Budding · Taking shape, still changing
How much power a hybrid inverter burns while doing nothing
I run a small PV system – 2.5 kWp on microinverters – that mainly covers the house’s daytime use. A friend with a bigger setup (about 5 kWp, also on microinverters) added a home battery with a hybrid inverter (a Deye, if I remember right) and mentioned that the inverter alone draws 80–120 W, depending on what it is doing.
The datasheet gave no figure for self-consumption. Online, everyone calculates how much energy you can store and get back, but almost nobody asks how much the device uses just to keep itself running.
The scale of it
100 W sounds harmless, but it runs around the clock:
- 2.4 kWh a day,
- about 72 kWh a month,
- about 876 kWh a year.
My house uses about 6 kWh a day, so the inverter idling alone would be 40% of that.
What specific devices draw
Manufacturers rarely publish self-consumption, so I collected the numbers forum users reported. These are individual measurements, not datasheet values:
| Device | Reported self-consumption |
|---|---|
| Deye 10 kW with battery | about 75 kWh a month (~2.5 kWh a day) |
| FoxESS 5 kW | about 170 W while working; 23–45 W on standby (the maker says 15 W) |
| Sumry 10 kW | 5–10 W when running from the grid, 40–80 W from the battery |
| Easun iSolar SMG II 4 kW | 25–60 W from the grid, 70–80 W under normal load |
Victron reportedly has the lowest self-consumption, but its price kills the business case for a system this small. Switching the inverter off at every passing cloud does not help either – waking up from standby usually takes tens of seconds.
When a battery next to PV makes sense
I scaled my winter production up to a system four times bigger (10 kWp) and subtracted 75 kWh a month for running the battery:
| Month | 10 kWp production | After subtracting 75 kWh |
|---|---|---|
| November | 168 kWh | +93 kWh |
| December | 108 kWh | +33 kWh |
| January | 137 kWh | +62 kWh |
With 10 kWp on the roof there is something left over even in winter. With my 2.5 kWp – 27 to 42 kWh a month in those months – the inverter alone would eat more than the panels produce.
What about a battery charged only on the cheap tariff?
I also worked out the case without PV: charge in the cheap zone, use in the expensive one. Energa gross prices from February 2025:
- G12r: 0.47 PLN/kWh in the cheap zone and 1.04 PLN/kWh in the expensive one,
- G11: 0.77 PLN/kWh around the clock.
I assume about 30% losses on charging and discharging, including the battery’s own consumption. A cheap kWh then really costs about 0.61 PLN, 16 grosz less than on G11 – that is what I earn on every kWh that passes through the battery. The inverter idling costs 2.4 kWh × 0.61 PLN ≈ 1.46 PLN a day. To break even, at least about 9.15 kWh a day has to pass through the battery. I use 6 kWh, so for me it does not add up.
Takeaways
- Inverter self-consumption belongs in any battery payback calculation – just like panels and batteries losing capacity with age.
- Inverters are optimised for running close to their rated power. A Polish winter means running at a few per cent of it, where the inverter’s own losses dominate.
- Annual production is a poor measure of payback: on a sunny June day, when RCE prices are negative and batteries are full, the production counter keeps turning but earns nothing.
- With a small microinverter system, a battery makes no sense for me yet. I hope batteries designed as an add-on to microinverters, or for dynamic tariffs, will appear.
Instead of a battery, I shift consumption to the hours when energy is cheap – with the help of my Android app for RCE energy prices.
Source: my thread on elektroda.pl (in Polish) – Zużycie energii przez falowniki hybrydowe – ile wynosi w trybie gotowości? (February 2025).