Solar made 30.6 MWh in the 12 months to Sep 2026, 7.9% more than the 12 months before. Since the expansion each kWh/m² of sunlight has made about 21.2 kWh.
1 day far outside the rest isn't drawn.
Click a monthTap a month twice to pick it out.
| J | F | M | A | M | J | J | A | S | O | N | D | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2023 | 548 | 1.5k | 1.5k | |||||||||
| 2024 | 1.7k | 2.2k | 2.5k | 2.8k | 2.1k | 2.3k | 710 | 2.5k | 2.2k | 2.2k | 2.1k | 1.8k |
| 2025 | 2k | 1.9k | 2.7k | 3k | 2.7k | 2.5k | 2.6k | 2.5k | 2.3k | 2.4k | 2.3k | 1.8k |
| 2026 | 2.1k | 2.3k | 2.4k | 3k | 2.4k | 2.7k | 3.1k | 2.9k | 3.2k |
| 2023 | 2024 | 2025 | 2026 | |
|---|---|---|---|---|
| Jan | 1.7k | 2k | 2.1k | |
| Feb | 2.2k | 1.9k | 2.3k | |
| Mar | 2.5k | 2.7k | 2.4k | |
| Apr | 2.8k | 3k | 3k | |
| May | 2.1k | 2.7k | 2.4k | |
| Jun | 2.3k | 2.5k | 2.7k | |
| Jul | 710 | 2.6k | 3.1k | |
| Aug | 2.5k | 2.5k | 2.9k | |
| Sep | 2.2k | 2.3k | 3.2k | |
| Oct | 548 | 2.2k | 2.4k | |
| Nov | 1.5k | 2.1k | 2.3k | |
| Dec | 1.5k | 1.8k | 1.8k |
Solar and sunlight plots each measured day's solar against the sunlight measured by the local Tempest station (AVG(solar_radiation_w_m2) × 24 ÷ 1000, kWh/m²), since the station's records begin. Days before the Jul 3, 2026 expansion are pale blue, because a smaller array made them. The trend is a line through zero fitted to the days since: Σ sunlight × solar ÷ Σ sunlight², 21.2 kWh per kWh/m². The ringed dot is the latest day. Days far below the line are worth a look; a day whose readings are clearly broken is left off the chart, and estimated days are left out.
Year over year adds up each year's bills from January, so this year's pace reads against the same point in earlier years. Only bills with solar for the whole billing period count, so a line starts at its year's first full solar bill (November 2023, the bill after the October 2023 install bill) and stops at the last bill on file. Each month across years puts one dot per year on each bill month, the same full solar bills only, with a dashed tick for the PVWatts projection for that month at each array size the system has had. The answer compares the latest 12 bills with the 12 before only when all of those had solar too.
Estimated bills count everywhere: the 2023–24 bills' solar and house use are round-number estimates, Jul '24's house use comes from the energy balance, and Jul '26 was partly metered. They are hatched in the dot plot and the heatmap, and named in the year-over-year legend and tooltips (a line can't be hatched, and a dashed line would read as a projection).
The answer's 12-month total adds up the latest bill and the 11 before it, so the seasons cancel out. Capacity factor is solar against what the array would make running flat out: solar_kwh ÷ (system_size_dc × hours_in_period). Over the last 12 bills it averaged 18%. Rooftop systems in warm climates usually run 15–20%; each bill is measured against the array in service on each of its days.