Why does my power station shut off when the pump or fridge starts?
Power basics Updated 2026-08-01
Because a motor does not draw its running watts at the instant it starts. For a fraction of a second the rotor is not turning yet, the motor looks close to a short circuit, and it pulls locked-rotor current — several times its running current. Your machine is comparing that spike against its surge ceiling, not against its continuous rating, and it trips before the appliance ever reaches its running load.
That is why a pump rated well under a station’s continuous output still refuses to start on it, and why the fix is a bigger surge ceiling or a lower inrush — never a bigger battery.
Three loads, three completely different starts
Published start-up figures
Each of these comes off a manufacturer document for a specific model, with the ratio the document itself supports. The spread is the point: "motors need two to three times their running watts" is a rule of thumb that is wrong about every row below.
Converted from the full-load amps on the pump maker’s engineering sheet.
This is volt-amperes, not watts: the manufacturer publishes current, not power factor.
See the load record →From the locked-rotor amps printed on the same row of the same sheet.
Locked-rotor amps divided by full-load amps, both from the maker.
A rotary compressor, and the reason an RV air conditioner is the hardest common load to start.
See the load record →Quoted by the compressor manufacturer, not estimated.
Also a compressor — but a variable-speed one with its own electronic drive, which ramps instead of slamming. Same appliance category, no inrush at all.
See the load record →Not a motor: this is the input capacitors of a switching supply charging, over tens of milliseconds. It is a much larger multiple than any pump and matters far less.
See the load record →officialverified 2026-08-01appliance records
Duration is why the last row is harmless and the first two are not. A capacitive inrush is over in a few line cycles; a motor start lasts hundreds of milliseconds, which is inside the window any surge rating is meant to cover. Our engine bands them by their published duration and enforces them differently, rather than treating every spike as equal.
The rule of thumb that oversizes by three times
The standard tables used to size breakers and motor controllers publish locked-rotor envelopes for each horsepower class. Those envelopes are deliberately generous — they exist so a protective device never nuisance-trips. Read as an appliance figure they are badly wrong in the expensive direction: for the pump above, the code envelope is more than twice what the pump maker measured on its own product. Where we have only an envelope value we label it as one and cap the confidence, because sizing a backup supply off it means buying two or three times the machine you need.
What that does to a verdict
One sump pump, three machines
Live engine runs. The percentage is the running load as a share of that machine’s continuous output — which is exactly why it explains nothing about the third row.
Fails before surge is even reached: the running load alone is above the inverter rating.
Show the working →Half the continuous rating, and the published surge ceiling clears the start.
Show the working →A bigger machine that says no. This record publishes no surge rating at all, so the engine substitutes the continuous output as a conservative floor — an inverter cannot surge below what it sustains — and the pump’s start is above it. We will not assume a ceiling nobody stated.
Show the working →The third row is not a bug and it is the most useful row on this page. The floor the engine used is 2600 W, taken from the continuous output because the brand published no surge figure. A site that quietly assumed "probably twice the rating" would have shown a green light there. We would rather tell you that the manufacturer has not answered the question.