# Power Compare

Compares your Sungrow solar system's actual usage against every electricity
plan the AER publicly lists for your Queensland distributor, so you can see
which plan would actually cost you the least given how much you generate,
consume, import and export -- no battery/storage assumed.

## How it works

- An always-on background scheduler polls your Sungrow iSolarCloud plant
  every few minutes and stores interval generation/load/import/export data
  in a local SQLite database, building up history over time.
- It also periodically refreshes a cache of every current QLD electricity
  plan from the AER's public "Get Generic Plans" / "Get Generic Plan
  Detail" APIs.
- A costing engine evaluates every cached plan (flat or time-of-use, with
  supply charge and feed-in tariff credit) against your actual usage
  history and ranks them by estimated annual cost, compared against your
  own plan (entered manually in Settings).
- A small web dashboard shows your usage over time and the plan comparison
  table.

## Quick start (no real solar data yet)

The app runs entirely on synthetic mock usage data until you've registered
a Sungrow developer app (see [SETUP.md](SETUP.md)) -- useful for seeing how
the comparison works today.

```bash
python3 -m venv .venv
source .venv/bin/activate
pip install -e ".[dev]"
cp .env.example .env

# optional: instantly backfill a year of synthetic usage data instead of
# waiting for the scheduler to accumulate it in real time
python scripts/seed_sample_data.py --days 365

uvicorn app.main:app --reload
```

Then open http://localhost:8000, set your distributor/postcode and current
plan under Settings, and check the Plan Comparison page.

## Running the tests

```bash
pytest
```

The costing engine and the AER plan mapper are both unit-tested, including
against real plan data fetched from the live AER API (see
`app/aer/tests/sample_*.json`).

## Connecting your real Sungrow system

See [SETUP.md](SETUP.md) for registering a developer app on the iSolarCloud
portal and connecting your account.

## Running as an always-on service

```bash
docker compose up -d
```

This persists the SQLite database in a named Docker volume and restarts
automatically. See `.env.example` for configuration.