README.md

metadata

title: Upfront_Embodied_Carbon_Calculator
emoji: 📚
colorFrom: purple
colorTo: gray
sdk: gradio
sdk_version: 5.47.2
app_file: app.py
pinned: false

A1–A5 Upfront Embodied Carbon Calculator (Gradio)

A single-page Gradio app to compute A1–A5 upfront embodied carbon for building materials and processes. Enter rows of data (materials, transport, site energy/waste), get per-row results, per-module summary, total tCO2e, and optional reduction suggestions powered by a small LLM via Hugging Face Inference API.

Features

• Deterministic A1–A5 backend (EN 15978 scope; RICS/IStructE/GHG Protocol units)
• Units and aliases:
  • A1–A3/A5 materials/waste: kg|t × kgCO2e/kg or tCO2e/t
  • A4 transport: ton_km × kgCO2e/ton_km (aliases: t-km, t·km, ton-km)
  • A5 site energy: kWh × kgCO2e/kWh
• Output: per-row kgCO2e|tCO2e, per-module summary, Total
• Optional suggestions via Hugging Face Inference API (falls back to deterministic tips if no token)

Run locally

pip install -r requirements.txt
python app.py

Open the local URL printed by Gradio.

Deploy to Hugging Face Spaces

• Create a new Space (Gradio Python)
• Upload these files:
  • app.py
  • requirements.txt
  • README.md (optional)
• In Space Settings → Secrets, set one of the following if you want LLM suggestions:
  • HUGGINGFACEHUB_API_TOKEN (recommended), or
  • HF_API_TOKEN

If no token is set, the app still runs and shows rule-based fallback suggestions.

Data columns (required)

module,name,quantity,unit,ef,ef_unit,notes

Examples:

• Concrete (A1): 10000 kg × 0.12 kgCO2e/kg
• Rebar (A1): 2000 kg → 2 t × 1.9 tCO2e/t
• CLT (A3): 5 t × 0.35 tCO2e/t
• Truck transport (A4): 800 ton_km × 0.1 kgCO2e/ton_km
• Construction waste (A5): 2 t × 0.25 tCO2e/t

Notes

• Backend validates missing columns, unsupported modules, and unit mismatches with clear errors
• Transport aliases are normalized to ton_km
• Keep provenance in notes (EPD/database/assumptions) for auditability

License

MIT (user content and EPD/IP remain their respective owners)