013. Water consumption, reuse and reduction strategies in food processing.\nIn: Sustainable Food Processing. John Wiley & Sons, Ltd, pp. 401\u2013437. https://doi.\norg/10.1002/9781118634301.ch17.\nPan, Z., Li, X., Bingol, G., Mchugh, T.H., Atungulu, G.G., 2009. Development of infrared\nradiation heating method for sustainable tomato peeling. Appl. Eng. Agric. 25,\n935-942.\nPanepinto, D., Viggiano, F., Genon, G., 2014. The potential of biomass supply for\nenergetic utilization in a small Italian region: basilicata. Clean Technol. Environ.\nPolicy 16, 833-845. https://doi.org/10.1007/s10098-013-0675-6.\nPataro, G., Ferrari, G., 2020. Limitations of Pulsed Electric Field Utilization in Food\nIndustry, Pulsed Electric Fields to Obtain Healthier and Sustainable Food for\nTomorrow. INC. https://doi.org/10.1016/B978-0-12-816402-0.00013-6.\nPataro, G., Dons\u00ec, G., Ferrari, G., 2011. Aseptic processing of apricots in syrup by means\nof a continuous pilot scale ohmic unit. Lwt 44, 1546-1554. https://doi.org/10.1016/\nj.lwt.2011.01.026.\nPataro, G., Barca, G.M.J., Pereira, R.N., Vicente, A.A., Teixeira, J.A., Ferrari, G., 2014.\nQuantification of metal release from stainless steel electrodes during conventional\nand pulsed ohmic heating. Innov. Food Sci. Emerg. Technol. 21, 66\u201373. https://doi.\norg/10.1016/j.ifset.2013.11.009.\nPataro, G., Carullo, D., Bakar Siddique, M.A., Falcone, M., Dons\u00ec, F., Ferrari, G., 2018.\nImproved extractability of carotenoids from tomato peels as side benefits of PEF\ntreatment of tomato fruit for more energy-efficient steam-assisted peeling. J. Food\nEng. 233, 65\u201373. https://doi.org/10.1016/j.jfoodeng. 2018.03.029.\nPataro, G., Carullo, D., Falcone, M., Ferrari, G., 2020. Recovery of lycopene from\nindustrially derived tomato processing by-products by pulsed electric fields-assisted\nextraction. Innov. Food Sci. Emerg. Technol. 63, 102369 https://doi.org/10.1016/j.\nifset.2020.102369.\nPatras, A., Brunton, N., Da Pieve, S., Butler, F., Downey, G., 2009. Effect of thermal and\nhigh pressure processing on antioxidant activity and instrumental colour of tomato\nand carrot pur\u00e9es. Innov. Food Sci. Emerg. Technol. 10, 16\u201322. https://doi.org/\n10.1016/j.ifset.2008.09.008.\nPereira, R.N., Vicente, A.A., 2010. Environmental impact of novel thermal and non-\nthermal technologies in food processing. Food Res. Int. 43, 1936\u20131943. https://doi.\norg/10.1016/j.foodres.2009.09.013.\nPereira, R.N., Rodrigues, R.M., Genisheva, Z., Oliveira, H., de Freitas, V., Teixeira, J.A.,\nVicente, A.A., 2016. Effects of ohmic heating on extraction of food-grade\nphytochemicals from colored potato. Lwt 74, 493-503. https://doi.org/10.1016/j.\nlwt.2016.07.074.\nPeterson, S.D., Am\u00f3n, R., Wong, T., Spang, E.S., Simmons, C.W., 2022. Material and\nenergy flow analysis of craft brewing: a case study at a California microbrewery.\nFront. Sustain. Food Syst. 6, 1028520 https://doi.org/10.3389/fsufs.2022.1028520.\nQu, W., Liu, Y., Feng, Y., Ma, H., 2022. Research on tomato peeling using flame-catalytic\ninfrared radiation. Lwt 163, 113542. https://doi.org/10.1016/j.lwt.2022.113542.\nRajak, A.R.A., 2022. Emerging technological methods for effective farming by cloud\ncomputing and IoT. Emerg. Sci. J. 6, 1017\u20131031. https://doi.org/10.28991/ESJ-\n2022-06-05-07.\nRaso, J., Frey, W., Ferrari, G., Pataro, G., Knorr, D., Teissie, J., Miklav\u010di\u010d, D., 2016.\nRecommendations guidelines on the key information to be reported in studies of\napplication of PEF technology in food and biotechnological processes. Innov. Food\nSci. Emerg. Technol. 37, 312\u2013321. https://doi.org/10.1016/j.ifset.2016.08.003.\nRathnakumar, K., Kalaivendan, R.G.T., Eazhumalai, G., Raja Charles, A.P., Verma, P.,\nRustagi, S., Bharti, S., Kothakota, A., Siddiqui, S.A., Manuel Lorenzo, J.,\nPandiselvam, R., 2023. Applications of ultrasonication on food enzyme inactivation-\nrecent review report (2017-2022). Ultrason. Sonochem. 96, 106407 https://doi.org/\n10.1016/j.ultsonch.2023.106407.\nReyes-De-Corcuera, J.I., Goodrich-Schneider, R.M., Barringer, S.A., Landeros-Urbina, M.\nA., 2014. Processing of fruit and vegetable beverages. In: Clark, S., Jung, S.,\nLamsal, B. (Eds.), Food Processing: Principles and Applications. Wiley, pp. 339\u2013362.\nhttps://doi.org/10.1002/9781118846315.ch15.\nRingler, C., Willenbockel, D., Perez, N., Rosegrant, M., Zhu, T., Matthews, N., 2016.\nGlobal linkages among energy, food and water: an economic assessment. J. Environ.\nStud. Sci. 6, 161\u2013171. https://doi.org/10.1007/s13412-016-0386-5.\nRock, C., Yang, W., Nooji, J., Teixeira, A., Feng, H., 2010. Evaluation of roma tomato\n(Solanum lycopersicum) peeling methods: conventional vs. Power ultrasound. In:\nProceedings of the Florida State Horticultural Society, pp. 241-245.\n\n\nHamidov, A., Helming, K., 2020. Sustainability considerations in water-energy-food\n\n\nnexus research in irrigated agriculture. Sustain. Times 12. https://doi.org/10.3390/\nSU12156274.\n\n\nHauptmann, P., Hoppe, N., P\u00fcttmer, A., 2002. Application of ultrasonic sensors in the\nprocess industry. Meas. Sci. Technol. 13 (8), R73-R83.\nIslam, M.F., Karim, Z., 2019. World's demand for food and water: the consequences of\nclimate change. In: Desalination - Challenges and Opportunities of Climate Change,\npp. 1-27. https://doi.org/10.5772/intechopen.85919.\nJayathunge, K.G.L.R., Stratakos, A.C., Delgado-Pando, G., Koidis, A., 2019. Thermal and\nnon-thermal processing technologies on intrinsic and extrinsic quality factors of\ntomato products: a review. J. Food Process. Preserv. 43, 1-15. https://doi.org/\n10.1111/jfpp.13901.\nJunqua, R., Carullo, D., Ferrari, G., Pataro, G., Ghidossi, R., 2021. Ohmic heating for\npolyphenol extraction from grape berries: an innovative prefermentary process.\nOeno One 55, 39-51. https://doi.org/10.20870/oeno-one.2021.55.3.4647.\nKarakaya, A., \u00d6zilgen, M., 2011. Energy utilization and carbon dioxide emission in the\nfresh, paste, whole-peeled, diced, and juiced tomato production processes. Energy\n36, 5101-5110. https://doi.org/10.1016/j.energy.2011.06.007.\nKoch, Y., Witt, J., Lammerskitten, A., Siemer, C., Toepfl, S., 2022. The influence of Pulsed\nElectric Fields (PEF) on the peeling ability of different fruits and vegetables. J. Food\nEng. 322, 110938 https://doi.org/10.1016/j.jfoodeng.2021.110938.\nKohli, D., Champawat, P.S., Mudgal, V.D., Jain, S.K., Tiwari, B.K., 2021. Advances in\npeeling techniques for fresh produce. J. Food Process. Eng. 44 https://doi.org/\n10.1111/jfpe.13826.\nKrebbers, B., Matser, A.M., Hoogerwerf, Sanne W., Moezelaar, R., Tomassen, M.M.M.,\nvan den Berg, R.W., 2003. Combined high-pressure and thermal treatments for\nprocessing of tomato puree: evaluation of microbial inactivation and quality\nparameters. Innov. Food Sci. Emerg. Technol. 4, 377-385. https://doi.org/10.1016/\nS1466-8564.\nLatini, A., Giagnacovo, G., Campiotti, C.A., Bernardini, A., Fuentes-Pila, J., 2017. Energy\nanalysis and current value stream map applied to a typical triple concentrate tomato\npaste production plant with an average production capacity of 500 ton/day. Open\nAccess J. Agric. Res. 2, 1-9. https://doi.org/10.23880/oajar-16000140.\nLauteri, C., Ferri, G., Piccinini, A., Pennisi, L., Vergara, A., 2023. Ultrasound technology\nas inactivation method for foodborne pathogens: a review. Foods 12, 1212. https://\ndoi.org/10.3390/foods12061212.\nLee, S.Y., Sagong, H.G., Ryu, S., Kang, D.H., 2012. Effect of continuous ohmic heating to\ninactivate Escherichia coli O157:H7, Salmonella Typhimurium and Listeria\nmonocytogenes in orange juice and tomato juice. J. Appl. Microbiol. 112, 723\u2013731.\nhttps://doi.org/10.1111/j.1365-2672.2012.05247.x.\nLi, X., Pan, Z., 2014a. Dry-peeling of tomato by infrared radiative heating: Part I. Model\ndevelopment. Food Bioprocess Technol. 7, 1996-2004. https://doi.org/10.1007/\ns11947-013-1203-8.\nLi, X., Pan, Z., 2014b. Dry peeling of tomato by infrared radiative heating: Part II. Model\nvalidation and sensitivity analysis. Food Bioprocess Technol. 7, 2005-2013. https://\ndoi.org/10.1007/s11947-013-1188-3.\n\n\nLi, X., Pan, Z., Atungulu, G.G., Wood, D., McHugh, T., 2014a. Peeling mechanism of\ntomato under infrared heating: peel loosening and cracking. J. Food Eng. 128,\n79-87. https://doi.org/10.1016/j.jfoodeng.2013.12.020.\n\n\nLi, X., Pan, Z., Atungulu, G.G., Zheng, X., Wood, D., Delwiche, M., McHugh, T.H., 2014b.\nPeeling of tomatoes using novel infrared radiation heating technology. Innov. Food\nSci. Emerg. Technol. 21, 123\u2013130. https://doi.org/10.1016/j.ifset.2013.10.011.\nLiu, C., Zhang, Z., Liu, S., Liu, Q., Feng, B., Tanzer, J., 2019. Evaluating agricultural\nsustainability based on the water-energy-food nexus in the Chenmengquan irrigation\ndistrict of China. Sustain. Times 11. https://doi.org/10.3390/sul1195350.\n\n\n15\n\n\nE. Eslami et al.\n\n\nJournal of Cleaner Production 425 (2023) 138996\n\n\nRock, C., Yang, W., Goodrich-Schneider, R., Feng, H., 2012. Conventional and alternative\nmethods for tomato peeling. Food Eng. Rev. 4, 1\u201315. https://doi.org/10.1007/\ns12393-011-9047-3.\nRossini, G., Toscano, G., Duca, D., Corinaldesi, F., Foppa Pedretti, E., Riva, G., 2013.\nAnalysis of the characteristics of the tomato manufacturing residues finalized to the\nenergy recovery. Biomass Bioenergy 51, 177\u2013182. https://doi.org/10.1016/j.\nbiombioe.2013.01.018.\nSawant, S.R., Pandey, J.P., Singh, A., Prakash, O., 2018. Performance and evaluation of\nohmic heating assisted lye and salt concentration on peeling quality of tomato. Int. J.\nCurr. Microbiol. Appl. Sci. 7, 3515-3524. https://doi.org/10.20546/\nijcmas.2018.709.436.\nSeybold, C., Fr\u00f6hlich, K., Bitsch, R., Otto, K., B\u00f6hm, V., 2004. Changes in contents of\ncarotenoids and vitamin E during tomato processing. J. Agric. Food Chem. 52,\n7005-7010. https://doi.org/10.1021/jf049169c.\nShams, R., Manzoor, S., Shabir, I., Dar, A.H., Dash, K.K., Srivastava, S., Pandey, V.K.,\nBashir, I., Khan, S.A., 2023. Pulsed Electric Field-Induced Modification of Proteins: A\nComprehensive Review, Food and Bioprocess Technology. Springer US. https://doi.\norg/10.1007/s11947-023-03117-x.\nSims, R., Flammini, A., Puri, M., Bracco, S., 2015. Opportunities for Agri-Food Chains to\nBecome Energy-Smart. FAO USAID.\n\n\nTrueblood, A.J., Wu, Y.Y., Ganji, A.R., 2013. Potential for energy, peak demand, and\nwater savings in California tomato processing facilities. Ind. Energy Technol. Conf.\n2013.\nUN-Water, 2021. Water, Food and Energy. https://www.unwater.org/water-facts/wate\nr-food-and-energy. accessed 6.21.23.\nVidyarthi, S.K., 2017. Study and Modeling of Infrared Heating for Tomato Dry-Peeling\n(Ph.D. Dissertation). Department of Biological and Agricultural Engineering,\nUniversity of California at Davis.\nVidyarthi, S.K., El-Mashad, H.M., Khir, R., Zhang, R., McHugh, T.H., Pan, Z., 2019.\nTomato peeling performance under pilot scale catalytic infrared heating. J. Food\nEng. 246, 224\u2013231. https://doi.org/10.1016/j.jfoodeng.2018.11.002.\nWongsa-Ngasri, P., Sastry, S.K., 2015. Effect of ohmic heating on tomato peeling. LWT-\nFood Sci. Technol. 61, 269\u2013274. https://doi.org/10.1016/j.lwt.2014.12.053.\nWongsa-Ngasri, P., Sastry, S.K., 2016a. Tomato peeling by ohmic heating: effects of lye-\nsalt combinations and post-treatments on weight loss, peeling quality and firmness.\nInnov. Food Sci. Emerg. Technol. 34, 148\u2013153. https://doi.org/10.1016/j.\nifset.2016.01.013.\nWongsa-ngasri, P., Sastry, S.K., 2016b. Tomato peeling by ohmic heating with lye-salt\ncombinations: effects of operational parameters on peeling time and skin diffusivity.\nJ. Food Eng. 186, 10-16. https://doi.org/10.1016/j.jfoodeng.2016.04.005.\nWu, J., Gamage, T.V., Vilkhu, K.S., Simons, L.K., Mawson, R., 2008. Effect of\nthermosonication on quality improvement of tomato juice. Innov. Food Sci. Emerg.\nTechnol. 9, 186\u2013195. https://doi.org/10.1016/j.ifset.2007.07.007.\nXu, Z., Fan, Y., Wang, T., Huang, Y., Dehkordy, F.M., Dai, Z., Xia, L., Dong, Q.,\nBagtzoglou, A., McCutcheon, J., Lei, Y., Li, B., 2019. Towards high resolution\nmonitoring of water flow velocity using flat flexible thin mm-sized resistance-typed\nsensor film (MRSF). Water Res. X 4, 100028.\nYildiz, H., Baysal, T., 2006. Effects of alternative current heating treatment on\nAspergillus Niger, pectin methylesterase and pectin content in tomato. J. Food Eng.\n75, 327-332. https://doi.org/10.1016/j.jfoodeng.2005.04.020.\nZhou, Y.H., Vidyarthi, S.K., Yang, X.H., Duan, X., Liu, Z.L., Mujumdar, A.S., Xiao, H.W.,\n2022. Conventional and novel peeling methods for fruits and vegetables: a review.\nInnov. Food Sci. Emerg. Technol. 77, 102961 https://doi.org/10.1016/j.\nifset.2022.102961.\nZhoua, H., Jia, T., Wang, R., Ge, X., Tang, X., Tang, S., 2018. Multipath ultrasonic gas\nflowmeter based on multiple reference waves. Ultrasonics 82, 145\u2013152. https://doi.\norg/10.1016/j.ultras.2017.07.010.\n\n\nSomavat, R., Mohamed, H.M.H., Sastry, S.K., 2013. Inactivation kinetics of Bacillus\n\n\ncoagulans spores under ohmic and conventional heating. Lwt 54, 194\u2013198. https://\ndoi.org/10.1016/j.lwt.2013.04.004.\n\n\nSteven, R., Hall, A., 2009. Orifice plate meter wet gas flow performance. Flow Meas.\nInstrum. 20, 141\u2013151. https://doi.org/10.1016/j.flowmeasinst. 2009.07.001.\nStrati, I.F., Oreopoulou, V., 2014. Recovery of carotenoids from tomato processing by-\nproducts - a review. Food Res. Int. 65, 311\u2013321. https://doi.org/10.1016/j.\nfoodres.2014.09.032.\nTerefe, N.S., Gamage, M., Vilkhu, K., Simons, L., Mawson, R., Versteeg, C., 2009. The\nkinetics of inactivation of pectin methylesterase and polygalacturonase in tomato\njuice by thermosonication. Food Chem. 117, 20-27. https://doi.org/10.1016/j.\n\n\nfoodchem.2009.03.067.\n\n\nTorkian Boldaji, M., Borghei, A.M., Beheshti, B., Hosseini, S.E., 2015. The process of\nproducing tomato paste by ohmic heating method. J. Food Sci. Technol. 52,\n3598-3606. https://doi.org/10.1007/s13197-014-1424-5.\n\n\n16\n16\n"}, "expected_output": {"claims": [{"unit": "kWh/ton", "value": 355, "evidence": ["355\n", "KPIs in triple tomato concentrate and peeled tomato production lines.", "Thermal energy\nconsumption per\nton of tomato\nproducts (kWh/\nton)\nElectrical energy\nconsumption per\nton of tomato\nproducts (kWh/\nton)\nWater consumption\nper ton of tomato\nproducts (m\u00b3/\nton)\n", "average KPIs for water, thermal, and electrical energy consumption per ton of final products in triple tomato paste, tomato puree, and peeled tomato production lines.", "Thermal energy\nconsumption per\nton of tomato\nproducts (kWh/\nton)"]}]}, "metadata": {"product_category": "Food & beverages", "request_id": "req_d40c092f4cff5a94"}}
{"id": "8b54a2c9f97cd28490065bd8", "input": {"query": "What are the fuel and steam consumption values for cold rolling operations including annealing? I need fuel consumption in GJ/tonne or MJ/kg and steam consumption for the annealing process. What are the world best practice values?", "source_url": "https://eta-publications.lbl.gov/sites/default/files/industrial_best_practice_en.pdf", "document_text": "LBNL-62806\n\n\nREV. 2\n\n\nERNEST ORLANDO LAWRENCE\nBERKELEY NATIONAL LABORATORY\n\n\nBERKELEY LAB\n\n\nWorld Best Practice Energy\nIntensity Values for Selected\nIndustrial Sectors\n\n\nErnst Worrell, Lynn Price, Maarten Neelis,\nChristina Galitsky, Zhou Nan\n\n\nEnvironmental Energy\nTechnologies Division\n\n\nFebruary 2008\n\n\nThis work was supported by the China Sustainable Energy Program of the Energy\nFoundation through the U.S. Department of Energy under Contract No. DE-AC02-\n05CH11231.\n\n\nDisclaimer\n\n\nThis document was prepared as an account of work sponsored by the United States\nGovernment. While this document is believed to contain correct information, neither\nthe United States Government nor any agency thereof, nor The Regents of the\nUniversity of California, nor any of their employees, makes any warranty, express or\nimplied, or assumes any legal responsibility for the accuracy, completeness, or\nusefulness of any information, apparatus, product, or process disclosed, or represents\nthat its use would not infringe privately owned rights. Reference herein to any specific\ncommercial product, process, or service by its trade name, trademark, manufacturer, or\notherwise, does not necessarily constitute or imply its endorsement, recommendation,\nor favoring by the United States Government or any agency thereof, or The Regents of\nthe University of California. The views and opinions of authors expressed herein do\nnot necessarily state or reflect those of the United States Government or any agency\nthereof, or The Regents of the University of California.\n\n\nErnest Orlando Lawrence Berkeley National Laboratory is an equal\nopportunity employer.\n\n\nAbstract\n\n\n\"World best practice\u201d energy intensity values, representing the most energy-efficient\nprocesses that are in commercial use in at least one location worldwide, are provided for\nthe production of iron and steel, aluminium, cement, pulp and paper, ammonia, and\nethylene. Energy intensity is expressed in energy use per physical unit of output for each\nof these commodities; most commonly these are expressed in metric tonnes (t). The\nenergy intensity values are provided by major energy-consuming processes for each\nindustrial sector to allow comparisons at the process level. Energy values are provided for\nfinal energy, defined as the energy used at the production facility as well as for primary\nenergy, defined as the energy used at the production facility as well as the energy used to\nproduce the electricity consumed at the facility. The \u201cbest practice\u201d figures for energy\nconsumption provided in this report should be considered as indicative, as these may\ndepend strongly on the material inputs.\n\n\nKey words: energy intensity, industry, steel, aluminium, cement, paper, ammonia,\nethylene\n\n\nTable of Contents\n\n\n1. Introduction.....\n\n\n1\n\n\n2. World Best Practice Energy Intensity Values..\n2.1. Iron and Steel...\n\n\n5\n\n\n5\n\n\n2.1.1 Blast Furnace \u2013 Basic Oxygen Furnace Route..\n2.1.2 Smelt Reduction \u2013 Basic Oxygen Furnace\u2026..\n2.1.3 Direct Reduced Iron - Electric Arc Furnace.\n2.1.4 Electric Arc Furnace..\n\n\n6\n\n\n10\n\n\n13\n\n\n15\n\n\n2.1.5 Casting...\n\n\n16\n\n\n2.1.6 Rolling and Finishing..\n\n\n17\n\n\n2.2. Aluminium...\n\n\n18\n\n\nAlumina Production..\n\n\n2.2.1\n\n\n19\n\n\n2.2.2 Anode Manufacture..\n\n\n19\n\n\n2.2.3 Aluminium Smelting (Electrolysis)\u2026\u2026.\n\n\n.20\n\n\n2.2.4 Ingot Casting.....\n\n\n.20\n\n\n2.2.5 Secondary Aluminium Production...\n\n\n.21\n\n\n2.3. Cement...\n\n\n22\n\n\nRaw Materials and Fuel Preparation..\n\n\n2.3.1\n\n\n.22\n\n\n2.3.2\n\n\nClinker Production...\n\n\n29\n\n\n2.3.3 Additive Preparation..\n\n\n29\n\n\n2.3.4 Cement grinding...\n\n\n29\n\n\n2.3.5 Other Production Energy..\n\n\n30\n\n\n2.4. Pulp and Paper........\n\n\n31\n\n\nNon-Wood Pulping...\n\n\n2.4.1\n\n\n.32\n\n\n2.4.2 Kraft Pulping\n\n\n.35\n\n\n2.4.3 Sulfite Pulping.....\n\n\n35\n\n\n2.4.4\n\n\nMechanical Pulping.\n\n\n.35\n\n\n2.4.5 Fiber Recovery..\n\n\n36\n\n\n2.4.6 Papermaking..\n\n\n.36\n\n\n2.4.7 Integrated Pulp and Paper Mills..\n\n\n..36\n\n\n2.5. Ammonia..\n\n\n..38\n\n\nNatural Gas Steam Reforming.\n\n\n2.5.1\n\n\n.38\n\n\n2.5.2\n\n\nCoal....\n\n\n.39\n\n\n2.6. Ethylene..\n\n\n.40\n\n\n2.6.1 Naphtha and Ethane\n\n\n40\n\n\n2.6.2 Other Feedstocks and Emerging Technologies.......\n\n\n.43\n\n\n3. Summary and Next Steps ....\n\n\n44\n\n\n4. Acknowledgments....\n\n\n.44\n\n\n1. Introduction\n\n\nThis report provides information on world best practice energy intensity values for\nproduction of iron and steel, aluminium, cement, pulp and paper, ammonia, and ethylene.\n\"World best practice\" values represent the most energy-efficient processes that are in\ncommercial use in at least one location worldwide.\u00b9\n\n\nThese values are expressed in energy use per physical unit of output for each of these\ncommodities; most commonly these are expressed in metric tonnes (t). Energy values are\nprovided in both syst\u00e8me international (SI) units (joules) and in kilograms of coal\nequivalent (kgce), a common unit in China.\u00b2\n\n\n2\n\n\nEnergy values are provided for final energy, defined as the energy used at the production\nfacility as well as for primary energy, defined as the energy used at the production facility\nas well as the energy used to produce the electricity consumed at the facility. For primary\nenergy values, the losses associated with conversion of fuels into electricity along with\nthe losses associated with transmission and distribution of the electricity are included. It\nis assumed that these losses are 67%. The energy values referenced in the text of this\ndocument are provided for final energy only; primary energy values can be found in the\ntables.\n\n\nTable 1.1 provides a summary of the world best practice final energy intensity values for\nthe sectors covered in this report. Table 1.2 provides a summary of the world best\npractice primary energy intensity values. Details regarding the calculation of these values\nand references are provided in the following sections.\n\n\n1 While this report describes best practices in energy efficiency for key processes, the integration of these\nindividual technologies is key to obtain the full benefits of these technologies. For example, combined heat\nand power would increase the efficiency of steam supply for the described processes, while by-product\nenergy flows may also be used more efficiently by implementing more efficient technologies (e.g. use of\nblast-furnace gas in a combined cycle instead of a boiler).\n\n\n2\n\n\nIn April 2006, China's central government launched the Top-1000 Enterprises Energy-Efficiency\nProgram (Top-1000 program), the goal of which is to improve industrial energy efficiency by targeting\nChina's 1000 highest energy-consuming enterprises. These enterprises currently account for approximately\n50% of total industrial sector energy consumption and 30% of total energy consumption in China. During\nthe summer of 2006, energy-saving agreements with targets for 2010 were signed with all Top-1000\nenterprises. The Top-1000 enterprises are from the iron and steel, petroleum and petrochemical, chemical,\nnon-ferrous metal, building materials, pulp and paper, electricity production, coal mining, and textile\nindustries. Chinese government officials have expressed a desire to understand how Chinese industrial\nenterprises compare to international best practice.\n\n\n1\n\n\nTable 1.1. Summary of World Best Practice Final Energy Intensity Values for Selected\nIndustrial Sectors\n\n\nUnit\n\n\nkgce/t\n\n\nGJ/t\n\n\nIron and Steel\n\n\nBlast Furnace \u2013 Basic Oxygen Furnace \u2013 Thin Slab Casting\nSmelt Reduction \u2013 Basic Oxygen Furnace \u2013 Thin Slab Casting\nDirect Reduced Iron - Electric Arc Furnace - Thin Slab Casting\nScrap - Electric Arc Furnace \u2013 Thin Slab Casting\n\n\n14.8\n\n\n504.5\n\n\nt steel\n\n\nt steel\n\n\n17.8\n\n\n606.4\n\n\n16.9\n\n\n576.2\n\n\nt steel\n\n\nt steel\n\n\n87.5\n\n\n2.6\n\n\nAluminium\n\n\nPrimary Aluminium\n\n\nt aluminium\n\n\n70.6\n\n\n2411\n\n\nSecondary Aluminium\n\n\nt aluminium\n\n\n2.5\n\n\n85\n\n\nCement\n\n\nPortland Cement\n\n\n2.9\n\n\n100\n\n\nt cement\n\n\nFly Ash Cement\n\n\n2.0\n\n\n70\n\n\nt cement\n\n\n1.7\n\n\nBlast Furnace Slag Cement\n\n\n57\n\n\nt cement\n\n\nPulp\n\n\nNon-wood Market Pulp\n\n\nair dried t\n\n\n264\n\n\n7.7\n\n\nWood Kraft Pulp\nWood Sulfite Pulp\n\n\nair dried t\n\n\n11.1\n\n\n380\n\n\nair dried t\n\n\n632\n\n\n18.5\n\n\nWood Thermo-mechanical Pulp\n\n\n224\n\n\nair dried t\n\n\n6.6\n\n\nRecovered Paper Pulp\n\n\nair dried t\n\n\n1.5\n\n\n51\n\n\nPaper\n\n\nUncoated Fine Paper\n\n\nair dried t\n\n\n9.0\n\n\n307\n\n\nCoated Fine Paper\n\n\nair dried t\n\n\n355\n\n\n10.4\n\n\nair dried t\n\n\n7.2\n\n\nNewsprint\n\n\n244\n\n\nair dried t\n\n\n327\n\n\nBoard\n\n\n9.6\n\n\nair dried t\n\n\nKraftliner\n\n\n7.8\n\n\n267\n\n\nTissue\n\n\nair dried t\n\n\n10.5\n\n\n358\n\n\nPulp and Paper\n\n\nair dried t\n\n\n18.3\n\n\nBleached Uncoated Fine\n\n\n625\n\n\nKrafliner (unbleached)/Bag Paper\n\n\nair dried t\n\n\n601\n\n\n17.6\n\n\nair dried t\n\n\n22.4\n\n\nBleached Coated Fine\n\n\n765\n\n\nair dried t\n\n\nBleached Uncoated Fine\n\n\n22.3\n\n\n762\n\n\nNewsprint\n\n\nair dried t\n\n\n6.6\n\n\n226\n\n\nMagazine Paper\n\n\nair dried t\n\n\n7.3\n\n\n248\n\n\nair dried t\n\n\n402\n\n\nBoard\n\n\n11.8\n\n\nair dried t\n\n\nRecovered Paper Board\n\n\n11.2\n\n\n384\n\n\nRecovered Paper Newsprint\n\n\nair dried t\n\n\n259\n\n\n7.6\n\n\nair dried t\n\n\nRecovered Paper Tissue\n\n\n11.3\n\n\n386\n\n\nAmmonia\n\n\nNatural Gas Feedstock (Steam Reforming)\nCoal Feedstock\n\n\nt ammonia\n\n\n28\n\n\n956\n\n\n34.8\n\n\nt ammonia\n\n\n1188\n\n\nEthylene\n\n\nt high value\n\n\nEthane Cracking\n\n\n12.5\n\n\n427\n\n\nchemicals\n\n\nt high value\n\n\nNaphtha Cracking\n\n\n11\n\n\n409\n\n\nchemicals\n\n\n2\n\n\nTable 1.2. Summary of World Best Practice Primary Energy Intensity Values for Selected\nIndustrial Sectors\n\n\nUnit\n\n\nkgce/t\n\n\nGJ/t\n\n\nIron and Steel\n\n\n16.3\n\n\n555.1\n\n\nBlast Furnace \u2013 Basic Oxygen Furnace \u2013 Thin Slab Casting\nSmelt Reduction \u2013 Basic Oxygen Furnace \u2013 Thin Slab Casting\nDirect Reduced Iron - Electric Arc Furnace - Thin Slab Casting\nScrap - Electric Arc Furnace \u2013 Thin Slab Casting\n\n\nt steel\n\n\nt steel\n\n\n19.2\n\n\n656.8\n\n\n18.6\n\n\nt steel\n\n\n635.8\n\n\n205.1\n\n\nt steel\n\n\n6.0\n\n\nAluminium\n\n\nPrimary Aluminium\n\n\nt aluminium\n\n\n174\n\n\n5940\n\n\nSecondary Aluminium\n\n\nt aluminium\n\n\n7.6\n\n\n259\n\n\nCement\n\n\nPortland Cement\n\n\n3.4\n\n\n115\n\n\nt cement\n\n\nFly Ash Cement\n\n\n2.5\n\n\n84\n\n\nt cement\n\n\n73\n\n\nBlast Furnace Slag Cement\n\n\n2.1\n\n\nt cement\n\n\nPulp\n\n\n364\n\n\n10.7\n\n\nNon-wood Market Pulp\n\n\nair dried t\n\n\n11.0\n\n\n377\n\n\nWood Kraft Pulp\nWood Sulfite Pulp\n\n\nair dried t\n\n\n23.6\n\n\n807\n\n\nair dried t\n\n\n770\n\n\n22.6\n\n\nWood Thermo-mechanical Pulp\n\n\nair dried t\n\n\nRecovered Paper Pulp\n\n\n3.9\n\n\n133\n\n\nair dried t\n\n\nPaper\n\n\n467\n\n\n13.7\n\n\nUncoated Fine Paper\n\n\nair dried t\n\n\n16.3\n\n\n558\n\n\nCoated Fine Paper\n\n\nair dried t\n\n\n11.3\n\n\n386\n\n\nair dried t\n\n\nNewsprint\n\n\n527\n\n\n15.4\n\n\nair dried t\n\n\nBoard\n\n\n11.7\n\n\n401\n\n\nair dried t\n\n\nKraftliner\n\n\n17.8\n\n\n608\n\n\nTissue\n\n\nair dried t\n\n\nPulp and Paper\n\n\n925\n\n\n27.1\n\n\nair dried t\n\n\nBleached Uncoated Fine\n\n\n24.9\n\n\nKrafliner (unbleached)/Bag Paper\n\n\nair dried t\n\n\n850\n\n\n24.9\n\n\n850\n\n\nair dried t\n\n\nBleached Coated Fine\n\n\n33.4\n\n\n1139\n\n\nair dried t\n\n\nBleached Uncoated Fine\n\n\n1061\n\n\n31.1\n\n\nNewsprint\n\n\nair dried t\n\n\n22.7\n\n\n775\n\n\nMagazine Paper\n\n\nair dried t\n\n\n772\n\n\n22.6\n\n\nair dried t\n\n\nBoard\n\n\n28.6\n\n\n976\n\n\nair dried t\n\n\nRecovered Paper Board\n\n\nRecovered Paper Newsprint\n\n\n17.8\n\n\n608\n\n\nair dried t\n\n\n14.9\n\n\n509\n\n\nair dried t\n\n\nRecovered Paper Tissue\n\n\nAmmonia\n\n\nNatural Gas Feedstock Steam Reforming\nCoal Feedstock\n\n\nt ammonia\n\n\n28\n\n\n956\n\n\nt ammonia\n\n\n34.8\n\n\n1188\n\n\nEthylene\n\n\nt high value\nchemicals\n\n\nEthane Cracking\n\n\n496\n\n\n14.5\n\n\nt high value\nchemicals\n\n\nNaphtha Cracking\n\n\n13\n\n\n478\n\n\nNote: Primary energy includes electricity generation, transmission, and distribution losses of 67%.\n\n\n3\n\n\n4\n\n\n2. World Best Practice Energy Intensity Values\n\n\nWorld best practice energy intensity values for production of iron and steel, aluminium,\ncement, pulp and paper, ammonia, and ethylene are provided in the following sections.\n\n\n2.1 Iron and Steel\n\n\nThis section provides world best practice energy intensity values by process for iron and\nsteelmaking based on four possible process configurations:\n\n\nBlast Furnace \u2013 Basic Oxygen Furnace\n\n\n\"\n\n\nSmelt Reduction - Basic Oxygen Furnace\n\n\n\u25a0\n\n\n\u25c9 Direct Reduced Iron - Electric Arc Furnace\n\n\nScrap Electric Arc Furnace\n\n\n\"\n\n\n-\n\n\nThe world best practice values for these four process configurations are provided\nseparately for hot rolled bars and for cold rolled and finished steel using a continuous\ncaster. In addition, world best practice values are provided for thin slab (near net shape)\ncasting. Tables 2.1.1 and 2.2.2, along with the accompanying text, provide more details\non these best practice final and primary energy intensity values, respectively, by process.\n\n\nTable 2.1.1. World Best Practice Final Energy Intensity Values for Iron and Steel (values\nare per metric ton of steel).\n\n\nSmelt Reduction-\nBasic Oxygen\nFurnace\n\n\nBlast Furnace\nBasic Oxygen\nFurnace\n\n\nDirect Reduced\nIron - Electric\nArc Furnace\nkgce/t\n\n\nScrap-\nElectric Arc\n\n\n-\n\n\nFurnace\n\n\nGJ/t\n\n\nGJ/t kgce/t\n\n\nkgce/t\n\n\nkgce/t\n\n\nGJ/t\n\n\nGJ/t\n\n\nSintering\n\n\nMaterial\n\n\n65.2\n\n\n65.2\n\n\n1.9\n\n\n1.9\n\n\nPelletizing\n\n\nPreparation\n\n\n0.6\n\n\n19.0\n\n\n0.6\n\n\n19.0\n\n\nCoking\n\n\n0.8\n\n\n28.6\n\n\nBlast Furnace\n\n\nIronmaking\n\n\n12.2\n\n\n414.9\n\n\nSmelt Reduction\n\n\n17.3\n\n\n591.6\n\n\nDirect Reduced Iron\n\n\n11.7\n\n\n399.6\n\n\nBasic Oxygen Furnace\n\n\nSteelmaking\n\n\n-15.4\n\n\n-0.4\n\n\n-15.4\n\n\n-0.4\n\n\nElectric Arc Furnace\n\n\n80.6\n\n\n2.5\n\n\n85.6\n\n\n2.4\n\n\nRefining\n\n\n0.1\n\n\n4.3\n\n\n0.1\n\n\n4.3\n\n\nContinuous Casting\n\n\nCasting\n\n\n0.1\n\n\n0.1\n\n\n2.0\n\n\n0.1\n\n\n2.0\n\n\n0.1\n\n\n2.0\n\n\n2.0\n\n\nand Rolling\n\n\nHot Rolling\n\n\n62.5\n\n\n1.8\n\n\n1.8\n\n\n62.5\n\n\n1.8\n\n\n62.5\n\n\n1.8\n\n\n62.5\n\n\n562.2\n\n\n664.0\n\n\n633.9\n\n\n16.5\n\n\n19.5\n\n\n4.3\n\n\n145.1\n\n\n18.6\n\n\nSub-Total\n\n\nCold Rolling\n\n\nCold Rolling\n\n\n0.4\n\n\n0.4\n\n\n13.7\n\n\n13.7\n\n\nand Finishing\n\n\nFinishing\n\n\n1.1\n\n\n38.1\n\n\n1.1\n\n\n38.1\n\n\n18.0\n\n\n715.8\n\n\n18.6\n\n\n633.9\n\n\n613.9\n\n\n21.0\n\n\n4.3\n\n\n145.1\n\n\nTotal\n\n\nReplace Continuous\n\n\nAlternative:\n\n\nCasting and\nRolling\nAlternative\n\n\nCasting and Rolling\n\n\n0.2\n\n\n0.2\n\n\n6.9\n\n\n0.2\n\n\n6.9\n\n\n0.2\n\n\n6.9\n\n\n6.9\n\n\nwith Thin Slab Casting\n\n\n16.9\n\n\n14.8\n\n\n504.5\n\n\n17.8\n\n\n606.4\n\n\n576.2\n\n\n2.6\n\n\n87.5\n\n\nTotal\n\n\nTotals for process routes depend on the feedstock and material flows and differ from plant to plant; totals should not be\nused to compare individual plants. Hot rolling values are based on energy use for production of hot rolled bars - see\nfollowing tables for data on hot rolling strip or wire.\n\n\n5\n\n\nTable 2.1.2. World Best Practice Primary Energy Intensity Values for Iron and Steel\n(values are per metric ton of steel).\n\n\nDirect Reduced\nIron - Electric\nArc Furnace\n\n\nBlast Furnace\n- Basic Oxygen\nFurnace\n\n\nSmelt Reduction-\n\n\nScrap -\nElectric Arc\n\n\nBasic Oxygen\nFurnace\n\n\n-\n\n\nFurnace\n\n\nGJ/t\n\n\nkgce/t\n\n\nkgce/t\n\n\nkgce/t\n\n\nGJ/t kgce/t\n\n\nGJ/t\n\n\nGJ/t\n\n\nMaterial\n\n\nSintering\n\n\n74.3\n\n\n2.2\n\n\n2.2\n\n\n74.3\n\n\nPelletizing\n\n\nPreparation\n\n\n0.8\n\n\n25.7\n\n\n0.8\n\n\n25.7\n\n\nCoking\n\n\n1.1\n\n\n36.3\n\n\nIronmaking\n\n\nBlast Furnace\n\n\n423.7\n\n\n12.4\n\n\nSmelt Reduction\n\n\n610.2\n\n\n17.9\n\n\nDirect Reduced Iron\n\n\n9.2\n\n\n315.6\n\n\nSteelmaking\n\n\nBasic Oxygen Furnace\nElectric Arc Furnace\n\n\n-0.3\n\n\n-9.5\n\n\n-0.3\n\n\n-9.5\n\n\n202.9\n\n\n5.5\n\n\n187.7\n\n\n5.9\n\n\nRefining\n\n\n0.4\n\n\n13.0\n\n\n0.4\n\n\n13.0\n\n\nContinuous Casting\n\n\nCasting\n\n\n0.1\n\n\n0.1\n\n\n3.9\n\n\n0.1\n\n\n3.9\n\n\n0.1\n\n\n3.9\n\n\n3.9\n\n\nand Rolling\n\n\nHot Rolling\n\n\n2.4\n\n\n2.4\n\n\n2.4\n\n\n80.4\n\n\n80.4\n\n\n80.4\n\n\n2.4\n\n\n80.4\n\n\n18.2\n\n\n622.0\n\n\n21.2\n\n\n723.7\n\n\n20.6\n\n\n702.7\n\n\n8.0\n\n\n272.0\n\n\nSub-Total\n\n\nCold Rolling\n\n\nCold Rolling\n\n\n32.1\n\n\n0.9\n\n\n32.1\n\n\n0.9\n\n\nand Finishing\n\n\nFinishing\n\n\n1.4\n\n\n1.4\n\n\n48.4\n\n\n48.4\n\n\n272.0\n\n\n20.6\n\n\n20.6\n\n\n702.5\n\n\n23.6\n\n\n804.2\n\n\n702.7\n\n\n8.0\n\n\nTotal\n\n\nReplace Continuous\n\n\nAlternative:\n\n\nCasting and\n\n\nCasting and Rolling\n\n\n17.3\n\n\n0.5\n\n\n0.5\n\n\n0.5\n\n\n17.3\n\n\n17.3\n\n\n0.5\n\n\n17.3\n\n\nwith Thin Slab Casting\n\n\nRolling\n\n\nAlternative\n\n\n205.1\n\n\n16.3\n\n\n6.0\n\n\n555.1\n\n\n19.2\n\n\n656.8\n\n\n18.6\n\n\n635.8\n\n\nTotal\n\n\nTotals for process routes depend on the feedstock and material flows and differ from plant to plant; totals should not be\nused to compare individual plants. Hot rolling values are based on energy use for production of hot rolled bars - see\nfollowing tables for data on hot rolling strip or wire.\n\n\nNote: Primary energy includes electricity generation, transmission, and distribution losses of 67%.\n\n\n2.1.1 Blast Furnace \u2013 Basic Oxygen Furnace Route\n\n\n-\n\n\nTable 2.1.3 provides best practice energy consumption values by fuel for the blast\nfurnace - basic oxygen furnace route. These values are based on the International Iron\nand Steel Institute's (IISI's) EcoTech plant which is defined as \"all those proven energy\nsaving technologies that are economically attractive\" except for the values for the basic\noxygen furnace which are based on IISI's AllTech plant which is defined as \u201call proven\nenergy saving technologies...regardless of financial viability\u201d.\u00b3\n\n\nFor this steelmaking route, the best practice calculations are based on the following\nassumptions: 1.389 t sinter are required to produce 1 t hot rolled steel, 90% pig iron and\n10% scrap, 0.9923 t pig iron required to produce 1 t hot rolled steel, 1.05 t crude steel\nrequired to make 1 t hot rolled steel.\n\n\n3 International Iron and Steel Institute (IISI) 1998. Energy Use in the Steel Industry. Brussels: IISI.\n\n\n6\n\n\nThe best practice coke plant is a modern coke plant using standard technology, including\nelectrical exhausters, high-pressure ammonia liquor spray for oven aspiration, as well as\nvariable speed drives on motors and fans. Coke dry quenching saves an additional 1.44\nGJ/t (49 kgce/t) coke (beyond the Ecotech value). The best practice does not include a\nJumbo Coke Reactor or non-recovery coke ovens. The best practice sinter plant is a state-\nof-the-art sinter plant using a bed depth of 500 mm on a moving grate, using coke and\nbreeze as fuel, and gas as ignition furnace fuel. Waste heat is recovered from the sinter\nexhaust cooler, and air leakage is controlled.\n\n\nDuring the ironmaking process, sintered or pelletized iron ore is reduced using coke in\ncombination with injected coal or oil to produce pig iron in a blast furnace.* Limestone is\nadded as a fluxing agent. Reduction of the iron ore is the largest energy-consuming\nprocess in the production of primary steel. The best practice blast furnace is a modern\nlarge scale blast furnace. Fuel injection rates are similar to modern practices found at\nvarious plants around the world (equivalent to about approximately 125 kg/t hot metal,\nslight oxygen enrichment, as well as pressurized operation (4 bar) allowing for power\nrecovery using a top gas power recovery turbine (wet type). Furthermore, the hot blast\nstoves have a heating efficiency of 85% using staggered parallel operation with three or\nfour stoves per furnace. Combustion air is preheated. The stoves use a mixture of coke\noven and blast furnace gas without oxygen enrichment.\n\n\nThe BOF process operates through the injection of oxygen, oxidizing the carbon in the\nhot metal. Several configurations exist depending on the way the oxygen is injected. The\nsteel quality can be improved further by ladle refining processes used in the steel mill.\nThe scrap input is rather small for the BOF-route, typically about 10-25%. The process\nneeds no net input of energy and can even be a net energy exporter in the form of BOF-\ngas and steam. In the best practice case BOF gas and sensible heat are recovered.\n\n\n4\n\n\nBest practice energy use is also determined by the concentration and quality of the ore used. As ore is\ntraded internationally (and to China), it is assumed that plants around the world have access to similar\nqualities of raw materials.\n\n\n7\n\n\nTable 2.1.3. World Best Practice Final and Primary Energy Intensity Values for the Blast\nFurnace \u2013 Basic Oxygen Furnace Steelmaking Route (values are per metric ton of steel)\n\n\nBlast Furnace - Basic Oxygen Furnace Route\n\n\nkgce/t\n\n\nGJ/t\n\n\n2.0\n\n\nSintering\n\n\nFuel\n\n\n67.8\n\n\nSteam\n\n\n-7.6\n\n\n-0.2\n\n\nElectricity\n\n\n5.1\n\n\n0.2\n\n\nFinal Energy\n\n\n65.2\n\n\n1.9\n\n\nMaterial Preparation\n\n\n74.3\n\n\nPrimary Energy\n\n\n2.2\n\n\nCoking\n\n\nFuel\n\n\n21.5\n\n\n0.6\n\n\nSteam\n\n\n3.5\n\n\n0.1\n\n\nElectricity\n\n\n3.5\n\n\n0.1\n\n\nFinal Energy\n\n\n28.6\n\n\n0.8\n\n\nPrimary Energy\n\n\n36.3\n\n\n1.1\n\n\nBlast Furnace\n\n\nFuel\n\n\n390.5\n\n\n11.4\n\n\nSteam\n\n\n13.6\n\n\n0.4\n\n\nElectricity\n\n\n0.1\n\n\nIronmaking\n\n\n3.1\n\n\nOxygen\n\n\n7.7\n\n\n0.2\n\n\nFinal Energy\n\n\n414.9\n\n\n12.2\n\n\nPrimary Energy\n\n\n423.7\n\n\n12.4\n\n\n-0.7\n\n\nBasic Oxygen Furnace\n\n\nFuel\n\n\n-25.4\n\n\nSteam\n\n\n-5.4\n\n\n-0.2\n\n\nElectricity\n\n\n3.4\n\n\n0.1\n\n\nOxygen\n\n\n12.1\n\n\n0.4\n\n\nFinal Energy\n\n\nSteelmaking\n\n\n-15.4\n\n\n-0.4\n\n\nPrimary Energy\n\n\n-9.5\n\n\n-0.3\n\n\nRefining\n\n\nElectricity\n\n\n4.3\n\n\n0.1\n\n\nFinal Energy\n\n\n4.3\n\n\n0.1\n\n\nPrimary Energy\n\n\n0.4\n\n\n13.0\n\n\nContinuous Casting\n\n\nFuel\n\n\n1.0\n\n\n0.0\n\n\nCasting\n\n\nElectricity\n\n\n0.9\n\n\n0.0\n\n\nFinal Energy\n\n\n2.0\n\n\n0.1\n\n\nPrimary Energy\n\n\n3.9\n\n\n0.1\n\n\nHot Rolling Strip\n\n\nFuel\n\n\n1.3\n\n\n44.8\n\n\nSteam\n\n\n0.7\n\n\n0.0\n\n\nElectricity\n\n\n10.2\n\n\n0.3\n\n\nFinal Energy\n\n\n55.7\n\n\n1.6\n\n\n2.2\n\n\nPrimary Energy\n\n\n76.5\n\n\nHot Rolling - Bars\n\n\nFuel\n\n\n53.8\n\n\n1.6\n\n\nHot Rolling\n\n\nElectricity\n\n\n8.8\n\n\n0.3\n\n\nFinal Energy\n\n\n62.5\n\n\n1.8\n\n\nPrimary Energy\n\n\n80.4\n\n\n2.4\n\n\nHot Rolling Wire\n\n\n57.3\n\n\nFuel\n\n\n1.7\n\n\nElectricity\n\n\n13.5\n\n\n0.4\n\n\nFinal Energy\n\n\n2.1\n\n\n70.9\n\n\n98.4\n\n\nPrimary Energy\n\n\n2.9\n\n\n14.9\n\n\nSub Total\n\n\nFuel\n\n\n509.1\n\n\n(based on hot rolling-bars)\n\n\nSteam\n\n\n4.2\n\n\n0.1\n\n\nElectricity\n\n\n29.1\n\n\n0.9\n\n\nOxygen\n\n\n19.8\n\n\n0.6\n\n\nFinal Energy\nPrimary Energy\n\n\n562.2\n\n\n16.5\n\n\n622.0\n\n\n18.2\n\n\nNote: Primary energy includes electricity generation, transmission, and distribution losses of 67%.\n\n\n8\n\n\nTable 2.1.3. (continued) World Best Practice Final and Primary Energy Intensity Values for\nthe Blast Furnace - Basic Oxygen Furnace Steelmaking Route (values are per metric ton of\nsteel)\n\n\nGJ/t\n\n\nBlast Furnace - Basic Oxygen Furnace Route\n\n\nkgce/t\n\n\nCold Rolling\n\n\nCold Rolling\n\n\nFuel\n\n\n1.8\n\n\n0.1\n\n\nSteam\n\n\n3.0\n\n\n0.1\n\n\nElectricity\n\n\n8.8\n\n\n0.3\n\n\nFinal Energy\n\n\n13.7\n\n\n0.4\n\n\nPrimary Energy\n\n\n32.1\n\n\n0.9\n\n\nFinishing\n\n\nFinishing\n\n\nFuel\n\n\n24.9\n\n\n0.7\n\n\nSteam\n\n\n0.3\n\n\n8.9\n\n\nElectricity\n\n\n4.3\n\n\n0.1\n\n\nFinal Energy\n\n\n38.1\n\n\n1.1\n\n\n48.4\n\n\n1.4\n\n\nPrimary Energy\n\n\nTotal\n\n\n535.8\n\n\nFuel\n\n\n15.7\n\n\n(based on hot rolling-bars)\n\n\nSteam\n\n\n16.0\n\n\n0.5\n\n\n42.3\n\n\nElectricity\n\n\n1.2\n\n\n0.6\n\n\nOxygen\n\n\n19.8\n\n\nFinal Energy\n\n\n613.9\n\n\n18.0\n\n\nPrimary Energy\n\n\n702.5\n\n\n20.6\n\n\nAlternative:\n\n\nReplace Continuous\nCasting, Hot Rolling, Cold\nRolling, and Finishing with\nThin Slab Casting\n\n\nCasting and Rolling\n\n\nFuel\n\n\n1.7\n\n\n0.1\n\n\nElectricity\n\n\n5.2\n\n\n0.2\n\n\nFinal Energy\n\n\n6.9\n\n\n0.2\n\n\nPrimary Energy\n\n\n17.3\n\n\n0.5\n\n\nTotal\n\n\nFuel\n\n\n456.0\n\n\n13.4\n\n\nSteam\n\n\n4.2\n\n\n0.1\n\n\nElectricity\n\n\n0.7\n\n\n24.5\n\n\nOxygen\nFinal Energy\nPrimary Energy\n\n\n19.8\n\n\n0.6\n\n\n504.5\n\n\n14.8\n\n\n555.1\n\n\n16.3\n\n\nNote: Primary energy includes electricity generation, transmission, and distribution losses of 67%.\n\n\n9\n\n\n2.1.2 Smelt Reduction - Basic Oxygen Furnace\n\n\nTable 2.1.4 provides best practice energy consumption values by fuel for the smelt\nreduction basic oxygen furnace (BOF) route. These values are based on the\nInternational Iron and Steel Institute's (IISI's) EcoTech plant which is defined as \u201call\nthose proven energy saving technologies that are economically attractive\".\n\n\n-\n\n\n5\n\n\nSmelt reduction processes are the latest development in pig iron production and omit\ncoke production by combining the gasification of coal with the melt reduction of iron\nore. Energy consumption is reduced because production of coke is abolished and iron\nore preparation is reduced. Processes under development include CCF, DIOS, AISI, and\nHISmelt. Currently, only the COREX process (Voest-Alpine, Austria) is commercial and\noperating in South Africa, South Korea and India, and under construction at Baosteel in\nChina (startup in 2007). The COREX process uses agglomerated ore, which is pre-\nreduced by gases coming from a hot bath. The pre-reduced iron is then melted in the bath.\nThe process produces excess gas, which is used for power generation, DRI-production, or\nas fuel gas. The FINEX technology allows the use of ore fines, but the first commercial\nplant is now under construction and hence not included in the best practice. Likewise the\nfirst commercial plant using the HISmelt process is under construction in Australia.\n\n\n6\n\n\nCurrently operating COREX plants show net energy consumption levels comparable to\nthe blast furnace routes. The coal consumption rate is higher than that of current blast\nfurnaces, but a large volume of offgas is produced that is used as fuel for power\ngeneration using a conventional steam cycle. The offgas can also be used to produce DRI\n(as in practice at Saldanha steel in South Africa), but this is not assumed as part of this\nbest practice.\n\n\nThe best practice values for the COREX plant are based on the commercially operating\nplant at POSCO's Pohang site in Korea. The plant coal consumption is around 29.4\nGJ/thm (100 kgce/t), 75 kWh/t (9.2 kgce/t) hot metal electricity and 526 Nm3/t hot metal\nof oxygen. It exports offgases with an energy value of 13.4 GJ/t (457 kgce/t) hot metal.\n\n\n8\n\n\nThe BOF process operates through the injection of oxygen, oxidizing the carbon in the\nhot metal. Several configurations exist depending on the way the oxygen is injected. The\nsteel quality can be improved further by ladle refining processes used in the steel mill.\nThe scrap input is rather small for the BOF-route, typically about 10-25%. The process\nneeds no net input of energy and can even be a net energy exporter in the form of BOF-\ngas and steam.\n\n\n5 International Iron and Steel Institute (IISI) 1998. Energy Use in the Steel Industry. Brussels: IISI.\n\n\n6\n\n\nBest practice energy use is also determined by the concentration and quality of the ore used; it is assumed\nthat all plants have access to similar qualities of raw materials.\nVoest Alpine Industrieanlagenbau, 1996. COREX, Revolution in Ironmaking, Linz, Austria: VAI.;de Beer,\nJ., Worrell, E., Blok, K., 1998. \"Future Technologies for Energy-Efficient Iron and Steel Making,\" Annual\nReview of Energy and Environment, 23: 123-205.\n\n\n7\n\n\n8 International Iron and Steel Institute (IISI) 1998. Energy Use in the Steel Industry. Brussels: IISI.\n\n\n10\n\n\nTable 2.1.4. World Best Practice Final and Primary Energy Intensity Values for the\nSmelt Reduction \u2013 Basic Oxygen Furnace Steelmaking Route (values are per metric\nton of steel)\n\n\n-\n\n\nSmelt Reduction- Basic Oxygen Furnace\n\n\nGJ/t\n\n\nkgce/t\n\n\nMaterial\nPreparation\n\n\nPelletizing\n\n\nFuel\nElectricity\nFinal Energy\n\n\n15.6\n\n\n0.5\n\n\n3.3\n\n\n0.1\n\n\n0.6\n\n\n19.0\n\n\nPrimary Energy\n\n\n25.7\n\n\n0.8\n\n\nSmelt Reduction\n\n\nFuel\n\n\n541.8\n\n\n15.9\n\n\nIronmaking\n\n\nElectricity\n\n\n9.1\n\n\n0.3\n\n\n40.6\n\n\nOxygen\n\n\n1.2\n\n\nFinal Energy\n\n\n591.6\n\n\n17.3\n\n\nPrimary Energy\n\n\n610.2\n\n\n17.9\n\n\nFuel\n\n\n-25.4\n\n\nBasic Oxygen Furnace\n\n\n-0.7\n\n\n-5.4\n\n\n-0.2\n\n\nSteam\n\n\nElectricity\n\n\n3.4\n\n\n0.1\n\n\nOxygen\n\n\n12.1\n\n\n0.4\n\n\nFinal Energy\n\n\nSteelmaking\n\n\n-15.4\n\n\n-0.4\n\n\n-9.5\n\n\n-0.3\n\n\nPrimary Energy\n\n\nElectricity\n\n\nRefining\n\n\n4.3\n\n\n0.1\n\n\nFinal Energy\n\n\n4.3\n\n\n0.1\n\n\nPrimary Energy\n\n\n0.4\n\n\n13.0\n\n\nContinuous Casting\n\n\n0.0\n\n\nCasting\n\n\nFuel\n\n\n1.0\n\n\n0.0\n\n\nElectricity\n\n\n0.9\n\n\nFinal Energy\n\n\n2.0\n\n\n0.1\n\n\nPrimary Energy\n\n\n0.1\n\n\n3.9\n\n\nHot Rolling Strip\n\n\nFuel\n\n\n44.8\n\n\n1.3\n\n\nSteam\n\n\n0.0\n\n\n0.7\n\n\nElectricity\n\n\n0.3\n\n\n10.2\n\n\nFinal Energy\n\n\n55.7\n\n\n1.6\n\n\n76.5\n\n\nPrimary Energy\n\n\n2.2\n\n\nFuel\n\n\n53.8\n\n\nHot Rolling - Bars\n\n\n1.6\n\n\nHot Rolling\n\n\nElectricity\n\n\n0.3\n\n\n8.8\n\n\nFinal Energy\n\n\n62.5\n\n\n1.8\n\n\nPrimary Energy\n\n\n80.4\n\n\n2.4\n\n\nHot Rolling Wire\n\n\nFuel\n\n\n57.3\n\n\n1.7\n\n\nElectricity\n\n\n13.5\n\n\n0.4\n\n\nFinal Energy\n\n\n2.1\n\n\n70.9\n\n\nPrimary Energy\n\n\n98.4\n\n\n2.9\n\n\nSub Total\n\n\nFuel\n\n\n586.8\n\n\n17.2\n\n\n(based on hot rolling-bars)\n\n\nSteam\n\n\n-5.4\n\n\n-0.2\n\n\nElectricity\n\n\n29.8\n\n\n0.9\n\n\n52.8\n\n\nOxygen\n\n\n1.5\n\n\nFinal Energy\nPrimary Energy\n\n\n664.0\n\n\n19.5\n\n\n723.7\n\n\n21.2\n\n\nNote: Primary energy includes electricity generation, transmission, and distribution losses of 67%.\n\n\n11\n\n\nTable 2.1.4. (continued) World Best Practice Final and Primary Energy Intensity\nValues for the Smelt Reduction \u2013 Basic Oxygen Furnace Steelmaking Route (values\nare per metric ton of steel)\n\n\n-\n\n\nSmelt Reduction- Basic Oxygen Furnace\n\n\nGJ/t\n\n\nkgce/t\n\n\nCold Rolling\n\n\nCold Rolling\n\n\nFuel\n\n\n1.8\n\n\n0.1\n\n\nSteam\n\n\n3.0\n\n\n0.1\n\n\nElectricity\n\n\n8.8\n\n\n0.3\n\n\nFinal Energy\n\n\n0.4\n\n\n13.7\n\n\nPrimary Energy\n\n\n32.1\n\n\n0.9\n\n\nFinishing\n\n\nFinishing\n\n\nFuel\n\n\n24.9\n\n\n0.7\n\n\nSteam\n\n\n8.9\n\n\n0.3\n\n\nElectricity\n\n\n4.3\n\n\n0.1\n\n\nFinal Energy\n\n\n38.1\n\n\n1.1\n\n\n48.4\n\n\nPrimary Energy\n\n\n1.4\n\n\nFuel\n\n\n18.0\n\n\n613.5\n\n\nTotal\n\n\n(based on hot rolling-bars)\n\n\nSteam\n\n\n6.5\n\n\n0.2\n\n\nElectricity\n\n\n43.0\n\n\n1.3\n\n\n52.8\n\n\nOxygen\n\n\n1.5\n\n\nFinal Energy\n\n\n715.8\n\n\n21.0\n\n\nPrimary Energy\n\n\n804.2\n\n\n23.6\n\n\nAlternative:\n\n\nReplace Continuous\nCasting, Hot Rolling, Cold\nRolling, and Finishing with\nThin Slab Casting\n\n\nCasting and Rolling\n\n\nFuel\n\n\n1.7\n\n\n0.1\n\n\nElectricity\n\n\n5.2\n\n\n0.2\n\n\nFinal Energy\n\n\n6.9\n\n\n0.2\n\n\nPrimary Energy\n\n\n17.3\n\n\n0.5\n\n\nFuel\n\n\n533.8\n\n\n15.6\n\n\nTotal\n\n\n-5.4\n\n\nSteam\n\n\n-0.2\n\n\n25.3\n\n\nElectricity\n\n\n0.7\n\n\nOxygen\nFinal Energy\n\n\n52.8\n\n\n1.5\n\n\n17.8\n\n\n606.4\n\n\nPrimary Energy\n\n\n656.8\n\n\n19.2\n\n\nNote: Primary energy includes electricity generation, transmission, and distribution losses of 67%.\n\n\n12\n\n\n2.1.3 Direct Reduced Iron - Electric Arc Furnace\n\n\nTable 2.1.5 provides best practice energy consumption values by fuel for the direct\nreduced iron - electric arc furnace (DRI \u2013 EAF) route. DRI, hot briquetted iron (HBI),\nand iron carbide are all alternative iron making processes.\" DRI, also called sponge iron,\nis produced by reduction of the ores below the melting point in small-scale plants (< 1\nMt/year) and has different properties than pig iron. DRI serves as a high-quality\nalternative for scrap in secondary steelmaking.\n\n\nDRI plants use either natural gas or coal as reductant. Globally, natural gas is widely\npreferred, and used in the leading processes Midrex and HyL-III. For China a coal-based\nprocess is more appropriate as natural gas availability is still limited to a small number of\nregions. The SL/RN process is the only commercial coal-based DRI process and in use in\nIndia and South Africa, as well as China. However, the new Circofer/Circored process\nhas now been demonstrated in Trinidad, while other technologies are actively being\ntested. For best practice the SL/RN process is used. The SL/RN process consumes 19.5\nGJ/t DRI (665 kgce/t DRI) of coal and 100 kWh of electricity. The waste heat from the\nfurnace is used to generate power in a conventional steam cycle, equivalent to 609 kWh/t\nDRI. Net power production is estimated at 509 kWh/t DRI (62.5 kgce/t DRI).\n\n\n10\n\n\nIn the EAF steelmaking process, the coke production, pig iron production, and steel\nproduction steps are omitted, resulting in much lower energy consumption. To produce\nEAF steel, scrap is melted and refined, using a strong electric current. The EAF can also\nbe fed with iron from the DRI route, but electricity consumption will increase by 40-120\nkWh/t liquid steel depending on the amount of DRI and degree of metallization of the\nDRI. DRI is used to enhance steel quality or if high quality scrap is scarce or expensive.\nSeveral process variations exist using either AC or DC currents, and fuels can be injected\nto reduce electricity use.\n\n\nThe best practice EAF plant is state-of-the-art facility with eccentric bottom tapping, ultra\nhigh power transformers, oxygen blowing, and carbon injection. The furnace uses a mix\nof 60% DRI and 40% high quality scrap. The high DRI charge rate limits the feasibility\nof fuel injection. The best practice excludes scrap preheating, although this is used in\nlarge scale furnaces. For the above charge the scrap preheater would achieve electricity\nsavings of 40 kWh/t liquid steel.\n\n\nThe best practice DRI-scrap-fed EAF consumes a mix of 60% DRI and 40% scrap. It\nconsumes 530 kWh/t (65 kgce/t) liquid steel for the EAF and 65 kWh/t (8 kgce/t) liquid\nsteel for gas cleaning and ladle refining, as well as 8 kg/t liquid steel of carbon. Installing\na scrap preheater will reduce power use in the EAF by 40 kWh/t (4.9 kgce/t) liquid steel,\nreducing total electricity use to 555 kWh/t (68.2 kgce/t) liquid steel.\n\n\n9 McAloon, T.P., 1994. \"Alternate Ironmaking Update,\u201d Iron & Steelmaker 21(2): 37-39 + 55.\n10 International Iron and Steel Institute (IISI) 1998. Energy Use in the Steel Industry. Brussels: IISI.\n\n\n13\n\n\nTable 2.1.5. World Best Practice Final and Primary Energy Intensity Values for Direct\nReduced Iron \u2013 Electric Arc Furnace Route (values are per metric ton of steel)\n\n\nDirect Reduced Iron - Electric Arc Furnace Route\n\n\nkgce/t\n\n\nGJ/t\n\n\nSintering\n\n\nFuel\n\n\n2.0\n\n\n67.8\n\n\n-7.6\n\n\nSteam\n\n\n-0.2\n\n\nElectricity\n\n\n0.2\n\n\n5.1\n\n\nFinal Energy\n\n\n65.2\n\n\n1.9\n\n\n2.2\n\n\nMaterial\n\n\nPrimary Energy\n\n\n74.3\n\n\nPelletizing\n\n\nPreparation\n\n\nFuel\n\n\n15.6\n\n\n0.5\n\n\nElectricity\n\n\n3.3\n\n\n0.1\n\n\nFinal Energy\n\n\n0.6\n\n\n19.0\n\n\n25.7\n\n\n0.8\n\n\nPrimary Energy\n\n\nDirect Reduced Iron\n\n\nFuel\n\n\n440.9\n\n\n12.9\n\n\nElectricity\n\n\n-41.4\n\n\nIronmaking\n\n\n-1.2\n\n\nFinal Energy\n\n\n399.6\n\n\n11.7\n\n\nPrimary Energy\n\n\n315.6\n\n\n9.2\n\n\nElectric Arc Furnace\n\n\nFuel\n\n\n19.2\n\n\n0.6\n\n\nSteelmaking\n\n\n57.8\n\n\nElectricity\n\n\n1.7\n\n\n8.6\n\n\nOxygen\n\n\n0.3\n\n\nFinal Energy\n\n\n85.6\n\n\n2.5\n\n\nPrimary Energy\n\n\n202.9\n\n\n5.9\n\n\nContinuous Casting\n\n\nCasting\n\n\nFuel\n\n\n1.0\n\n\n0.03\n\n\nElectricity\n\n\n0.03\n\n\n0.9\n\n\nFinal Energy\n\n\n2.0\n\n\n0.1\n\n\nPrimary Energy\n\n\n3.9\n\n\n0.1\n\n\nHot Rolling Strip\n\n\nFuel\n\n\n1.3\n\n\n44.8\n\n\nSteam\n\n\n0.7\n\n\n0.02\n\n\nElectricity\n\n\n0.3\n\n\n10.2\n\n\nFinal Energy\n\n\n55.7\n\n\n1.6\n\n\n76.5\n\n\n2.2\n\n\nPrimary Energy\n\n\nHot Rolling - Bars\n\n\n53.8\n\n\n1.6\n\n\nFuel\n\n\nHot Rolling\n\n\nElectricity\n\n\n8.8\n\n\n0.3\n\n\nFinal Energy\n\n\n62.5\n\n\n1.8\n\n\n2.4\n\n\n80.4\n\n\nPrimary Energy\n\n\nHot Rolling Wire\n\n\nFuel\n\n\n57.3\n\n\n1.7\n\n\nElectricity\n\n\n13.5\n\n\n0.4\n\n\nFinal Energy\n\n\n2.1\n\n\n70.9\n\n\nPrimary Energy\n\n\n98.4\n\n\n2.9\n\n\nTotal\n\n\nFuel\n\n\n598.3\n\n\n17.5\n\n\n(based on hot rolling-bars)\n\n\n-7.6\n\n\nSteam\n\n\n-0.2\n\n\nElectricity\n\n\n1.0\n\n\n34.6\n\n\nOxygen\n\n\n8.6\n\n\n0.3\n\n\nFinal Energy\nPrimary Energy\n\n\n18.6\n\n\n633.9\n\n\n702.7\n\n\n20.6\n\n\nNote: Primary energy includes electricity generation, transmission, and distribution losses of 67%.\n\n\n14\n\n\nTable 2.1.5. (continued) World Best Practice Final and Primary Energy Intensity Values for\nDirect Reduced Iron \u2013 Electric Arc Furnace Route (values are per metric ton of steel)\n\n\n-\n\n\nDirect Reduced Iron - Electric Arc Furnace Route\n\n\nkgce/t\n\n\nGJ/t\n\n\nAlternative:\nCasting and Rolling\n\n\nReplace Continuous\nCasting, Hot Rolling, Cold\nRolling, and Finishing with\nThin Slab Casting\n\n\nFuel\n\n\n1.7\n\n\n0.1\n\n\nElectricity\n\n\n5.2\n\n\n0.2\n\n\nFinal Energy\n\n\n6.9\n\n\n0.2\n\n\n17.3\n\n\n0.5\n\n\nPrimary Energy\n\n\nTotal\n\n\nFuel\n\n\n545.2\n\n\n16.0\n\n\nSteam\n\n\n-7.6\n\n\n-0.2\n\n\nElectricity\n\n\n30.0\n\n\n0.9\n\n\nOxygen\nFinal Energy\nPrimary Energy\n\n\n8.6\n\n\n0.3\n\n\n576.2\n\n\n16.9\n\n\n635.8\n\n\n18.6\n\n\nNote: Primary energy includes electricity generation, transmission, and distribution losses of 67%.\n\n\n2.1.4 Electric Arc Furnace\n\n\nIn the EAF steelmaking process, the coke production, pig iron production, and steel\nproduction steps are omitted, resulting in much lower energy consumption. To produce\nEAF steel, scrap is melted and refined, using a strong electric current. Several process\nvariations exist, using either AC or DC currents and fuels can be injected to reduce\nelectricity use.\n\n\nTable 2.1.6 provides best practice energy consumption values by fuel for the EAF route.\nThe best practice EAF plant is state-of-the-art facility using 100% high quality scrap. The\nEAF is equipped with eccentric bottom tapping, ultra high power transformers, oxygen\nblowing, full foamy slag operation, oxy-fuel burners, and carbon injection. Scrap\npreheating is not assumed, although economically attractive, especially for large scale\nfurnaces. Scrap preheating will reduce power consumption by 70 kWh/t (8.6 kgce/t)\nliquid steel.\n\n\nThe \u201cbest practice\u201d DRI-scrap-fed EAF consumes 100% scrap. It consumes 409 kWh/t\n(50.3 kgce/t) liquid steel for the EAF and 65 kWh/t (8 kgce/t) liquid steel for gas cleaning\nand ladle refining, as well as 0.15 GJ/t (5.1 kgce/t) liquid steel of natural gas and 8 kg/t\nliquid steel of carbon. Installing a scrap preheater would reduce power use in the EAF by\n70 kWh/t (8.6 kgce/t), reducing total electricity use to 404 kWh/t (49.6 kgce/t) liquid\nsteel.\n\n\n15\n\n\nTable 2.1.6. World Best Practice Final and Primary Energy Intensity Values for Electric\nArc Furnace Route (values are per metric ton of steel)\n\n\nElectric Arc Furnace Route\n\n\nkgce/t\n\n\nGJ/t\n\n\nFuel\n\n\nElectric Arc Furnace\n\n\n19.2\n\n\n0.6\n\n\nSteelmaking\n\n\nElectricity\n\n\n1.5\n\n\n52.8\n\n\n0.3\n\n\nOxygen\n\n\n8.6\n\n\nFinal Energy\n\n\n80.6\n\n\n2.4\n\n\nPrimary Energy\n\n\n187.7\n\n\n5.5\n\n\nContinuous Casting\n\n\nCasting\n\n\nFuel\n\n\n1.0\n\n\n0.03\n\n\nElectricity\n\n\n0.9\n\n\n0.03\n\n\nFinal Energy\n\n\n0.1\n0.1\n\n\n2.0\n\n\nPrimary Energy\n\n\n3.9\n\n\nHot Rolling Strip\n\n\nFuel\n\n\n44.8\n\n\n1.3\n\n\n0.02\n\n\nSteam\n\n\n0.7\n\n\n10.2\n\n\nElectricity\n\n\n0.3\n\n\nFinal Energy\n\n\n1.6\n\n\n55.7\n\n\n76.5\n\n\n2.2\n\n\nPrimary Energy\n\n\n1.6\n\n\nHot Rolling - Bars\n\n\n53.8\n\n\nFuel\n\n\nHot Rolling\n\n\nElectricity\n\n\n8.8\n\n\n0.3\n\n\nFinal Energy\n\n\n62.5\n\n\n1.8\n\n\nPrimary Energy\n\n\n80.4\n\n\n2.4\n\n\nHot Rolling Wire\n\n\nFuel\n\n\n1.7\n\n\n57.3\n\n\nElectricity\n\n\n13.5\n\n\n0.4\n\n\nFinal Energy\n\n\n70.9\n\n\n2.1\n\n\nPrimary Energy\n\n\n98.4\n\n\n2.9\n\n\n2.2\n\n\nTotal\n\n\nFuel\n\n\n74.0\n\n\n(based on hot rolling-bars)\n\n\nElectricity\n\n\n62.5\n\n\n1.8\n\n\nOxygen\n\n\n8.6\n\n\n0.3\n\n\nFinal Energy\n\n\n145.1\n\n\n4.3\n\n\nPrimary Energy\n\n\n272.0\n\n\n8.0\n\n\nAlternative:\n\n\nReplace Continuous\nCasting, Hot Rolling, Cold\nRolling, and Finishing with\nThin Slab Casting\n\n\nCasting and Rolling\n\n\nFuel\n\n\n1.7\n\n\n0.1\n\n\nElectricity\n\n\n5.2\n\n\n0.2\n\n\nFinal Energy\n\n\n6.9\n\n\n0.2\n\n\n17.3\n\n\nPrimary Energy\n\n\n0.5\n\n\nTotal\n\n\n0.6\n\n\n20.9\n\n\nFuel\n\n\nElectricity\n\n\n57.9\n\n\n1.7\n\n\nOxygen\n\n\n8.6\n\n\n0.3\n\n\nFinal Energy\nPrimary Energy\n\n\n87.5\n\n\n2.6\n\n\n205.1\n\n\n6.0\n\n\nNote: Primary energy includes electricity generation, transmission, and distribution losses of 67%.\n\n\n2.1.5 Casting\n\n\nContinuous casting values are based on the International Iron and Steel Institute's\nEcoTech plant which includes \u201call those proven energy saving technologies that are\neconomically attractive\"11\nand the thin slab/near net shape casting values are based on\nWorrell et al. (2004). 12 Casting can be either continuous casting or thin slab/near net\n\n\n11 International Iron and Steel Institute (IISI) 1998. Energy Use in the Steel Industry. Brussels: IISI.\n\n\n12\n\n\nWorrell, E., Price, L., and Galitsky, C., 2004. \u201cEmerging Energy-Efficient Technologies in Industry:\nCase Study of Selected Technologies,\u201d Technical Appendix Chapter 3: Improving Energy Efficiency of\n\n\n16\n\n\nshape casting. Best practice continuous casting uses 0.06 GJ/t (2.0 kgce/t) steel of final\nenergy. Energy is only used to dry and preheat the ladles, heat the tundish, and for\nmotors to drive the casting equipment. Thin slab/near net shape casting is a more\nadvanced casting technique which reduces the need for hot rolling because products are\ninitially cast closer to their final shape using a simplified rolling strand positioned behind\nthe caster's reheating tunnel furnace, eliminating the need for a separate hot rolling mill.\nFinal energy used for casting and rolling using thin slab casting is 0.20 GJ/t (6.9 kgce/t)\nsteel.\n\n\n13\n\n\n2.1.6 Rolling and Finishing\n\n\nHot Rolling\n\n\nRolling of the cast steel begins in the hot rolling mill where the steel is heated and passed\nthrough heavy roller sections to reduce the thickness. Best practice values for hot rolling\nare 1.55 GJ/t (53.0 kgce/t), 1.75 GJ/t (59.6 kgce/t), and 1.98 GJ/t (67.5 kgce/t) of steel of\nfinal energy for rolling strip, bars, and wire, respectively. 14 Electricity consumption for\nthe best practice hot strip mill is based on hot strip mill 2 at Corus, IJmuiden,\nNetherlands. The best practice values assume 100% cold charging, a walking beam\nfurnace with furnace controls and energy efficient burners, and efficient motors. Hot\ncharging and premium efficiency motors may further reduce the rolling mill energy use.\n\n\n15\n\n\nCold Rolling\n\n\nThe hot rolled sheets may be further reduced in thickness by cold rolling. The coils are\nfirst treated in a pickling line followed by treatment in a tandem mill. The best practice\nfinal energy intensity for cold rolling is 0.09 GJ/t (3.0 kgce/t) steam, fuel use of 0.053\nGJ/t (1.8 kgce/t) and electricity use of 87 kWh/t (10.7 kgce/t) cold rolled sheet,\"\nequivalent to 0.47 GJ/t (13.7 kgce/t) cold sheet.\n\n\n16\n\n\nFinishing\n\n\nFinishing is the final production step, and may include different processes such as\nannealing and surface treatment. The best practice final energy intensity for batch\nannealing is steam use of 0.173 GJ/t, fuel use of 0.9 GJ/t and 35 kWh/t of electricity,\nequivalent to 1.2 GJ/t (41.0 kgce/t). Best practice energy use for continuous annealing is\nassumed to be equal to fuel use of 0.73 GJ/t, steam use of 0.26 GJ/t, and electricity use of\n35 kWh/t, equivalent to final energy use of 1.1 GJ/t (or 38.1 kgce/t). Continuous\nannealing is considered the state-of-the-art technology, and therefore assumed to be best\npractice technology.\n\n\nNational Commission on Energy Policy report Ending the Energy Stalemate: A Bipartisan Strategy to Meet\nAmerica's Energy Challenges (http://www.energycommission.org/)\n\n\n13 International Iron and Steel Institute (IISI) 1998. Energy Use in the Steel Industry. Brussels: IISI.\n14 International Iron and Steel Institute (IISI) 1998. Energy Use in the Steel Industry. Brussels: IISI.\n\n\n15 Worrell, E. 1994. Potentials for Improved Use of Industrial Energy and Materials, Ph.D. Thesis, Utrecht\nUniversity, June 1994.\n\n\n16 International Iron and Steel Institute (IISI) 1998. Energy Use in the Steel Industry. Brussels: IISI.\n\n\n17\n\n\n2.2. Aluminium\n\n\nThere are five steps in the primary aluminium production process: bauxite extraction,\nalumina production, anode manufacture, aluminium smelting, and ingot casting. This\nassessment excludes bauxite extraction because the energy use will primarily depend on\nthe ore deposit characteristics. Secondary aluminium production is based on melting and\nreshaping scrap aluminium. Table 2.2.1 provides best practice final energy intensity\nvalues for the process steps for primary aluminium production along with the best\npractice energy intensity value for secondary aluminium production. Table 2.2.2 provides\nprimary energy values for these two aluminium production processes.\n\n\nTable 2.2.1. World Best Practice Final Energy Intensity Values for Aluminium Production\n(values are per metric tonne aluminium).\n\n\nPrimary\nAluminium\n\n\nSecondary\nAluminium\n\n\nGJ/t kgce/t\n\n\nkgce/t\n\n\nGJ/t\n\n\nDigesting (fuel)\n\n\n12.1\n\n\nAlumina Production\n\n\n414\n\n\nCalcining Kiln (fuel)\n\n\n6.5\n\n\n223\n\n\n(Bayer)\n\n\n1.4\n\n\nElectricity\n\n\n48\n\n\nAnode Manufacture\n\n\n35\n\n\nFuel\n\n\n1.0\n\n\nElectricity\n\n\n(Carbon)\n\n\n7\n\n\n0.21\n\n\nAluminium Smelting (Electrolysis)\nIngot Casting\n\n\n1671\n\n\nElectricity\n\n\n49.0\n\n\nElectricity\n\n\n12\n\n\n0.35\n\n\nTotal\n\n\n85\n\n\n2.5\n\n\n2411\n\n\n70.6\n\n\nTable 2.2.2. World Best Practice Primary Energy Intensity Values for Aluminium\nProduction (values are per metric tonne aluminium).\n\n\nPrimary\nAluminium\n\n\nSecondary\nAluminium\n\n\nkgce/t GJ/t kgce/t\n\n\nGJ/t\n\n\nAlumina Production\n\n\nDigesting (fuel)\n\n\n414\n\n\n12.1\n\n\nCalcining Kiln (fuel)\n\n\n223\n\n\n(Bayer)\n\n\n6.5\n\n\nElectricity\n\n\n145\n\n\n4.3\n\n\nFuel\n\n\n35\n\n\nAnode Manufacture\n\n\n1.0\n\n\n(Carbon)\n\n\nElectricity\n\n\n22\n\n\n0.64\n\n\nAluminium Smelting (Electrolysis)\n\n\nElectricity\n\n\n5064\n\n\n148.4\n\n\n36\n\n\nIngot Casting\n\n\nElectricity\n\n\n1.06\n\n\n174.0\n\n\nTotal\n\n\n5940\n\n\n259\n\n\n7.6\n\n\nNote: Primary energy includes electricity generation, transmission, and distribution losses of 67%.\n\n\n18\n\n\n2.2.1 Alumina Production\n\n\nBauxite ore 17 is converted to alumina through the Bayer process, whereby the ore is\ncrushed and dissolved in a hot sodium hydroxide solution. Iron oxides and other oxides\nare removed as insoluble \u201cred mud\" and the solution is precipitated and then calcined to\nproduce anhydrous alumina. The Bayer process is energy-intensive, especially the\ndigestion and calcination processes. Calcination can be done in rotary or stationary kilns.\nThe resulting alumina is cooled in rotary or satellite coolers, or fluidized bed coolers.\n\n\nElectricity and fuel account for an average of 13% and 85% of total energy, respectively.\nBest practice electricity use of an alumina plant is estimated to be 203 kWh/t (24.9\nkgce/t) alumina or 391 kWh/t (48 kgce/t) aluminium assuming 1.925 t alumina equals 1 t\naluminium. Energy use for digesting can vary between 6.3 and 12.6 GJ/t (215 and 430\nkgce/t) alumina or 12.1 and 24.3 GJ/t (414 and 828 kgce/t) aluminium, while the fuel\nconsumption for the calcining kiln will vary from 3.4 GJ/t to 4.2 GJ/t (116 to 143 kgce/t)\nalumina or 6.5 to 8.1 GJ/t (223 to 276 kgce/t) aluminium.\u201d\n\n\n18\n\n\nFor the best practice alumina Bayer plant for alumina production, total fuel consumption\nof 9.7 GJ/t (331 kgce/t) alumina or 18.7 GJ/t (637 kgce/t) aluminium and electricity\nconsumption of 203 kWh/t alumina, for a total consumption of 10.4 GJ/t (356 kgce/t)\nalumina or 20.1 GJ/t (685 kgce/t) aluminium is assumed.\n\n\n2.2.2 Anode Manufacture\n\n\nThe most energy-efficient aluminum electrolysis process uses pre-baked anodes. While\nresearch is ongoing in the development of inert anodes (consisting of Titanium boride,\nTiB2) these technologies are not yet commercially used. Hence this report assumes that\nbest practice is pre-baked carbon anodes.\n\n\nAnodes are produced by heating ground and pressed tar pitch or coke from refineries at\nhigh temperatures in gas-heated furnaces. Anodes can be produced onsite at the smelter\nor in separate plants specialized in the manufacture of carbon anodes for various\nindustries and applications.\n\n\nThe furnaces can be fired with any fuel. In most countries natural gas is used as fuel. The\nspecific fuel consumption for anode production is estimated to be 2.45 GJ/t (84 kgce/t)\nanode and 140 kWh/t anode. The most efficient smelters consume 400-440 kg of\n\n\n19, 20\n\n\n17\n\n\nBest practice energy use is also determined by the concentration and quality of the ore used; it is assumed\nthat all plants have access to similar qualities of raw materials.\nWorrell, E. and de Beer, J., 1991. Energy Requirements in Relation to Prevention and Re-Use of Waste\nStreams. Report: (in Dutch). Utrecht, The Netherlands: Novem.\n19 International Aluminium Institute (2003). Lifecycle Assessment of Aluminium: Inventory Data for the\nWorldwide Primary Aluminium Industry.\n20 Worrell, E. and de Beer, J., 1991. Energy Requirements in Relation to Prevention and Re-Use of Waste\nStreams. Report: Aluminium (in Dutch). Utrecht, The Netherlands: Novem.\n\n\n18\n\n\n19\n\n\nanode per tonne of aluminium.21 Assuming 0.42 t anode to produce 1 t aluminium, fuel\nconsumption is 1.0 GJ/t (35 kgce/t) aluminium and electricity consumption is 0.21 GJ/t (7\nkgce/t) aluminium for a total consumption for anode manufacture of 1.2 GJ/t (42 kgce/t)\naluminium.\n\n\n2.2.3 Aluminum Smelting (Electrolysis)\n\n\nThe Hall-Heroult process serves as the basis for commercial aluminum smelting. The\naluminum industry currently uses two types of smelting technology: cells with prebaked\nanodes and cells with baked-in-situ anodes (S\u00f8derberg). Over the years five aluminum\nsmelter types have become widespread:\n\n\nIn-situ (S\u00f8derberg):\n\n\nVertical Stud S\u00f8derberg (VSS)\nHorizontal Stud S\u00f8derberg (HSS).\n\n\n-\n\n\nPre-baked:\n\n\nPoint Feed Prebake (PFPB)\n\n\nCenter Feed Prebake (CFPB)\nSide Work Prebake (SWPB).\n\n\nS\u00f8derberg cell plants are more energy-intensive and environmentally problematic than\nplants that use prebaked cells, and are hence not considered best practice. The best\npractice technology is a Center-Feed Prebaked (CFBP) cell. The current best practice\nCFPB designs use 300-315 kA currents (current densities of 0.8 - 0.85 A/cm\u00b2), and\nconsume 400-440 kg anode/t aluminium.\n\n\n22\n\n\nThe lowest theoretical energy requirement for electrolysis is 6,360 kWh per t of\naluminium roduct. 23,24 However, no current cell design comes close to the\nthermodynamic minimum. The current best practice of Hall-Heroult electrolysis cells\n(using currents of 300-315 kA) is estimated to be 12.9 to 13.0 MWh/t aluminium. Losses\nof rectifiers, auxillaries, and pollution control demand an additional 0.7 - 1.0 MWh/t\nprimary aluminum. Hence, the total best practice energy consumption of the aluminum\nsmelter is estimated to be 13.6 MWh/t or 49 GJ/t (1671 kgce/t) aluminium, including all\nutilities.\n\n\n2.2.4 Ingot Casting\n\n\nThe molten aluminium is most often cast into ingots. Ingot casting also allows the\naluminium to be alloyed with other metals to produce a specific alloy. Ingots can have\nvarious shapes and forms (e.g. slabs, rolls, bars, and blocks). After casting the ingots are\n\n\n21\n\n\nEuropean Commission. Integrated Pollution Prevention & Control: Reference Document on Best\nAvailable Techniques in the Non-Ferrous Metals Industries. Brussels/Sevilla, December 2001.\n22 European Commission. Integrated Pollution Prevention & Control: Reference Document on Best\nAvailable Techniques in the Non-Ferrous Metals Industries. Brussels/Sevilla, December 2001.\n\n\n23 Choate, W.T., Green, J.A.S., 2003. U.S. Energy Requirements for Aluminum Production: Historical\n\n\nPerspectives, Theoretical Limits and New Opportunities. Washington DC: BCS, Inc.\n\n\n24 Beck, T.R., 2001. \"Electrolytic Production of Aluminum,\u201d Electrochemistry Encyclopedia\nElectrochemical Technology Corp. (http://electrochem.cwru.edu/ed/encycl/art-a01-al-prod.htm)\n\n\n20\n20\n\n\ncooled and transported to the end user. The end user may process the ingots to the final\nproduct through casting and rolling (e.g. sheets, castings).\n\n\nAlloying takes place in a furnace. The furnace can be heated by fuel or electrically. Best\npractice electricity use is estimated to be 0.35 GJ/t (12 kgce/t) aluminium ingot, assuming\ndirect casting with aluminium transferred hot to the alloying furnace.25 In practice, energy\nconsumption will depend on the aluminium temperature, holding time and casting\nsequence.\n\n\n2.2.5 Secondary Aluminium Production\n\n\nSecondary smelting of aluminium using scrap only requires roughly 5% of the energy of\nprimary smelting due to the relatively low melting temperature of 700-800 \u00b0C. Secondary\naluminium may not be suitable for all applications because the purity of the product is\nharder to control since the scrap may consist of many different alloying elements, and\nsome elements are hard to remove.\n\n\nVarious technologies are used to recycle aluminium scrap, including reverbatory and\ninduction furnaces. A number of new and emerging technologies are being investigated\nincluding rotary arc and plasma furnaces. The choice of the most appropriate technology\nwill depend in the scrap to be used.\n\n\nThe theoretical energy consumption for aluminium melting is 1.1 GJ/t (38 kgce/t).\nHowever, no melting furnace comes close to this level. The best practice assumes a\nnatural gas fired reverbatory furnace. Reverbatory furnaces consume between 3 and 9\nGJ/t (102 and 307 kgce/t) of fuel. For the best practice performance a large reverbatory\nfurnace using recuperative burners and state-of-the-art computer controls, consuming 2.5\nGJ (85 kgce) of natural gas/t aluminium, is assumed.\n\n\n26\n\n\n25 International Aluminium Institute (2003). Lifecycle Assessment of Aluminium: Inventory Data for the\nWorldwide Primary Aluminium Industry.\n26 Flannagan, J.M., 1993. Process Heating in the Metals Industry. Sittard, The Netherlands: IEA-Caddet.\n\n\n21\n\n\n2.3 Cement\n\n\nBest practice values for each step of the cement making process raw materials\npreparation (limestone and fuels), clinker making (fuel use and electricity use), additive\ndrying, cement grinding and, where applicable, other production energy, which includes\nquarrying, auxiliaries, conveyors and packaging - are provided for final energy in Tables\n2.3.1 to 2.3.3 and for primary energy in Tables 2.3.4 to 2.3.5. Other non-production\nenergy (lighting, office equipment, etc.) is based on production throughput and a study\ndone by Warshawsky (1996). 27\n\n\n-\n\n\nBecause clinker making accounts for about 90% of the energy consumed in the cement\nmaking process, reducing the ratio of clinker to final cement by mixing clinker with\nadditives can greatly reduce the energy used for manufacture of cement. Best practice\nvalues for additive use are based on the following European ENV 197-2 standards: for\ncomposite Portland cements (CEM II), up to 35% can be fly ash and 65% clinker; for\nblast furnace slag cements (CEM III/A), up to 65% can be blast furnace slag and 35%\nclinker. Best practice final and primary energy use for three types of cement (Portland\ncement, fly ash cement and blast furnace slag cement) are given in Tables 2.3.1 to 2.3.6.\n\n\n28\n\n\n2.3.1 Raw Materials and Fuel Preparation\n\n\nEnergy used in preparing the raw material consists of crushing, grinding and drying (if\nnecessary) the raw meal which is mostly limestone. Solid fuels input to the kiln must also\nbe crushed, ground, and dried. Best practice for raw materials preparation is based on the\nuse of a gyratory crusher as 0.38 kWh/t raw meal, 2 a longitudinal preblending store with\neither bridge scraper or bucket wheel reclaimer or a circular preblending store with\nbridge scraper reclaimer for preblending (prehomogenization and proportioning) at 0.5\nkWh/t raw meal, an integrated vertical roller mill system with four grinding rollers and\na high-efficiency separator at 11.45 kWh/t raw meal for grinding,\u00b3 and a gravity (multi-\noutlet silo) dry system at 0.10 kWh/t raw meal for homogenization. 32 Based on the above\nvalues, the overall best practice value for raw materials preparation is 12.05 kWh/t raw\nmaterial. Ideally this value should take into account the differences in moisture content of\nthe raw materials as well as the hardness of the limestone. Higher moisture content\nrequires more energy for drying and harder limestone requires more crushing and\ngrinding energy. If drying is required, best practice is to install a preheater to dry the raw\nmaterials, which decreases the efficiency of the kiln. For this analysis, it is assumed that\npre-heating of wet raw materials is negligible and does not decrease the efficiency of the\nkiln.\n\n\n29\n\n\n30\n\n\n31\n\n\n27 Warshawsky, J. of CMP. 1996. TechCommentary: Electricity in Cement Production. EPRI Center for\nMaterials Production, Carnegie Mellon Research Institute, Pittsburgh, PA.\n\n\n28 CEM I is Portland cement, set at \u2264 5% additives, 95% clinker.\n\n\n29 Portland Cement Association, 2004. Innovations in Portland Cement Manufacturing. Skokie, IL: PCA.\n\n\n30 Cembureau, 1997. Best Available Techniques for the Cement Industry, Brussels: Cembureau.\n\n\n31 Schneider, U., \"From ordering to operation of the first quadropol roller mill at the Bosenberg Cement\nWorks,\" ZKG International, No.8, 1999: 460-466.\n\n\n32 Portland Cement Association, 2004. Innovations in Portland Cement Manufacturing. Skokie, IL: PCA.\n\n\n22\n22\n\n\nSolid fuel preparation also depends on the moisture content of the fuel. It is assumed that\nonly coal needs to be dried and ground and that the energy required for drying or grinding\nof other materials is insignificant or unnecessary. Best practice is to use the waste heat\nfrom the kiln system, e.g., the clinker cooler (if available) to dry the coal. 55 Best practice\nusing a MPS vertical roller mill is 10-36 kWh/t anthracite, 6-12 kWh/t pit coal, 8-19\nkWh/t lignite, and 7-17 kWh/t petcoke 34 or using a bowl mill is 10-18 kWh/t product.\u00b3\nBased on the above, it is assumed that best practice for solid fuel preparation is 10 kWh/t\nproduct.\n\n\n33\n\n\n35\n\n\n2.3.2 Clinker Production\n\n\nClinker production can be split into the electricity required to run the machinery,\nincluding the fans, the kiln drive, the cooler and the transport of materials to the top of\nthe preheater tower, and the fuel needed to dry, to calcine and to clinkerize the raw\nmaterials. Best practice for clinker making mechanical requirements is estimated to be\n22.5 kWh/t clinker,36 while fuel use has been reported as low as 2.85 GJ/t (97.3 kgce/t)\nclinker.\n\n\n37\n\n\n2.3.3 Additive Preparation\n\n\nIn addition to clinker, some plants use additives in the final cement product. While this\nreduces the most energy intensive stage of production (clinker making), as well as the\ncarbonation process which produces additional CO2 as a product of the reaction,\nadditional electricity is required to blend and grind the additives, while additional fuel is\nrequired to dry some additives like blast furnace and other slags.\n\n\n33 Worrell, E. and Galitsky, C., 2004. Energy Efficiency Improvement Opportunities for Cement Making:\nAn ENERGY STAR\u00ae Guide for Energy and Plant Managers. Berkeley, CA: Lawrence Berkeley National\nLaboratory (LBNL-54036).\nKraft, B. and Reichardt, Y., 2005. \u201cGrinding of Solid Fuels Using MPS Vertical Roller Mills,\" ZKG\nInternational 58:11 (pp 36-47).\n\n\n34\n\n\n35 Portland Cement Association, 2004. Innovations in Portland Cement Manufacturing. Skokie, IL: PCA.\n\n\n36 COWIconsult, March Consulting Group and MAIN, 1993. Energy Technology in the Cement Industrial\nSector, Report prepared for CEC - DG-XVII, Brussels, April.\n37 Park, H. 1998. Strategies for Assessing Energy Conservation Potentials in the Korean Manufacturing\nSector. In: Proceedings 1998 Seoul Conference on Energy Use in Manufacturing: Energy Savings and CO2\nMitigation Policy Analysis. 19-20 May, POSCO Center, Seoul, Republic of Korea.\n\n\n23\n\n\nTable 2.3.1. World Best Practice Final Energy Intensity Values for Portland Cement\n\n\nGJ/t kWh/t\nclinker\n\n\nGJ/t\n\n\nGJ/t\nproduct\n\n\nkWh/t kgce/t\nproduct product\n\n\nkgce/t\n\n\nProduct\nunit\n\n\nkWh/t kgce/t\nclinker clinker\n\n\ncement\n\n\ncement\n\n\ncement\n\n\nRaw Materials Preparation\n\n\nt raw meal\nt coal\n\n\nElectricity\nElectricity\n\n\n0.07\n\n\n1.5\n\n\n0.04\n\n\n2.62\n\n\n12.05\n\n\n21.3\n\n\n0.08\n\n\n20.3\n\n\n2.49\n\n\nSolid Fuels Preparation\n\n\n0.12\n\n\n10\n\n\n1.2\n\n\n0.04\n\n\n0.97\n\n\n0.92\n\n\n0.11\n\n\nClinker Making\n\n\nt clinker\n\n\nFuel\n\n\n97\n\n\n2.85\n\n\n2.71\n\n\n92\n\n\nElectricity\n\n\nt clinker\n\n\n2.76\n\n\n22.5\n\n\n0.08\n\n\n2.63\n\n\n0.08\n\n\n21.4\n\n\nAdditives Preparation\n\n\nt additive\n\n\nFuel\n\n\nt additive\n\n\nElectricity\n\n\nCement grinding\n\n\n325 cement Electricity\n\n\nt cement\n\n\n16\n\n\n2.0\n\n\n0.06\n\n\n425 cement\n525 cement\n625 cement\n\n\nElectricity\nElectricity\nElectricity\n\n\nt cement\n\n\n17.3\n\n\n2.1\n\n\n0.06\n\n\nt cement\n\n\n19.2\n\n\n2.4\n\n\n0.07\n\n\nt cement\n\n\n19.8\n\n\n2.4\n\n\n0.07\n\n\nTotal\n\n\n325 cement\n\n\nt cement\n\n\n59\n\n\n99.6\n\n\n2.92\n\n\n425 cement\n\n\nt cement\n\n\n99.8\n\n\n60\n\n\n2.92\n\n\n525 cement\n\n\nt cement\n\n\n100.0\n\n\n2.93\n\n\n62\n\n\n625 cement\n\n\nt cement\n\n\n100.1\n\n\n2.93\n\n\n62\n\n\nNotes: all values in final energy. Assumes ratio of 1.77 t raw materials per t clinker; ratio of coal to clinker is 0.97; ratio of additives to cement is\n0.05 for Portland cement; clinker to cement ratio is 0.095 for Portland cement. Electricity required for grinding and blending additives (in addition\nto the electricity required to blend and grind into final product) varies depending on the material ground. See text for more details.\n\n\n24\n24\n\n\nTable 2.3.2. World Best Practice Final Energy Intensity Values for Fly Ash Cement\n\n\nGJ/t kWh/t\nclinker\n\n\nGJ/t\n\n\nkgce/t\n\n\nGJ/t\nproduct\n\n\nProduct\nunit\n\n\nkWh/t\nproduct\n\n\nkgce/t\nproduct\n\n\nkWh/t kgce/t\nclinker clinker\n\n\ncement\n\n\ncement\n\n\ncement\n\n\nRaw Materials Preparation\n\n\nElectricity\nElectricity\n\n\nt raw meal\n\n\n0.04\n\n\n0.05\n\n\n2.62\n\n\n12.05\n\n\n1.5\n\n\n21.3\n\n\n0.08\n\n\n1.70\n\n\n13.9\n\n\nSolid Fuels Preparation\n\n\nt coal\n\n\n0.12\n\n\n0.08\n\n\n0.63\n\n\n10\n\n\n1.2\n\n\n0.04\n\n\n0.97\n\n\nClinker Making\n\n\nt clinker\n\n\nFuel\n\n\n63\n\n\n97\n\n\n2.85\n\n\n1.9\n\n\nt clinker\n\n\nElectricity\n\n\n22.5\n\n\n2.8\n\n\n0.08\n\n\n14.6\n\n\n1.80\n\n\n0.05\n\n\nAdditives Preparation\n\n\nt additive\n\n\nFuel\n\n\nt additive\n\n\nElectricity\n\n\n0.86\n\n\n7\n\n\n0.03\n\n\nCement grinding\n\n\n325 cement\n\n\nElectricity t cement\nElectricity t cement\nElectricity\nElectricity\n\n\n23\n\n\n2.8\n\n\n0.08\n\n\n425 cement\n525 cement\n625 cement\n\n\n25\n\n\n3.1\n\n\n0.09\n\n\nt cement\n\n\n28\n\n\n3.4\n\n\n0.10\n\n\nt cement\n\n\n3.5\n\n\n28\n\n\n0.10\n\n\nTotal\n\n\n325 cement\n\n\nt cement\n\n\n52\n\n\n69.6\n\n\n2.04\n\n\n425 cement\n\n\nt cement\n\n\n2.05\n\n\n54\n\n\n69.9\n\n\n525 cement\n\n\nt cement\n\n\n70.2\n\n\n57\n\n\n2.06\n\n\n625 cement\n\n\nt cement\n\n\nna\n\n\nna\n\n\nna\n\n\nNotes: all values in final energy. Assumes ratio of 1.77 t raw materials per t clinker; ratio of coal to clinker is 0.097; ratio of additives to cement is\n0.35 for fly ash cement (5% is gypsum and anhydrites; 30% is fly ash); clinker to cement ratio is 0.65 for fly ash cement. Electricity required for\ngrinding and blending additives (in addition to the electricity required to blend and grind into final product) varies depending on the material\nground. See text for more details.\n\n\n25\n\n\nTable 2.3.3. World Best Practice Final Energy Intensity Values for Blast Furnace Slag Cement\n\n\nGJ/t kWh/t\nclinker\n\n\nkgce/t\ncement\n\n\nGJ/t\n\n\nkgce/t GJ/t\nproduct product\n\n\nkWh/t kgce/t\nclinker clinker\n\n\nProduct\nunit\n\n\nkWh/t\nproduct\n\n\ncement\n\n\ncement\n\n\nRaw Materials Preparation\n\n\nt raw meal\nt coal\n\n\nElectricity\nElectricity\n\n\n1.5\n\n\n21.33\n\n\n2.62\n\n\n0.92\n\n\n12.05\n\n\n0.03\n\n\n0.04\n\n\n0.08\n\n\n7.5\n\n\nSolid Fuels Preparation\n\n\n0.12\n\n\n0.04\n\n\n10\n\n\n1.2\n\n\n0.04\n\n\n0.97\n\n\n0.34\n\n\nClinker Making\n\n\nt clinker\n\n\nFuel\n\n\n2.85\n\n\n97.3\n\n\n34.0\n\n\n1.00\n\n\nt clinker\n\n\nElectricity\n\n\n22.5\n\n\n2.8\n\n\n0.08\n\n\n7.9\n\n\n0.03\n\n\n1.0\n\n\nAdditives Preparation\n\n\nt additive\n\n\nFuel\n\n\n15.4\n\n\n25.6\n\n\n0.45\n\n\nt additive\n\n\nElectricity\n\n\n0.09\n\n\n3.07\n\n\n25\n\n\nCement grinding\n\n\n325 cement\n\n\nElectricity t cement\nElectricity t cement\nElectricity\nElectricity\n\n\n5.0\n\n\n0.15\n\n\n41\n\n\n425 cement\n525 cement\n625 cement\n\n\n44\n\n\n5.4\n\n\n0.16\n\n\nt cement\n\n\n49\n\n\n6.0\n\n\n0.18\n\n\nt cement\n\n\n51\n\n\n6.2\n\n\n0.18\n\n\nTotal\n\n\n325 cement\n\n\nt cement\n\n\n1.65\n\n\n57\n\n\n56.4\n\n\n425 cement\n\n\nt cement\n\n\n1.66\n\n\n60\n\n\n56.8\n\n\n525 cement\n\n\nt cement\n\n\n65\n\n\n57.4\n\n\n1.68\n\n\n625 cement\n\n\nt cement\n\n\nna\n\n\nna\n\n\nna\n\n\nNotes: all values in final energy. Assumes ratio of 1.77 t raw materials per t clinker; ratio of coal to clinker is 0.097; ratio of additives to cement is\n0.65 for blast furnace slag cement (5% is gypsum or anhydrites; 60% is slags); clinker to cement ratio is 0.35 for blast furnace slag cement.\nElectricity required for grinding and blending additives (in addition to the electricity required to blend and grind into final product) varies\ndepending on the material ground. See text for more details.\n\n\n26\n\n\nTable 2.3.4. World Best Practice Primary Energy Intensity Values for Portland Cement\n\n\nGJ/t kWh/t\nclinker clinker\ncement\n\n\nGJ/t\nproduct\n\n\nGJ/t\n\n\nProduct\nunit\n\n\nkgce/t\n\n\nkgce/t\nproduct product\n\n\nkWh/t\n\n\nkWh/t kgce/t\nclinker\n\n\ncement\n\n\ncement\n\n\nRaw Materials Preparation\n\n\nElectricity\n\n\nt raw meal\n\n\n64.6\n\n\n0.13\n\n\n7.54\n\n\n37\n\n\n7.94\n\n\n4.5\n\n\n0.23\n\n\n61.4\n\n\n0.22\n\n\nSolid Fuels Preparation\n\n\nElectricity\n\n\nt coal\n\n\n0.34\n\n\n3.7\n\n\n0.01\n\n\n30\n\n\n0.11\n\n\n2.95\n\n\n0.36\n\n\n2.80\n\n\n0.01\n\n\nClinker Making\n\n\nt clinker\n\n\nFuel\n\n\n2.85\n\n\n97\n\n\n92\n\n\n2.71\n\n\nt clinker\n\n\nElectricity\n\n\n68.2\n\n\n0.25\n\n\n8.4\n\n\n64.8\n\n\n7.96\n\n\n0.23\n\n\nAdditives Preparation\n\n\nt additive\n\n\nFuel\n\n\nt additive\n\n\nElectricity\n\n\nCement grinding\n\n\n325 cement\n\n\nElectricity t cement\nElectricity t cement\nElectricity\nElectricity\n\n\n6.0\n\n\n0.17\n\n\n48\n\n\n425 cement\n525 cement\n625 cement\n\n\n52\n\n\n6.4\n\n\n0.19\n\n\nt cement\n\n\n58\n\n\n7.1\n\n\n0.21\n\n\nt cement\n\n\n0.22\n\n\n60\n\n\n7.4\n\n\nTotal\n\n\n325 cement\n\n\nt cement\n\n\n177\n\n\n114.2\n\n\n3.35\n\n\n425 cement\n\n\nt cement\n\n\n3.36\n\n\n181\n\n\n114.7\n\n\n525 cement\n\n\nt cement\nt cement\n\n\n115.4\n\n\n187\n\n\n3.38\n\n\n625 cement\n\n\n189\n\n\n115.6\n\n\n3.39\n\n\nNotes: all values in primary energy. Primary energy includes electricity generation, transmission, and distribution losses of 67%. Assumes ratio of\n1.77 t raw materials per t clinker; ratio of coal to clinker is 0.097 for Portland cement; ratio of additives to cement is 0.05 for Portland cement;\nclinker to cement ratio is 0.95 for Portland cement. Electricity required for grinding and blending additives (in addition to the electricity required\nto blend and grind into final product) varies depending on the material ground. See text for more details.\n\n\n27\n\n\nTable 2.3.5. World Best Practice Primary Energy Intensity Values for Fly Ash Cement\n\n\nGJ/t\n\n\nkgce/t GJ/t kWh/t\nclinker clinker cement\n\n\nkgce/t\ncement\n\n\nProduct\nunit\n\n\nGJ/t\n\n\nkWh/t\nproduct\n\n\nkgce/t\n\n\nkWh/t\nclinker\n\n\nproduct product\n\n\ncement\n\n\nRaw Materials Preparation\n\n\nt raw meal\nt coal\n\n\nElectricity\nElectricity\n\n\n64.6\n\n\n0.15\n\n\n37\n\n\n0.2\n\n\n42.0\n\n\n5.16\n\n\n4.5\n\n\n0.13\n\n\n7.94\n\n\nSolid Fuels Preparation\n\n\n3.7\n\n\n0.01\n\n\n30\n\n\n0.11\n\n\n2.95\n\n\n0.36\n\n\n1.92\n\n\n0.24\n\n\n0.01\n\n\nClinker Making\n\n\nt clinker\n\n\nFuel\n\n\n2.85\n\n\n97\n\n\n63\n\n\n1.9\n\n\nt clinker\n\n\nElectricity\n\n\n68.2\n\n\n8.4\n\n\n0.25\n\n\n5.44\n\n\n0.16\n\n\n44.3\n\n\nAdditives Preparation\n\n\nt additive\n\n\nFuel\n\n\nt additive\n\n\nElectricity\n\n\n21.2\n\n\n2.61\n\n\n0.08\n\n\nCement grinding\n\n\n325 cement\n\n\nElectricity t cement\nElectricity t cement\nElectricity\nElectricity\n\n\n70\n\n\n8.6\n\n\n0.25\n\n\n425 cement\n525 cement\n625 cement\n\n\n75\n\n\n9.3\n\n\n0.27\n\n\nt cement\n\n\n10.3\n\n\n0.30\n\n\n84\n\n\nt cement\n\n\n86\n\n\n10.6\n\n\n0.31\n\n\nTotal\n\n\n325 cement\n\n\nt cement\n\n\n158\n\n\n82.6\n\n\n2.42\n\n\n425 cement\n\n\n164\n\n\nt cement\n\n\n83.3\n\n\n2.44\n\n\n525 cement\n\n\nt cement\nt cement\n\n\n84.3\n\n\n2.47\n\n\n172\n\n\n625 cement\n\n\nna\n\n\nna\n\n\nna\n\n\nNotes: all values in primary energy. Primary energy includes electricity generation, transmission, and distribution losses of 67%. Assumes ratio of\n1.77 t raw materials per t clinker; ratio of coal to clinker is 0.097; ratio of additives to cement is 0.35 for fly ash cement (5% is gypsum and\nanhydrites; 30% is fly ash); clinker to cement ratio is 0.65 for fly ash cement. Electricity required for grinding and blending additives (in addition\nto the electricity required to blend and grind into final product) varies depending on the material ground. See text for more details.\n\n\n28\n\n\nTable 2.3.6. World Best Practice Primary Energy Intensity Values for Blast Furnace Slag Cement\n\n\nGJ/t kWh/t\nclinker\n\n\nGJ/t\n\n\nkgce/t\ncement\n\n\nProduct\nunit\n\n\nGJ/t\nproduct\n\n\nkWh/t\nproduct\n\n\nkgce/t\nproduct\n\n\nkWh/t kgce/t\nclinker clinker\n\n\ncement\n\n\ncement\n\n\nRaw Materials Preparation\n\n\nElectricity\nElectricity\n\n\nt raw meal\n\n\n7.54\n\n\n64.6\n\n\n7.94\n\n\n37\n\n\n0.13\n\n\n4.5\n\n\n0.23\n\n\n0.22\n\n\n61.4\n\n\nSolid Fuels Preparation\n\n\nt coal\n\n\n3.7\n\n\n0.01\n\n\n30\n\n\n0.11\n\n\n2.95\n\n\n0.36\n\n\n1.03\n\n\n0.13\n\n\n0.00\n\n\nClinker Making\n\n\nt clinker\n\n\nFuel\n\n\n2.85\n\n\n97.3\n\n\n34.0\n\n\n1.00\n\n\nt clinker\n\n\nElectricity\n\n\n0.25\n\n\n68.2\n\n\n8.4\n\n\n23.9\n\n\n2.93\n\n\n0.09\n\n\nAdditives Preparation\n\n\nt additive\n\n\nFuel\n\n\n15.36\n\n\n0.45\n\n\nt additive\n\n\nElectricity\n\n\n9.31\n\n\n0.27\n\n\n75.8\n\n\nCement grinding\n\n\n325 cement\n\n\nElectricity t cement\nElectricity t cement\nElectricity\nElectricity\n\n\n91\n\n\n11.2\n\n\n0.33\n\n\n425 cement\n525 cement\n625 cement\n\n\n98\n\n\n12.1\n\n\n0.35\n\n\nt cement\n\n\n109\n\n\n0.39\n\n\n13.4\n\n\nt cement\n\n\n113\n\n\n13.8\n\n\n0.41\n\n\nTotal\n\n\n325 cement\n\n\nt cement\n\n\n177\n\n\n71.2\n\n\n2.09\n\n\n425 cement\n\n\nt cement\n\n\n2.11\n\n\n185\n\n\n72.1\n\n\n525 cement\n\n\n73.4\n\n\nt cement\nt cement\n\n\n195\n\n\n2.15\n\n\n625 cement\n\n\nna\n\n\nna\n\n\nna\n\n\nNotes: all values in primary energy. Primary energy includes electricity generation, transmission, and distribution losses of 67%. Assumes ratio of\n1.77 t raw materials per t clinker; ratio of coal to clinker is 0.097; ratio of additives to cement is 0.65 for blast furnace slag cement (5% is gypsum\nor anhydrites; 60% is slags); clinker to cement ratio is 0.35 for blast furnace slag cement. Electricity required for grinding and blending additives\n(in addition to the electricity required to blend and grind into final product) varies depending on the material ground. See text for more details.\n\n\n29\n29\n\n\nAdditional requirements from use of additives are based on the differences between\nblending and grinding Portland cement (5% additives) and other types of cement (up to\n65% additives). Portland Cement typically requires about 55 kWh/t for clinker grinding,\nwhile fly ash cement (with 25% fly ash) typically requires 60 kWh/t and blast furnace\nslag cement (with 65% slag) 80 kWh/t (these are typical grinding numbers only used to\ndetermine the additional grinding energy required by additives, not best practice; for best\npractice refer to data below in cement grinding section). 38 It is assumed that only fly ash,\nblast furnace and other slags and natural pozzolans need additional energy. Based on the\ndata above, fly ash will require an additional 20 kWh/t of fly ash and slags will require an\nadditional 38 kWh/t of slag. It is assumed that natural pozzolans have requirements\nsimilar to fly ash. These data are used to calculate cement grinding requirements. For\nadditives which are dried, best practice requires 0.75 GJ/t (26 kgce/t) of additive.\"\nGenerally, only blast furnace and other slags are dried.\n\n\n39\n\n\n2.3.4 Cement Grinding\n\n\nBest practice for cement grinding depends on the cement being produced, measured as\nfineness or Blaine (cm\u00b2/g). In 1997, it was reported that the Horomill required 25\nkWh/tonne of cement for 3200 Blaine and 30 kWh/tonne cement for 4000 Blaine. 40 We\nmake the following assumptions re: Chinese cement types - 325 = a Blaine of less than or\nequal to 3200, 425 = a Blaine of approximately 3500, 525 = a Blaine of about 4000, and\n625\n= a Blaine of approximately 4200. More recent estimates of Horomill energy\nconsumption range between 16 and 19 kWh/tonne. 41 We used best practice values for the\nHoromill for 3200 and 4000 Blaine and interpolated and extrapolated values based on an\nassumed linear distribution for 3500 and 4200 Blaine. We estimated lowest quality\ncement requires 16 kWh/tonne and that 3500 Blaine is 8% more than 3200 Blaine (17.3\nkWh/tonne), 4000 Blaine is 20% more than 3200 Blaine (19.2 kWh/tonne), and 4200\nBlaine is 24% more than 3200 Blaine (19.8 kWh/tonne). We then used these values to\nestimate the values of other types of cement, based on more or less grinding that would\nbe needed for any additives.\n\n\n38 Van Heijningen, R.J.J., J.F.M de Castro and E. Worrell (ed.). 1992. Energiekentallen in relatie tot\npreventie en hergebruik van afvalstromen, Rapport in opdracht van Nationaal Onderzoeks Programma\nHergebruik van Afvalstoffen, Utrecht/Bilthoven, February.\nE. Worrell, R.J.J. van Heijningen, J.F.M. de Castro, J.H.O. Hazewinkel, J.G. de Beer, A.P.C. Faaij and\nK. Vringer, 1994. \"New Gross Energy-Requirement Figures for Materials Production\", Energy, the\nInternational Journal 6 19 pp.627-640.\n40 Buzzi, S. 1997. Die Horomill\u00ae - Eine Neue M\u00fchle f\u00fcr die Feinzerkleinerung, ZKG International 3 50:\n\n\n39\n\n\n127-138.\n\n\n41 Hendricks, C.A., Worrell, E., de Jager, D., Blok, K., and Riemer, P., 2004. \"Emission Reduction of\nGreenhouse Gases from the Cement Industry,\" Proceedings of Greenhouse Gas Control Technologies\nConference. http://www.wbcsd.org/web/projects/cement/tf1/prghgt42.pdf\n\n\n30\n30\n\n\n2.3.5 Other Production Energy\n\n\nSome cement enterprises have quarries on-site, and those generally use both trucks and\nconveyors to move raw materials. If applicable to the cement facility, quarrying is\nestimated to use about 1% of the total electricity at the facility.4\n\n\n42\n\n\nOther production energy includes power for auxiliaries, conveyors within the facility, and\npackaging equipment. Total power use for auxiliaries is estimated to require about 10\nkWh/t of clinker at a cement enterprise. Packaging and conveyors together are estimated\nto use about 5% of the total electricity at a cement enterprise; of that 5%, power use for\nconveyors is estimated to require about 1 to 2 kWh/t of cement.4\n\n\n43\n\n\n42 Warshawsky, J. of CMP. 1996. TechCommentary: Electricity in Cement Production. EPRI Center for\nMaterials Production, Carnegie Mellon Research Institute, Pittsburgh, PA.\n43 Worrell, E. and Galitsky, C., 2004. Energy Efficiency Improvement Opportunities for Cement Making:\nAn ENERGY STAR\u00ae Guide for Energy and Plant Managers. Berkeley, CA: Lawrence Berkeley National\nLaboratory (LBNL-54036).\n\n\n31\n\n\n2.4. Pulp and Paper\n\n\nThe pulp and paper industry converts fibrous raw materials into pulp, paper, and\npaperboard. The processes involved in papermaking include raw materials preparation,\npulping (chemical, semi-chemical, mechanical, or waste paper), bleaching, chemical\nrecovery, pulp drying, and papermaking. The most significant energy-consuming\nprocesses are pulping and drying. Globally, wood is the main fiber source in the paper\nindustry and most mills are quite large, producing over 300,000 t/year for typical paper\nmills.\n\n\nInternational best practice energy use in pulp and papermaking technology is based on\nwood-based fibers.\" Hence, the identified best practice technologies may not be\napplicable to non-wood fiber based pulp mills. Most papermaking technology is\ndeveloped and manufactured in Europe (Metso and Voith) and Japan (Mitsubishi), and\nspecialized products from North America (e.g. felts). There is limited experience with\nnon-wood fiber outside of China and India, with the last mills in Europe closing down\n(Dunavarosc in Hungary (1980s), Fredericia in Denmark (1991), and SAICA in Spain\n(1999) due to tightening environmental regulations. Even though there is increased\ninterest in the use of non-wood fibers internationally, only a few best practice\ntechnologies are available. Only one \"non-wood\" best practice pulping technology\noutside China has been identified. Other clean modern non-wood pulping technology has\nnot yet been demonstrated on commercial scales. Although the use of non-wood fibers\nmay affect the characteristics of the pulp (e.g. runnability, water retention), it is hard to\nevaluate ex-ante the impact on the energy use of the paper machine. The energy use of\nthe paper machine is generally dependent on the pulp quality and paper grade produced,\nand hence the best practice values apply to paper machines, indiscriminate of the source\nof the virgin pulp (given a specific quality). Note that the variation of the pulp\ncharacteristics and paper grades is so large, that it will affect the best practice energy\nintensity values.\n\n\n44\n\n\nTables 2.4.1 and 2.4.2 provide best practice final and primary energy intensity values,\nrespectively, for stand-alone pulp mills. Tables 2.4.3 and 2.4.4 provide best practice final\nand primary energy intensity values, respectively, for stand-alone paper mills. The best\npractice energy figures are only indicative, as energy use will depend on the specific\n\n\n44\n\n\nChina's paper industry is unique in that it is one of the largest users of non-wood fibers. In fact, the share\nof wood fiber used in China has declined since the 1990s to about 7% of the input of the paper production,\nwith recovered paper representing 36%, and the remaining 57% covered by imported waste paper and non-\nwood fiber (http://faostat.fao.org/). In the late 1990s there were over 5,000 pulp and paper mills in China,\nof which over 70% used non-wood fibers, mainly straw. Most non-wood fiber mills are small scale. In\n1998 there were only 43 non-wood mills with a capacity exceeding 30,000 t/year, and the vast majority\nproduced less than 10,000, or even 5,000 t/year (Ren, X., 1998. Cleaner Production in China's Pulp and\nPaper Industry. Journal of Cleaner Production 6 pp.349-355). Since 2000, the Chinese government has\nstarted to close down the small polluting and inefficient mills. In recent years modern large-scale paper\nmachines have been installed in China (e.g. Hebei Norske Skog Long 300,000 t/year plant, Dagang's paper\nmachine 3 with a capacity of 1.1 million t/year). Non-wood fibers are expected to continue to play an\nimportant role in China's future paper industry.\n\n\n32\n\n\nproperties of the raw materials and products. While the main factors affecting the best\npractice energy use are discussed, the figures should be interpreted with care.\nFurthermore, while the pulping and papermaking processes are discussed separately\nbelow, integration of the pulp and paper mill will result in energy savings due to the\nreduced need to dry pulp and opportunities to provide a better heat integration. The best\npractice energy intensities for the main processes and the factors affecting energy use and\nintensity are discussed below. Only the lime kiln in the Kraft recovery processes uses\nfuel. All other processes only use steam and electricity. Below energy use data is\nexpressed as steam (GJ/t) and electricity (kWh/t).45 Best practice assumes that the steam\nand electricity are generated in a cogeneration (combined heat and power) installation.\n\n\n2.4.1 Non-Wood Pulping\n\n\nThe current international best practice is based on the Chempolis process developed in\nFinland. It provides a clean process that recovers the chemicals. 40 A first demonstration\nplant with a capacity of 65,000 air dry t (ADt)/year has been designed for construction in\nChina, but construction has been delayed. The pulp has similar characteristics as hard-\nwood pulp, resulting in similar behavior (e.g. runnability, water retention) in the paper\nmachine (see below).\n\n\n46\n\n\nThe design assumes a steam consumption of 5 to 6 t/ADt pulp, or equivalent to\napproximately 10.5 to 12.6 GJ/ADt (358 to 430 kgce/ADt) pulp. These values vary with\nthe process lay-out. The above values assume conventional water treatment, and exclude\npulp drying (for preparation of market pulp). Electricity consumption is estimated to be\n400 kWh/ADt. Note that the process uses no fuel directly, as there is no need for a\ncalcination kiln (as with kraft pulping).\n\n\n47\n\n\nThe lignin generates steam of about 7 to 9t/ADt pulp, depending on the lignin yield,\nrequired steam pressure and feed water temperature. Hence, the plant can have an excess\nsteam production of 2 to 3 t/ADt, equivalent to approximately 4.2 to 6.3 GJ/ADt (143 to\n215 kgce/ADt) that can be exported for use in the paper machine.4\n\n\n48\n\n\n45\n\n\nEnergy use in the paper industry is typically expressed per tonne of air dried material (ADt).\n\n\n46\n\n\nAnttila, J.R., P.P. Rousu, P. Rousu, K.J.E. Hytonen and J.P. Tanskanen. 2006. Design of an\nEnvironmentally Benign Non-wood Pulp Plant. Chempolis.\n\n\n47 Rousu, P. 2006. Personal communication from Pasi Rousu, Chempolis, Finland. August 16th, 2006.\n\n\n48 Rousu, P., P. Rousu and J. Antila. 2002. Sustainable Pulp Production form Agricultural Waste.\nResources, Conservation & Recycling 35 pp.85-103.\n\n\n33\n\n\nTable 2.4.1. World Best Practice Final Energy Intensity Values for Stand-Alone Pulp Mills (values are per air dried metric tons). 49, 50\nSteam Exported Electricity Electricity\n\n\nTotal\n\n\nProcess\n\n\nFuel Use for Steam\n\n\nProduct\n\n\nRaw\nMaterial\n\n\nProduced\n\n\nUse\nkWh/Adt\n400\n\n\nGJ/Adt kgce/Adt\n-4.2\n\n\nkWh/Adt GJ/Adt kgce/Adt\n\n\nGJ/ADt kgce/Adt\n\n\nMarket Pulp\nMarket Pulp\n\n\nNon-wood\nWood\n\n\nPulping\nKraft\nSulfite\n\n\n358\n\n\n-143\n\n\n10.5\n\n\n7.7\n\n\n264\n\n\n382\n\n\n640\n\n\n-655\n\n\n11.1\n\n\n380\n\n\n11.2\n\n\n700\n\n\n632\n\n\n16\n\n\n546\n\n\n18.5\n\n\nThermo-mechanical\n\n\n-1.3\n\n\n-45\n\n\n6.6\n\n\n224\n\n\n2190\n\n\nRecovered Pulp\n\n\nPaper\n\n\n0.3\n\n\n330\n\n\n10\n\n\n1.5\n\n\n51\n\n\n51,52\n\n\nTable 2.4.2. World Best Practice Primary Energy Intensity Values for Stand-Alone Pulp Mills (values are per air dried metric tons).\n\n\nSteam Exported Electricity Electricity\n\n\nFuel Use for Steam\n\n\nProduct\n\n\nTotal\n\n\nRaw\nMaterial\n\n\nProcess\n\n\nUse\nkWh/Adt\n1212\n\n\nProduced\n\n\nkgce/Adt\n-143\n\n\nkWh/Adt GJ/Adt kgce/Adt\n\n\nGJ/ADt kgce/Adt\n\n\nGJ/Adt\n-4.2\n\n\nMarket Pulp\nMarket Pulp\n\n\n10.5\n\n\nNon-wood\nWood\n\n\nPulping\n\n\n358\n\n\n364\n\n\n10.7\n\n\nKraft\n\n\n382\n\n\n-1985\n\n\n377\n\n\n11.2\n\n\n1939\n\n\n11.0\n\n\nSulfite\n\n\n16\n\n\n546\n\n\n807\n\n\n2121\n\n\n23.6\n\n\nThermo-mechanical\n\n\n-1.3\n\n\n-45\n\n\n6636\n1000\n\n\n22.6\n\n\n770\n\n\nRecovered Pulp\n\n\n0.3\n\n\n133\n\n\nPaper\n\n\n10\n\n\n3.9\n\n\nNote: Primary energy includes electricity generation, transmission, and distribution losses of 67%.\n\n\n49 IPPC, 2001. Reference Document on Best Available Techniques in the Pulp and Paper Industries. Integrated Pollution Prevention & Control. European\nCommission, Brussels/Sevilla, 2001.\n\n\n50 Francis, D.W., M.T. Towers, T.C. Browne. 2002. Energy Cost Reduction in the Pulp and Paper Industry: An Energy Benchmarking Perspective. Ottawa: NRCan.\nIPPC, 2001. Reference Document on Best Available Techniques in the Pulp and Paper Industries. Integrated Pollution Prevention & Control. European\nCommission, Brussels/Sevilla, 2001.\n52 Francis, D.W., M.T. Towers, T.C. Browne. 2002. Energy Cost Reduction in the Pulp and Paper Industry: An Energy Benchmarking Perspective. Ottawa: NRCan.\n\n\n51\n\n\n34\n4\n\n\n53, 54, 55\n\n\nTable 2.4.3. World Best Practice Final Energy Intensity Values for Stand-Alone Paper Mills (values are per air dried metric tons).\nRaw Material\n\n\nElectricity Use\nkWh/ADt\n\n\nProduct\n\n\nProcess\n\n\nFuel Use for Steam\n\n\nTotal\n\n\nGJ/ADt\n\n\nkgce/Adt\n\n\nGJ/Adt\n\n\nkgce/ADt\n\n\nPaper Machine\n\n\nPulp\n\n\nUncoated Fine (wood free)\n\n\n6.7\n\n\n229\n\n\n640\n\n\n307\n\n\n9.0\n\n\nPaper Machine\n\n\nCoated Fine (wood free)\n\n\n10.4\n\n\n256\n\n\n810\n\n\n355\n\n\n7.5\n\n\nPaper Machine\n\n\nNewsprint\n\n\n174\n\n\n5.1\n\n\n570\n\n\n7.2\n\n\n244\n\n\nPaper Machine\n\n\nBoard\n\n\n6.7\n\n\n229\n\n\n800\n\n\n9.6\n\n\n327\n\n\nPaper Machine\n\n\n5.9\n\n\n201\n\n\n535\n\n\nKraftliner\n\n\n7.8\n\n\n267\n\n\nPaper Machine\n\n\n1000\n\n\nTissue\n\n\n6.9\n\n\n235\n\n\n10.5\n\n\n358\n\n\nTable 2.4.4. World Best Practice Primary Energy Intensity Values for Stand-Alone Paper Mills (values are per air dried metric tons).\nRaw Material\n\n\n56, 57, 58\n\n\nFuel Use for Steam\n\n\nElectricity Use\nkWh/ADt\n\n\nTotal\n\n\nProcess\n\n\nProduct\n\n\nGJ/ADt\n\n\nkgce/Adt\n\n\nGJ/Adt kgce/ADt\n\n\nPaper Machine\n\n\nPulp\n\n\nUncoated Fine (wood free)\n\n\n6.7\n\n\n229\n\n\n1939\n\n\n13.7\n\n\n467\n\n\nPaper Machine\n\n\nCoated Fine (wood free)\n\n\n7.5\n\n\n256\n\n\n2455\n\n\n16.3\n\n\n558\n\n\nPaper Machine\n\n\nNewsprint\n\n\n1727\n\n\n386\n\n\n5.1\n\n\n174\n\n\n11.3\n\n\nPaper Machine\n\n\nBoard\nKraftliner\nTissue\n\n\n6.7\n\n\n229\n\n\n527\n\n\n2424\n\n\n15.4\n\n\nPaper Machine\n\n\n5.9\n\n\n201\n\n\n1621\n\n\n11.7\n\n\n401\n\n\nPaper Machine\n\n\n17.8\n\n\n6.9\n\n\n235\n\n\n3030\n\n\n608\n\n\nNote: Primary energy includes electricity generation, transmission, and distribution losses of 67%.\n\n\n53 IPPC, 2001. Reference Document on Best Available Techniques in the Pulp and Paper Industries. Integrated Pollution Prevention & Control. European\nCommission, Brussels/Sevilla.\n\n\n54 Karlsson, M., 2005. The Dutch Innovation Transition, Small/Large Paper/Board Machine Concepts, Automation. Presentation at Meeting of the Royal\nNetherlands Paper and Board Industry Association (VNP), Beekbergen, The Netherlands, February 23rd, 2005.\n\n\n55 Francis, D.W., M.T. Towers, T.C. Browne. 2002. Energy Cost Reduction in the Pulp and Paper Industry: An Energy Benchmarking Perspective. Ottawa: NRCan.\nIPPC, 2001. Reference Document on Best Available Techniques in the Pulp and Paper Industries. Integrated Pollution Prevention & Control. European\nCommission, Brussels/Sevilla.\n\n\n56\n\n\n57 Karlsson, M., 2005. The Dutch Innovation Transition, Small/Large Paper/Board Machine Concepts, Automation. Presentation at Meeting of the Royal\nNetherlands Paper and Board Industry Association (VNP), Beekbergen, The Netherlands, February 23rd, 2005.\n\n\n58 Francis, D.W., M.T. Towers, T.C. Browne. 2002. Energy Cost Reduction in the Pulp and Paper Industry: An Energy Benchmarking Perspective. Ottawa: NRCan.\n\n\n35\n\n\n2.4.2. Kraft Pulping\n\n\nA best practice Kraft mill produces excess electricity that can be exported. The export is\nthe result of balancing the energy used in the pulping process and the energy recovered\nfrom the black liquor recovery process (combusting the lignin). The energy consumption\nof the process itself varies between 10 and 12.2 GJ/ADt (341-416 kgce/ADt) pulp, while\nelectricity use is around 610 kWh/ADt (75 kgce/ADt). The lime kiln uses 1.2 GJ/ADt in\nfuels and 30 kWh/ADt for total energy consumption of 11.2 GJ/ADt (382 kgce/ADt) in\nfuels and 640 kWh/ADt in electricity.\n\n\n59\n\n\nHowever, the recovery process is a net producer of 15.8 GJ/ADt of steam. It is assumed\nthat the steam is used in a back-pressure steam turbine to generate electricity (around 655\nkWh/ADt), resulting in a net export of power of 15 to 20 kWh/ADt. This leads to a total\noverall energy consumption value of 11.1 GJ/ADt (380 kgce).\n\n\nResearch and development in black liquor gasification has not yet resulted in a\ncommercially operating process, and is hence not included in the best practice energy\nconsumption figures. However, when this technology is available it could result in\nsignificant energy savings, due to large amounts of excess power production.\n\n\n2.4.3 Sulfite Pulping\n\n\nSulfite pulping is used much less than Kraft pulping, and mainly used for specialty\npapers. Also, most of the sulfite pulp is bleached. The wood is cooked using a solution of\nsulfur dioxide with an alkaline solution. The process can be operated to produce a wide\nrange of specialty products, which will also result in a wide range of energy use. Energy\ncan be recovered from the \"green liquor\u201d, similar to the black liquor recovery process,\nproducing about 15 GJ/ADt (512 kgce/ADt) pulp. The best practice assumes optimization\nof power use, state-of-the-art controls, efficient evaporation and concentration of the\ngreen liquor. Due to the variety of pulps to be produced, steam use is estimated to be 16\nto 18 GJ/ADt (546 to 614 kgce/ADt) and electricity use to be 700 kWh/ADt.\"\n\n\n60\n\n\n2.4.4 Mechanical Pulping\n\n\nEnergy use in mechanical pulping is determined by the wood type used and the\n\u201cfreeness\u201d of the pulp. The \u201cfreeness\u201d is an expression for the fiber quality and water\nretention. Hence, energy use may vary widely on the basis of the desired pulp quality\ngiven a specific wood type used.\n\n\nThere are several types of processes that can be used for mechanical pulping, i.e.\ngroundwood (GW), thermo-mechanical pulping (TMP) and chemo-thermo-mechanical\npulping (CTMP). The best practice assumes TMP. TMP allows the recovery of heat from\n\n\n59 Francis, D.W., M.T. Towers and T.C. Browne. 2002. Energy Cost Reduction in the Pulp and Paper\nIndustry - An Energy Benchmarking Perspective. NRCan, Ottawa, ON, Canada.\n60 IPPC, 2001. Reference Document on Best Available Techniques in the Pulp and Paper Industries.\nIntegrated Pollution Prevention & Control. European Commission, Brussels/Sevilla, 2001.\n\n\n36\n\n\nthe process in the form of hot water and steam, as only a fraction of the energy is actually\nused to separate the fibers. TMP allows the recovery of 60-65% of the heat generated in\nthe process (45% as steam, 20% as hot water). However, a TMP mill consumes more\npower than a groundwood mill. The best practice integrated TMP newsprint mill consists\nof a pressurized TMP mill consuming about 2190 kWh/ADt and generating 1.33 GJ/ADt\n(45 kgce/ADt) of heat. For an non-integrated pulp mill electricity use is estimated at 2420\nkWh/ADt with generation of 5.5 GJ/ADt of steam.'\n\n\n61, 62\n\n\n2.4.5 Fiber Recovery\n\n\nFiber recycling is an important option for reducing pulping energy use. China also\nimports paper from other countries (notably the US and Europe) to provide its fiber\nneeds. Recycled fiber has become a global market in which China is an important\nconsumer. The used fibers are pulped and (optionally) de-inked before being fed to stock\npreparation for the paper machine. Based on the performance of Swedish mills, the best\npractice is estimated to be 0.3 GJ/ADt (10 kgce/ADt) use of steam and electricity use of\n330 kWh/ADt.\n\n\n63\n\n\n2.4.6. Papermaking\n\n\nEnergy use in the paper machine is determined by the specific grade of paper to be\nproduced and the fiber quality (e.g. water retention) in the pulp. Moreover, not all energy-\nefficient technologies are suitable for all paper grades. The best practice values assume\nthat an effective control system is in place, long nip (or shoe) press is being used (not\nsuitable for tissue mills), use of efficient motors, condensate recovery, a closed hood for\nheat recovery, as well as integration of the various steam and hot water flows in the mill.\nNote that small scale mills may have a steam consumption that is 10-25% higher and an\nelectricity consumption that is 5-20% higher than the figures presented in Table 2.4.1.64\n\n\n2.4.7 Integrated Pulp and Paper Mills\n\n\nIntegrated mills can be more energy efficient than stand-alone mills, as no drying energy\nis needed for the intermediate drying of the pulp. This will result in energy savings at the\npulp mill. Furthermore, process integration of the different processes may result in a\nfurther optimization of the steam use on site. Finally, while stand-alone pulp mills may\nhave excess steam that cannot be used (due to black/green liquor recovery or from heat\nrecovery of the TMP), an integrated mill can use this excess heat to serve the additional\nheat use of the paper machine. Tables 2.4.5 and 2.4.6 summarize the best practice final\nand primary energy intensity, respectively, of various integrated mill types.\n\n\n61 Francis, D.W., M.T. Towers and T.C. Browne. 2002. Energy Cost Reduction in the Pulp and Paper\nIndustry - An Energy Benchmarking Perspective. NRCan, Ottawa, ON, Canada.\nIPPC, 2001. Reference Document on Best Available Techniques in the Pulp and Paper Industries.\nIntegrated Pollution Prevention & Control. European Commission, Brussels/Sevilla, 2001.\n63 Francis, D.W., M.T. Towers and T.C. Browne. 2002. Energy Cost Reduction in the Pulp and Paper\n\n\n62\n\n\nIndustry - An Energy Benchmarking Perspective. NRCan, Ottawa, ON, Canada.\n\n\n64\n\n\nIPPC, 2001. Reference Document on Best Available Techniques in the Pulp and Paper Industries.\n\n\nIntegrated Pollution Prevention & Control. European Commission, Brussels/Sevilla, 2001.\n\n\n37\n\n\nTable 2.4.5. World Best Practice Final Energy Intensity Values for Integrated Pulp and\nPaper Mills (values are per air dried metric tons). 65, 66\n\n\nTotal\n\n\nProduct\n\n\nFuel Use for Steam\n\n\nRaw\n\n\nProcess\n\n\nElectricity\n\n\nMaterial\n\n\nkgce/ADt\n\n\nkWh/ADt GJ/ADt\n\n\nGJ/ADt\n\n\nkgce/ADt\n\n\nKraft\n\n\nWood\n\n\n625\n\n\nBleached Uncoated Fine\n\n\n14\n\n\n478\n\n\n1200\n\n\n18.3\n\n\nKraftliner (unbleached)\n\n\n1000\n\n\nKraft\n\n\n14\n\n\n478\n\n\n17.6\n\n\n601\n\n\nand Bag Paper\n\n\nBleached Coated Fine\n\n\nSulfite\n\n\n17\n\n\n580\n\n\n1500\n\n\n22.4\n\n\n765\n\n\nBleached Uncoated Fine\n\n\nSulfite\n\n\n18\n\n\n614\n\n\n1200\n\n\n22.3\n\n\n762\n\n\nNewsprint\n\n\n2200\n\n\nTMP\n\n\n226\n\n\n-1.3\n\n\n-44\n\n\n6.6\n\n\nMagazine Paper\n\n\n-0.3\n\n\n-10\n\n\n2100\n\n\n7.3\n\n\n248\n\n\nTMP\n\n\n3.5\n\n\nBoard\n\n\n50% TMP\n\n\n119\n\n\n2300\n\n\n11.8\n\n\n402\n\n\nBoard (no de-inking)\n\n\nRecovered\nPaper\n\n\n273\n\n\n900\n\n\n384\n\n\n8\n\n\n11.2\n\n\n1000\n\n\nNewsprint (de-inked)\n\n\n137\n\n\n7.6\n\n\n259\n\n\n4\n\n\nTissue (de-inked)\n\n\n7\n\n\n239\n\n\n1200\n\n\n11.3\n\n\n386\n\n\nTable 2.4.6. World Best Practice Primary Energy Intensity Values for Integrated Pulp and\nPaper Mills (values are per air dried metric tons).\u201c\n\n\n67, 68\n\n\nFuel Use for Steam Electricity\n\n\nProduct\n\n\nRaw\nMaterial\n\n\nProcess\n\n\nTotal\n\n\nGJ/ADt\n\n\nGJ/ADt\n\n\nkgce/ADt\n\n\nkWh/ADt\n\n\nkgce/ADt\n\n\nWood\n\n\nKraft\n\n\nBleached Uncoated Fine\n\n\n3636\n\n\n27.1\n\n\n925\n\n\n14\n\n\n478\n\n\nKraftliner (unbleached)\n\n\nKraft\n\n\n478\n\n\n3030\n\n\n24.9\n\n\n850\n\n\n14\n\n\nand Bag Paper\n\n\nSulfite\n\n\nBleached Coated Fine\n\n\n24.9\n\n\n850\n\n\n14\n\n\n478\n\n\n3030\n\n\nSulfite\n\n\n1139\n\n\nBleached Uncoated Fine\n\n\n4545\n\n\n580\n\n\n33.4\n\n\n17\n\n\nNewsprint\n\n\nTMP\n\n\n3636\n\n\n31.1\n\n\n1061\n\n\n614\n\n\n18\n\n\n6667\n\n\nMagazine Paper\n\n\nTMP\n\n\n-1.3\n\n\n775\n\n\n-44\n\n\n22.7\n\n\n50% TMP\n\n\n772\n\n\n-0.3\n\n\n22.6\n\n\n-10\n\n\n6364\n\n\nBoard\n\n\nBoard (no de-inking)\n\n\n3.5\n\n\n28.6\n\n\nRecovered\nPaper\n\n\n6970\n\n\n976\n\n\n119\n\n\nNewsprint (de-inked)\n\n\n273\n\n\n2727\n\n\n8\n\n\n17.8\n\n\n608\n\n\nTissue (de-inked)\n\n\n3030\n\n\n137\n\n\n509\n\n\n4\n\n\n14.9\n\n\nNote: Primary energy includes electricity generation, transmission, and distribution losses of 67%.\n\n\n65 IPPC, 2001. Reference Document on Best Available Techniques in the Pulp and Paper Industries.\nIntegrated Pollution Prevention & Control. European Commission, Brussels/Sevilla.\n66 Francis, D.W., M.T. Towers and T.C. Browne. 2002. Energy Cost Reduction in the Pulp and Paper\nIndustry - An Energy Benchmarking Perspective. NRCan, Ottawa, ON, Canada.\n67 IPPC, 2001. Reference Document on Best Available Techniques in the Pulp and Paper Industries.\nIntegrated Pollution Prevention & Control. European Commission, Brussels/Sevilla.\n68 Francis, D.W., M.T. Towers and T.C. Browne. 2002. Energy Cost Reduction in the Pulp and Paper\nIndustry - An Energy Benchmarking Perspective. NRCan, Ottawa, ON, Canada.\n\n\n38\n\n\n2.5 Ammonia\n\n\nAmmonia (NH3) manufacture is the most energy-intensive production step in the\nproduction of nitrogenous fertilizers. Ammonia is made by the Haber-Bosch process,\ncombining nitrogen and hydrogen. The hydrogen is most often produced from synthesis\ngas. Synthesis gas can be produced from natural gas, oil (residues), coal or any\nhydrocarbon feedstock. Natural gas is the preferred feedstock due to the high hydrogen\ncontent. Today, over 80% of the world ammonia capacity is produced from natural gas.\nHowever, China is one of the largest ammonia producers in the world and in 2004, 70.3%\nof the feedstock for ammonia production in China was coal, 22.7% was natural gas, and\n7% was oil.\u00b0\n\n\n69\n\n\nThe energy intensity of anhydrous ammonia production is dependent on feedstock,\nprocesses, and technology. Natural gas is the most energy-efficient feedstock, followed\nby heavy oil, which requires an average 30% more input energy per t of output, and coal,\nwhich requires an average 70% more input energy per ton output.' Within each of these\nfeedstocks, the most common production processes are steam reforming of natural gas,\npartial oxidization of heavy fuel oil, gasification of coal, and electrolysis.\n\n\n70\n\n\nThe minimum theoretical energy required for ammonia production depends on the\ncomposition of the natural gas feedstock, but it can be as low as 19.2 GJ/t ammonia or\n23.3 GJ/t nitrogen (both in lower heating value, LHV).' The current best practice will\ndepend strongly on the feedstock. The best practice for natural gas and coal feedstocks\nare provided below. Table 2.5.1 summarizes best practice final energy intensity values\nfor specific ammonia production processes. Since electricity use in ammonia production\nis negligible, primary energy intensity values are assumed to be the same as final energy\nintensity values.\n\n\n71\n\n\nTable 2.5.1. World Best Practice Final Energy Intensity Values for Ammonia Production\n(values are per t ammonia and t nitrogen)\n\n\nEnergy Intensity\nGJ/t N kgce/t NH3\n34\n956\n\n\nFeedstock\n\n\nkgce/t N\n\n\nGJ/t NH3\n28\n\n\nNatural gas steam reforming\n\n\n1160\n\n\nCoal72\n\n\n42.3\n\n\n1188\n\n\n34.8\n\n\n1444\n\n\n2.5.1 Natural Gas Steam Reforming\n\n\nIn 1998 the most energy-efficient recorded ammonia production from natural gas\nrequired 28 GJ/t (1160 kgce/t) NH3 or 34 GJ/t (956 kgce/t) N. Limited power imports are\n\n\n69 China Chemical Technology Industry Association, 2006.\n\n\n70 European Fertilizer Manufacturers Association, 1997. Production of Ammonia: Description of\nProduction Processes, Brussels, Belgium.\n71 Ramirez, C.A., Worrell, E., 2006. \"Feeding fossil fuels to the soil; an analysis of energy embedded and\ntechnological learning in the fertilizer industry,\u201d Resource Conservation & Recycling 46 (2006): 75-93.\n72 Sinopec, 2004. \u201c2004 Parameter and Data,\u201d Beijing: Sinopec, p.5-51; most Chinese ammonia is made\nfrom heavy oil and coal, which is much less energy-efficient than natural gas.\n\n\n39\n39\n\n\nnecessary, assuming the ammonia loop compressor uses a steam turbine, using internally\ngenerated steam.\n\n\nVarious suppliers offer process designs that can attain such an efficiency level. These\nprocesses are characterized by a highly integrated primary and secondary reformer (e.g.\nthe KRES system offered by KBR), CO\u2082 removal using a physical absorption process\n(e.g. selexol), low-pressure ammonia synthesis loop, high-efficiency catalysts, as well as\n(membrane) methane (from the methanator) and hydrogen recovery.\n\n\n2.5.2 Coal\n\n\nThe best practice ammonia plant using coal as feedstock would use a coal gasifier to\nconvert the coal to synthesis gas. Most coal gasifier-based ammonia plants are\nconstructed in China.\n\n\nThe most recent plants use Shell or ChevronTexaco gasification technology, e.g. CNTIC\nNanjing Chemical's ammonia plant (start-up in 2003), Jilin and Haolianghe (2004) and\nSinopec's plants in Hubei, Anqing, and Dongtinq (2006). The processes consist of a\nmodern entrained bed gasifier, selexol gas cleanup and a low-pressure ammonia synthesis\nloop. Part of the CO2 is used for the production of urea in sites that produce ammonia and\n\n\nurea.\n\n\nBased on the performance of the ammonia plant in Coffeyville (Kansas, US) that uses\npetroleum coke, the best practice specific energy consumption is estimated to be 34.8\nGJ/t (1188 kgce/t) ammonia.\n\n\n40\n\n\n2.6 Ethylene\n\n\nEthylene is produced from various hydrocarbon feedstocks with the steam cracking\nprocess. Along with ethylene, other high value products such as propylene, butadiene and\naromatics are co-produced in the process. In an absolute sense, steam cracking is the most\nenergy-intensive process in the petrochemical industry with an estimated worldwide\nenergy use (excluding feedstock use) of approximately 2.8 EJ.73 The dominant feedstock\nfor worldwide ethylene production is naphtha (55%), followed by ethane (30%), liquefied\npetroleum gas (10%) and gas oil (5%). 14 Regional differences are substantial with ethane\ncracking being the dominant technology in the U.S. and naphtha cracking the dominant\ntechnology in most other world regions, including China.75\n\n\n74\n\n\n2.6.1 Naphtha and Ethane\n\n\nTables 2.6.1 and 2.6.2 provide best practice final and primary energy intensity values,\nrespectively, for naphtha and ethane cracking, the two most widely used feedstocks in\nconventional ethylene production. Energy use of the processes is allocated to all High\nValue Chemicals (HVC) to allow a fair comparison between various technologies. This is\na method also followed by Solomon Associates Ltd., a company performing international\nbenchmarks for the petrochemical industry. Allocating all energy use to ethylene alone\nwould yield confusing results, because the ethylene yield differs widely per process.\nBased on a survey of European steam crackers, an actual energy use of 14 to 22 GJ/t (478\nto 751 kgce/t) HVCs is found for naphtha crackers and 12.5 to 21.0 GJ/t (427 to 717\nkgce/t) HVCs for ethane crackers.'\n76\n\n\n77\n\n\n78\n\n\n73 Neelis, M.L., Patel, M.K., Bach, P.W. and Haije, W.G., 2005. Analysis of energy use and carbon losses\nin the chemical and refinery industries. Report ECN-I-05-008, Energy Research Centre of the Netherlands,\nPetten, the Netherlands.\nRen, T., Patel, M., and Blok, K., 2006. \u201cOlefins from conventional and heavy feedstocks: Energy use in\nsteam cracking and alternative processes,\u201d Energy 31 (2006), pp. 425-451.\n75 From international energy statistics (International Energy Agency, 2005. Extended energy balances of\nNon-OECD countries., Paris, France: IEA), we can conclude that in China, mainly naphtha cracking is\napplied, because approximately 90% of the oil feedstock delivery to the chemical industry was naphtha in\n2003.\n76 'Worrell, E., Phylipsen, D., Einstein, D. and Martin, N., 2000. Energy use and energy intensity in the US\nchemical industry. Berkeley, California: Lawrence Berkeley National Laboratory, LBNL-44314.\n77 Phylipsen, G.J.M., 2000. \u201cA methodology for international comparisons of the energy efficiency in the\npetrochemical industry,\u201d Chapter 3 in: International Comparisons and National Commitments, PhD thesis,\nUtrecht University, Utrecht, the Netherlands.\n78 Institute for Prospective Technological Studies, 2003. Reference document on best available technologies\nin the large volume organic chemical industry, Seville, Spain: European Commission, Joint Research\nCentre, IPTS.\n\n\n74\n\n\n41\n\n\nTable 2.6.1. World Best Practice Final Energy Intensity Values for Ethane and Naphtha Cracking (values are per t high value chemicals).\n\n\nEthane\n\n\nNaphtha\nTotal kgce/t kWh/t\nGJ/t HVC HVC\n\n\nkgce/t kWh/t\nHVC\n\n\nTotal kgce/t\nHVC\n\n\nGJ/t kWh/t\n\n\nTotal\nkgce/t HVC\n\n\nTotal\nGJ/t HVC\n\n\nkWh/t\n\n\nGJ/t\nHVC\n\n\nUnit process\nCracker\n\n\nHVC\n\n\nHVC HVC\n\n\nHVC\n\n\nHVC\n\n\n278\n\n\n5.9\n\n\n278\n\n\n219\n\n\n6.5\n\n\n6.5\n\n\n4.9\n\n\n184\n\n\n244\n\n\n244\n\n\nHeat of reaction\n\n\n75\n\n\n75\n\n\n2.6\n\n\n90\n\n\n2.0\n\n\n2.0\n\n\nSteam, heating and\n\n\nlosses\n\n\n94\n\n\n4.5\n\n\n4.5\n\n\n169\n\n\n169\n\n\n2.8\n\n\nFractionation and\n\n\ncompression\n\n\n2.8\n\n\n2.8\n\n\n86\n\n\n86\n\n\n1.5\n\n\n56\n\n\n56\n\n\n1.5\n\n\nSeparation\n\n\n3.9\n\n\n3.9\n\n\n122\n\n\n122\n\n\n75\n\n\n75\n\n\n2.0\n\n\n2.0\n\n\nTotal\n\n\n278\n\n\n12.5\n\n\n392\n\n\n427\n\n\n278\n\n\n11.5\n\n\n375\n\n\n10.0\n\n\n11.0\n\n\n409\n\n\nTable 2.6.2. World Best Practice Primary Energy Intensity Values for Ethane and Naphtha Cracking (values are per t high value\nchemicals).\n\n\nEthane\nkgce/t kWh/t\n\n\nNaphtha\n\n\nTotal kgce/t\nHVC\n\n\nTotal\n\n\nGJ/t kWh/t\nHVC HVC\n\n\nTotal kgce/t kWh/t\nGJ/t HVC HVC\n\n\nTotal\nkgce/t HVC\n\n\nGJ/t\n\n\nkWh/t\n\n\nUnit process\n\n\nGJ/t HVC HVC HVC\n\n\nHVC\n\n\nHVC\n\n\nHVC\n\n\nCracker\n\n\n288\n\n\n6.5\n\n\n244\n\n\n4.9\n\n\n842\n\n\n7.9\n\n\n6.5\n\n\n184\n\n\n842\n\n\n244\n\n\nHeat of reaction\n\n\n90\n\n\n75\n\n\n75\n\n\n2.6\n\n\n2.0\n\n\n2.0\n\n\nSteam, heating and\n\n\nlosses\n\n\n2.8\n\n\n94\n\n\n4.5\n\n\n4.5\n\n\n169\n\n\n169\n\n\nFractionation and\n\n\ncompression\n\n\n2.8\n\n\n86\n\n\n86\n\n\n1.5\n\n\n56\n\n\n56\n\n\n2.8\n\n\n1.5\n\n\nSeparation\n\n\n3.9\n\n\n3.9\n\n\n122\n\n\n122\n\n\n2.0\n\n\n75\n\n\n75\n\n\n2.0\n\n\nTotal\n\n\n392\n\n\n842\n\n\n11.5\n\n\n14.5\n\n\n496\n\n\n13.0\n\n\n375\n\n\n842\n\n\n478\n\n\n10.0\n\n\nNote: Primary energy includes electricity generation, transmission, and distribution losses of 67%.\n\n\n42\n42\n\n\nAlthough difficult, because of the intensive energy integration applied in a steam cracker,\nit is possible to roughly divide the total energy use into the various sections of the\ncracker.79\nAs most processes in the chemical industry, steam cracking can be regarded as\na combination of a reaction section, where the feedstock is converted to the desired\nproducts and a separation section, where the various products are separated into chemical\ngrade sellable commodities. In the cracker, naphtha or ethane are cracked at high\ntemperatures (750-900 \u00b0C) and quenched to lower temperatures to stop the reaction. In\nthe quench, high-pressure steam is generated that is used for driving compressors etc. In\nthe separation section, first the heavy fraction is condensed (the hot separation train, only\nfor naphtha and gas oil cracking) and the gaseous fraction is compressed. In the cold\nseparation, the various lighter products (ethylene, propylene, and butadiene) are separated\nusing cryogenic distillation. The exact process layout depends heavily on the feedstock\nprocesses. Typical yields of the various products are given in Table 2.6.3.\n\n\nThe conversions in the cracker are endothermic and the heat of reaction is the minimal\namount of energy required to convert the feedstock to the products (both at standard\nconditions of 1 bar and 25 \u00b0C) and is equivalent to approximately 20-25% of the process\nenergy. The separation is responsible for 20-30% of the energy consumption with the\nremaining energy being consumed in the cracker (e.g. the heat embodied in the flue gases\nof the cracker furnace) and in the compression section.\n\n\n80\n\n\nTable 2.6.3. Yields (%) for Ethane and Naphtha Cracking\nProduct\n\n\nNaphtha\n29-34 (30% typical)\n13-16\n\n\nEthane\n\n\nEthylene - HVC\n\n\n80-84\n\n\nPropylene - HVC\nButadiene-HVC\n\n\n1-1.6\n\n\n4-5\n10-16\n\n\n1-1.4\n\n\n2-3\n\n\nAromatics and C4+ -HVC\n\n\n55 (typical)\n13-14\n\n\n82 (typical)\n4.2\n\n\nTotal yield of HVC's\n\n\nMethane yield\nHydrogen yield\n\n\n4.3\n\n\n1\n\n\nBackflows to refineries\n\n\n0\n\n\n9-10\n\n\nLosses\n\n\n1-2\n\n\n1-2\n\n\nA recent report provides an excellent overview of state-of-the-art naphtha cracking\ntechnologies based on information from technology suppliers as well as the various\nfurnace and separation related features offered by the licensors. An overview of the\nethylene yields and specific energy consumption values are provided in Table 2.6.4.\n\n\n81\n\n\n79 The figures should be considered only indicative, because they were taken from various sources partly\napplying different system boundaries and assumptions.\n\n\n80 Ren, T., Patel, M., and Blok, K., 2006. \u201cOlefins from conventional and heavy feedstocks: Energy use in\nsteam cracking and alternative processes,\u201d Energy 31 (2006), pp. 425-451.\n\n\n81 Ren, T., Patel, M., and Blok, K., 2006. \"Olefins from conventional and heavy feedstocks: Energy use in\nsteam cracking and alternative processes,\u201d Energy 31 (2006), pp. 425-451.\n\n\n43\n\n\nTable 2.6.4. State-of the-Art Specific Final Energy Use Values for Naphtha Steam per\nlicensor.\n\n\n82\n\n\nKellog Brown\n& Root\n\n\nTechnip-\nCoflexip\n\n\nABB\n\n\nLinde\n\n\nStone &\nWebster\n\n\nLummus\n\n\nAG\n\n\nEthylene\n\n\n34.4%\n\n\n35%\n\n\n38%\n\n\n35%\n\n\nNo data\nfound\n\n\nYield wt.%\n\n\nBest: 18.820\nTypical: 21.6 -\n25.2\n\n\nBest: 21\n\n\n20-25\n\n\nNo data found\n\n\nEnergy use\n\n\nBest: 18\n\n\nGJ/t ethylene\n\n\n(w/gas\nturbine)\nTypical: 21\n\n\nIf a yield factor of 1.83 ton HVCs/t ethylene is applied to the energy consumption figures\npresented in Table 2.6.4., then the current best-practice naphtha crackers offered by\nlicensors have a specific energy consumption of 11 to 14 GJ/t (375 to 478 kgce/t). These\nbest-practice naphtha crackers focus on the design optimization of the furnace coils to\nimprove heat transfer, minimize coking and maximize the yield of olefins. In the\nseparation process, front-end demethanization can reduce refrigeration energy needs. The\ngas turbine technology also mentioned in Table 2.6.4 refers to a process option where a\nsteam cracker is operated integrated with a gas turbine, producing steam and electricity.\nThe hot off-gases from the turbine are used for feedstock heating. This option could save\n3 GJ/t (102 kgce/t) ethylene, but this option has not been used widely so far. No\ncomparable detailed data could be found on best-practice ethane cracking. Therefore,\nthe\nlowest value found from the literature (12.5 GJ/t HVCs or 427 kgce/t HVCs) is regarded\nas the best practice ethane cracking technology.\n\n\n2.6.2 Other Feedstocks and Emerging Technologies\n\n\nBesides naphtha and ethane, also other feedstocks are applied in steam cracking such as\nLPG (mixtures of propane and butane) and gas oil (heavier feedstock compared to\nnaphtha). Generally speaking, the specific energy consumption of gas oil crackers is\nsomewhat higher compared to naphtha crackers (e.g. a range of 18 to 23 GJ/t HVCs or\n614 to 785 kgce/t HVCs). LPG has process characteristics between ethane and naphtha\ncracking.\n\n\n83\n\n\nTwo recent publications provide an excellent overview of wide variety of possible\nethylene and propylene technologies from conventional and heavy feedstocks, including\nadvanced naphtha cracking technologies and from natural gas.' They conclude that all\n\n\n84, 85\n\n\n82\n\n\nRen, T., Patel, M., and Blok, K., 2006. \u201cOlefins from conventional and heavy feedstocks: Energy use in\nsteam cracking and alternative processes,\u201d Energy 31 (2006), pp. 425-451.\n83 Institute for Prospective Technological Studies, 2003. Reference document on best available technologies\nin the large volume organic chemical industry, Seville, Spain: European Commission, Joint Research\nCentre, IPTS.\n84 Ren, T., Patel M., and Blok, K., 2005. Steam Cracking and Natural Gas-to-Olefins: A Comparison of\nEnergy Use, CO2 Emissions and Economics. 2005 Spring National Meeting of American Institute of\nChemical Engineers (AIChE), Atlanta, USA.\n85 Ren, T., Patel, M., and Blok, K., 2006. \u201cOlefins from conventional and heavy feedstocks: Energy use in\nsteam cracking and alternative processes,\u201d Energy 31 (2006), pp. 425-451.\n\n\n44\n\n\npossible routes from natural gas via methanol production consume about twice the\namount of energy as in conventional steam cracking. Of the processes starting from\nconventional steam cracking, some have very low specific energy consumption values (8\nto 10 GJ/t HVCs or 273 to 341 kgce/t HVCs), but these technologies (e.g. shockwave\nreactors) are still in the laboratory phase. Catalytic cracking and hydro-pyrolisis of\nnaphtha also offer lower specific energy consumption figures (10 to 13 GJ/t or 341 to 444\nkgce/t), but these processes are either not commercially available yet (catalytic cracking\nis in the pilot plant stage) or is not offered by major licensors.\n\n\n3. Summary and Next Steps\n\n\n\"World best practice\u201d energy intensity values, representing the most energy-efficient\nprocesses that are in commercial use in at least one location worldwide, have been\nprovided in this report for the production of iron and steel, aluminium, cement, pulp and\npaper, ammonia, and ethylene. Energy intensity is expressed in energy use per physical\nunit of output for each of these commodities; most commonly these are expressed in\nmetric tonnes (t). The energy intensity values are provided by major energy-consuming\nprocesses for each industrial sector to allow comparisons at the process level. Energy\nvalues are provided for final energy, defined as the energy used at the production facility\nas well as for primary energy, defined as the energy used at the production facility as well\nas the energy used to produce the electricity consumed at the facility.\n\n\nThe \"best practice\" figures for energy consumption provided in this report should be\nconsidered as indicative, as these may depend strongly on the material inputs. For\nexample, energy consumption in steel, cement, and paper production depends strongly on\nthe amount of primary materials (e.g. iron ore, wood/straw) versus secondary materials\n(e.g. scrap, waste paper). These may vary over time depending on the availability, costs,\ncharacteristics and quality, as well as product type and quality. For sectors where such\nvariety exists it may be worthwhile to develop a \u201cbenchmarking\" tool to calculate the\n\"best practice\u201d energy consumption for a specific plant configuration and materials used\nand produced. In such a tool, the user could input the production characteristics and\ncalculate \u201cbest practice\u201d energy use and intensity for a specific plant and its production\nvariables. Also, such a tool could help plants to identify the key areas in a plant for\nenergy efficiency improvement.\n\n\n4. Acknowledgments\n\n\nThis work was funded by the Energy Foundation through the U.S. Department of Energy\nContract No. No. DE-AC02-05CH11231. This report does not necessarily reflect the\nopinion of the Energy Foundation. We wish to thank Dolf Gielen of the International\nEnergy Agency and Klaas Jan Kramer for their helpful comments on a previous draft. We\nalso thank Pasi Rousu for providing additional information on the Chempolis process,\nDavid Fridley of LBNL for his assistance on the section on ethylene production, and\nNathaniel Aden for his contribution to previous drafts. Despite their efforts, any\nremaining errors remain the sole responsibility of the authors.\n\n\n45\n"}, "expected_output": {"claims": [{"unit": "GJ/t", "value": 0.053, "evidence": ["The best practice final energy intensity for cold rolling is 0.09 GJ/t (3.0 kgce/t) steam, fuel use of 0.053 GJ/t (1.8 kgce/t) and electricity use of 87 kWh/t (10.7 kgce/t) cold rolled sheet,", "fuel use of 0.053 GJ/t (1.8 kgce/t)"]}, {"unit": "GJ/t", "value": 0.26, "evidence": ["Best practice energy use for continuous annealing is assumed to be equal to fuel use of 0.73 GJ/t, steam use of 0.26 GJ/t, and electricity use of 35 kWh/t", "Finishing is the final production step, and may include different processes such as\nannealing and surface treatment. The best practice final energy intensity for batch\nannealing is steam use of 0.173 GJ/t, fuel use of 0.9 GJ/t and 35 kWh/t of electricity,\nequivalent to 1.2 GJ/t (41.0 kgce/t). Best practice energy use for continuous annealing is\nassumed to be equal to fuel use of 0.73 GJ/t, steam use of 0.26 GJ/t, and electricity use of\n35 kWh/t, equivalent to final energy use of 1.1 GJ/t (or 38.1 kgce/t). Continuous\nannealing is considered the state-of-the-art technology, and therefore assumed to be best\npractice technology.\n", "steam use of 0.26 GJ/t"]}]}, "metadata": {"product_category": "Electricity, steam & fuels", "request_id": "req_0c64181253b0f53c"}}