Real-World Runtime Simulator

Spec-sheet math says a battery lasts capacity ÷ watts. This simulator also charges it for the power the inverter burns just by being on. It is a simulation, not a test, and the assumptions are on the page.

1Power station and device

Measured with a plug-in meter and nothing connected. Blank = assumed.
Simulated runtimeDetails

What the simulation does

It steps the battery down every 2 minutes until the protection cuts off, 200 times, each time drawing the uncertain inputs from a range. The battery supplies the load, the inverter's idle draw, 4–8% of the load as conversion loss, and a loss that grows with the square of the load. Usable capacity is 92–98% of the rating. A fridge cycles on and off, and the inverter keeps drawing its idle power while the compressor rests.

Where spec-sheet math misleads most

Take a Wi-Fi router and modem at 15 W. The fixed idle draw is a small share of a small station's budget and a large share of a big one's. Across 54 models, the simulation puts the spec-sheet figure 177% to 318% above its result for this load. Because the idle draw is an assumption, treat this as an estimate of the size of the effect, not a verdict on any one product.

Router + modem: spec-sheet math vs simulated typical runtime (simulated, not measured)
ModelSpec-sheet mathSimulatedSpec sheet overshoots by
Jackery Explorer 5000 Plus11.9 days2.8 days318%
Anker SOLIX F3800 Plus9.1 days2.4 days272%
Anker SOLIX F38009.1 days2.5 days270%
EcoFlow DELTA Pro 39.7 days3.3 days196%
EcoFlow DELTA Pro8.5 days3 days180%
Anker SOLIX F30007.3 days2.6 days177%
Jackery Explorer 24014 hours9 hours51%
BLUETTI Elite 200 V24.9 days3.2 days54%
BLUETTI Elite 100 V22.4 days38 hours54%
EcoFlow RIVER 314 hours9 hours55%

Six worst-affected models first, then the four least affected. The full dataset is in the model pages; the method is in How we calculate.

Help replace the guess with a measurement

The biggest assumption here is idle draw. If you own one of these units and have a plug-in watt meter, the reading with the AC output on and nothing connected is the number that matters. Send it with the model and your meter's make and we'll show it as a reported measurement, labeled as such.

Questions

Is this a test of the power station?
No. It is a computer simulation with assumed inputs. Nobody ran the unit. Where we have a measured result, it is shown separately on the model page under “Measured runtimes”.
Why is the simulated runtime shorter than capacity ÷ watts?
An inverter uses power just by being switched on, and it wastes a bit more as the load rises. With a big load that barely matters. With a 15 W router on a big station it can be most of the draw.
Where does the idle draw come from?
It is an assumption: about 8 W plus 0.6% of the rated output, varied from 60% to 150% of that. Makers rarely publish it. Victron lists 8 W for its 1,200 VA Phoenix inverter with nothing plugged in, which is the one published reference we use. Enter your own measured figure to replace the guess.
How do I measure my own idle draw?
Plug a watt meter between the power station’s AC outlet and nothing else: switch the AC output on with no device connected and read the watts. Use a meter that works at your outlet voltage.

More tools