Why split it out by array
The "solar generation" number I usually quote is really four different arrays lumped together, and lumping them together hides the interesting part. The newer Enphase system is 18 panels across the North and South faces of the main roof, plus 4 more on the Garage — all 510W Aiko panels on individual IQ8HC microinverters, so every single panel reports its own output. Sitting alongside that is the much older array on the lower roof, wired into the original Sonnen hybrid inverter and 15kWh battery from the 2022 install. Four arrays, four orientations, four different stories — so for this half-year update I pulled the numbers apart properly.
The four arrays, H1 2026
Enphase's per-panel monitoring lets me sum every individual panel reading for the half-year and attribute it to the right roof face. Combined with the Sonnen array's own generation meter, here's how the half-year actually split:
South comes out on top for the half-year, with the Sonnen array a close second despite being the oldest kit on the roof. North still pulls a very usable 1,549 kWh — proof again that a north-facing roof with microinverters is worth using, not writing off. And the Garage array, at just 4 panels, is the smallest total simply because there are fewer of them.
About those Garage panels: they weren't part of the original spec, and they went in earlier than the rest of the Enphase system — June 2025, versus October 2025 for the North and South main-roof arrays. When the 18 main-roof Aiko panels were ordered, four arrived scratched; rather than have them scrapped, the supplier let me keep them, so I put them up on the garage roof nobody can see from the ground, four months ahead of the rest of the system. That head start — including a full extra summer of 2025 generation before North and South even existed — is the real reason the Garage array shows the highest lifetime per-panel output on site. It's not that damaged panels perform better; it's that they've simply had longer to rack up kWh. Once you look at a window where all three arrays were running side by side, like this H1, the picture evens out: Garage's 267 kWh/panel sits just behind South's 278 kWh/panel and comfortably ahead of North's 172 kWh/panel — still a very respectable return for panels that would otherwise have gone to landfill.
The seasonal ramp
None of this will surprise anyone who's lived through a British winter, but the swing is still stark laid out month by month. January and February combined managed less than a quarter of what May produced on its own.
| Month | Enphase (kWh) | Sonnen (kWh) | Combined (kWh) | Grid export | Grid import |
|---|---|---|---|---|---|
| January | 307.9 | 114.7 | 422.6 | 85.7 | 1,400 |
| February | 344.6 | 160.5 | 505.1 | 116.7 | 1,200 |
| March | 743.7 | 369.6 | 1,113.3 | 558.4 | 992 |
| April | 1,100 | 548.7 | 1,648.7 | 912.0 | 791 |
| May | 1,300 | 612.4 | 1,912.4 | 1,100 | 843 |
| June | 1,300 | 574.1 | 1,874.1 | 1,000 | 597 |
| H1 total | 5,100 | 2,380.0 | 7,480 | 3,700 | 5,800 |
May turned out to be the standout month at 1,912 kWh combined, just ahead of June's 1,874 kWh. From January's 423 kWh to May's peak, output climbed roughly 4.5× — exactly the kind of ramp you'd expect as day length and sun angle improve through spring.
The battery: 1,090 kWh in, 1,348 kWh out
The Sonnen battery (15kWh, 3.6kW inverter) charged 1,090 kWh and discharged 1,348 kWh across H1. At first glance that looks like it's conjuring energy from nowhere — discharge exceeding charge shouldn't be possible. It isn't: some solar charging happens on a path that isn't fully captured by the "charged" meter total I pulled from Homey, and there's also whatever state of charge was already sitting in the battery on 1 January carrying into the count. It's a metering quirk rather than a mystery, and a reminder that reconciling three separate monitoring systems (Enphase, Sonnen/Homey, and the grid meter) rarely lines up to the decimal place.
Grid position and the money
Across H1 the house imported 5,800 kWh from the grid and exported 3,700 kWh back out, for a net import of 2,100 kWh over the half-year. Import still outweighs export by a fair margin — expected for a system offsetting demand through a UK winter and spring but not yet net-positive. That's much more likely in H2 once summer daylight properly kicks in.
// The financial headline
Using flat rates of 15p/kWh for the Smart Export Guarantee and 24.5p/kWh as the import rate avoided: roughly £555 in SEG export income, plus roughly £926 in avoided import cost from self-consumed solar, for a combined H1 benefit of about £1,481. These are conservative estimates on flat rates rather than actual time-of-use billing, but they're a solid read on what four arrays and a battery are worth so far this year.
What I'm watching for in H2
Three things: whether June's momentum carries into a genuinely net-positive July and August; whether the Sonnen array holds its position against the newer Enphase roofs as the seasons turn again; and getting the battery charge/discharge metering properly reconciled so H2's numbers are tighter than H1's. If you're running a similar multi-array-plus-battery setup and have solved that metering puzzle, I'd love to hear how.