The Days My Solar Beat the Grid — 2026

On 112 days so far this year, the ground-level solar panels in my garden made more electricity across the day than the grid spent running my flat. Not chipped in — beat it outright, midnight to midnight. For a small array behind a one-bed flat, that is a scoreboard I will happily take.

The wins bunch where the light is: 38 of 80 in spring, 57 of 79 in summer, 16 of 47 in autumn, 1 of 49 in winter. Winter, as a garden array well knows, contributes almost none.

July was the strongest month so far — 70 kWh of sun against 48 from the grid, with solar ahead outright on 23 separate days.

Two things about the table. The Solar and Grid figures are the win-days only — the sun and the grid on just the days the panels came out ahead. And the grid figure is what actually ran the flat: I set aside the cheap power I bank in the battery for later, since stockpiling is not the mains doing the day's work.

One deliberate exclusion. I set 34 plunge days aside entirely — days when Agile prices fell to zero or below and it made no sense not to gorge on grid. Racing the sun against a grid column I've deliberately inflated would be rigged, so those days count neither way. Of the 380 kWh I pulled in on them, 282 came in plunge half-hours priced at or below zero — free, or paid to me, not by me.

None of that grid-gorging is luck. The calls about when to pull from the mains come from a charge planner I built in the spring — it hunts the cheapest half-hours, dives on plunge windows, keeps a buffer for the morning, and defends a hard battery floor. I wrote up how it grew from a crude timer into that here.

I am not off-grid, and the year in full keeps me honest: across every fair-race day (plunge days set aside), not just the wins, about 434 kWh of sun against 503 the grid spent running the flat, so it still does most of the lifting. But the win-days are the good days earning their keep, and each brighter week adds another.

The months, side by side