Sezibwa Solar + Battery Operating Plan
8 × 600 W panels (4.8 kWp), one string on inverter PV1 · prepared 26 Oct 2026
| ≈19.5 kWh solar per average day (15–18 in rainy weeks) |
≈28 kWh site load per day (19–25 Oct, measured) |
≈9.6 kWh midday surplus to store in the battery |
≈65% of the load from solar + battery on a good day |
A day at Sezibwa: solar vs load
| Time of day | Load | Share | Who supplies it (plan) |
|---|---|---|---|
| Night 00–07 | 5.8 kWh | 21% | Grid (cheap night rate) |
| Day 07–18 | 10.6 kWh | 38% | Solar directly; surplus charges the battery |
| Evening 18–24 | 11.8 kWh | 42% | Battery (solar-charged), then grid |
The plan
- Prove the output. Remote PV1 read 27 Oct 09:30 EAT, then midday checks: expect ~300–360 V and 2–4 kW in sun.
- Make room for the sun (key decision). Today the grid charges the battery 00:00–06:00, so it's near full at sunrise and midday surplus is wasted (no export). Stop or cap the night charge (~50%) so solar fills the battery for free.
- Keep the evening discharge. The solar-charged battery covers the 18:00–24:00 peak, the dearest hours.
- Keep a backup floor. Hold 15% (higher in the rainy season) so outages stay covered.
- Make solar visible. OpenEMS reads PV as production (logger #95), then add solar to the daily brief.
- Later: cloudy-day logic and OpenEMS solar-aware charging controllers.
Still open on site: earthing (no earth point yet) for the surge protector and panel rails. Watch: trees beside the array can shade the string; one shaded panel limits all eight.
Decision with Aaron: where to set the 00:00–06:00 grid charge once PV is proven: stop it, or cap it at ~50% for cloudy days.
Commissioning log (26 Oct 2026)
- 8 × 600 Wp in one row on two continuous rail lines (vent removed, end clamps used at one joint for the missing inter-clamp pair).
- 6 mm² PV cable (25 m −, 15 m +); cable sizing checked: 0.4% drop.
- FMPV32 4-pole DC isolator wired per the manual's 4B layout (cables on top 1/3/5/7, links 2–4 and 6–8, go by stamped numbers); 1000 V DC SPD across +/−.
- Open-circuit string 379 V at 16:45 EAT; inverter PV1 read 151.8 V / 0 A at 18:39 EAT (sunset). PV1 registers 0x0107 V×0.1, 0x0108 A×0.1, 0x0109 W.
- Open: earthing (no earth point on site), full-output check 27 Oct, tree shading at one end of the array.