Build the application

## Build
Compare commercial building envelope glazing and insulation upgrades against peak chiller electrical demand in EnergyPlus models. Formulate envelope retrofit comparisons and recommend chiller downsizing scenarios.

## User
Commercial Energy Retrofit Consultant

## Workflow
1. Ingest baseline EnergyPlus epJSON building envelope properties and design day cooling peak loads.
2. Model synthetic window glazing U-factor and roof insulation retrofit packages to isolate load reduction impacts.
3. Calculate avoided peak cooling kW and downsized chiller capacity requirements based on envelope thermal improvements.
4. Generate structured engineering comparison tables for facility owner review, explicitly avoiding ungrounded financial payback claims.

## Interface
Streamlit web application featuring retrofit envelope sliders and simulated load curve charts.

## Data
Use open-source standard commercial building EnergyPlus epJSON archetypes from NREL repositories modified for ASHRAE Zone 1B.
Starter examples below are synthetic. They are not downloaded records.

## Source
[EnergyPlus](https://raw.githubusercontent.com/NREL/EnergyPlus/develop/README.md)
Extracts building envelope constructions and peak design cooling loads directly from EnergyPlus epJSON schema objects to isolate peak demand impacts.

## Tools
langgraph, fastapi, streamlit.
LangGraph manages the multi-step evaluation state across envelope modeling, sizing calculations, and engineering synthesis.
Hosted or local LLM with high context reasoning for building system parameters and engineering logic.

## Deliverables
Submit runnable source, pinned dependencies, setup commands, a working interface, example outputs, and a recorded end-to-end demo. Include tests and an action log.

## Acceptance
- Parses valid EnergyPlus epJSON Construction and ZoneInfiltration definitions.
- Calculates simulated chiller capacity reduction in refrigeration tons (RT) and kW draw based on provided conversion factors.
- Compares retrofit capacity limits against peak demand loads without making unsupported simple payback generalizations.
- Flags envelope retrofits yielding insufficient peak demand drop for human re-evaluation.

## Starter Examples
```json
[
  {
    "input": {
      "record_type": "Simplified synthetic app record",
      "facility_id": "TOWER-AL-NOOR",
      "baseline_peak_kw": 1850,
      "baseline_chiller_tons": 600,
      "u_factor_delta_absolute": 1.8,
      "kw_reduction_per_u_factor_drop": 200,
      "tons_reduction_per_avoided_kw": 0.333333
    },
    "expected_output": {
      "facility_id": "TOWER-AL-NOOR",
      "projected_peak_kw": 1490,
      "avoided_peak_kw": 360,
      "recommended_chiller_tons": 480,
      "status": "APPROVED_FOR_ENGINEERING_REVIEW"
    }
  },
  {
    "input": {
      "record_type": "Simplified synthetic app record",
      "facility_id": "WAREHOUSE-SOUTH",
      "baseline_peak_kw": 320,
      "baseline_chiller_tons": 90,
      "u_factor_delta_absolute": 0.4,
      "kw_reduction_per_u_factor_drop": 55,
      "tons_reduction_per_avoided_kw": 0.333333
    },
    "expected_output": {
      "facility_id": "WAREHOUSE-SOUTH",
      "projected_peak_kw": 298,
      "avoided_peak_kw": 22,
      "recommended_chiller_tons": 83,
      "status": "INSUFFICIENT_PEAK_DROP"
    }
  }
]
```

## Safety
HVAC equipment downsizing numbers are simulation estimates; final equipment procurement requires licensed professional mechanical engineering sign-off.
Use mock external writes. Require human review before any real deployment.