Solar ROI Analysis

A Jupyter notebook series analyzing the return on investment for a residential solar installation, using real weather station and PG&E utility data.

Notebooks

1. 01 - Solar Generation & Usage Offset — Combines solar radiation data from an AmbientWeather station with PG&E 15-minute interval usage data to estimate how much energy 16 solar panels (20.3% efficiency, 31.52 m²) would have generated and offset over a year.

2. 02 - Energy Cost Calculation — Models PG&E rate plans (E-TOU-C and E-1) including time-of-use tiers, seasonal baselines, and 3CE generation charges to compare annual energy costs with and without solar.

Data

Raw (data/raw/)

• ambient-weather-*.csv — Solar radiation (W/m²) at 5-minute resolution from AmbientWeather station
• pge_electric_interval_data_*.csv — PG&E electricity usage (kWh) at 15-minute resolution via Green Button Download

Processed (data/processed/)

• usage_data_with_generation.csv — Usage data enriched with solar generation and offset columns
• usage_data_with_generation_and_energy_cost.csv — Further enriched with modeled energy costs per rate plan

Setup

pip install pandas matplotlib tqdm pytz
jupyter notebook

Key Findings

Based on Feb 2023 – Jan 2024 data:

• 16-panel system would have generated ~7,385 kWh/year
• ~41% reduction in grid consumption
• ~4,586 kWh sold back to the grid
• ~$1,402/year in energy cost savings (~38%) switching from E-TOU-C to E-1 with solar