A system that carries the house for six hours and then goes dark the instant a compressor starts has not run out of energy, and those two events do not share a cause.
The assumption, stated plainly
The call is usually reported as a capacity problem. The house went dark, a big appliance was starting, and the battery is assumed to be too small for the house.
The timing is what contradicts it. An energy shortage announces itself as time: the lights dim, the state of charge falls, the house runs on for a while and then stops. What is being described here is a shutdown inside the first second of a load step, with the state of charge nowhere near empty. An energy figure cannot explain an event that finishes before it begins.
Where that assumption breaks
The limit that a load step meets appears in the published specification, and it is stated inconsistently across this range, which is where the assumption fails rather than simply being incomplete.
The standalone inverters publish their overload allowance as twice the rated power for ten seconds. The all-in-one units publish a surge magnitude instead, 9 kW on the 6 kW model and 15 kW on the 10 kW model, and neither page attaches a duration to it. Two further models in this range publish no surge figure at all, so a buyer cannot check the same question against those pages by any route.
The transfer behaviour is published with the same kind of variation. Two inverters state a backup switching time below 10 milliseconds and describe it as seamless. The split-phase model states below 20 milliseconds and adds a qualifier: seamless for most loads. That qualifier is a published admission that the transfer is not seamless for every load, and a compressor starting around the moment of transfer is the case it leaves open.
So the surge number a buyer relied on is complete on four pages, incomplete on two, absent on two, and the switching time carries a condition on one. Those figures are not on the same basis, because the measurement condition behind each one differs: a ratio with a ten-second window, a kilowatt figure with no window, and a switching time with a qualifier. No page Ruibit publishes states both the surge magnitude in kilowatts and the duration behind it, which is why the question returns to the supplier rather than to the catalogue.
Published figures are quoted from the product pages in this range, verified as of 2026-09. The signature column describes the distinguishing pattern this article proposes for each owner; the published pages do not state these patterns.
Worked example
The scenario below is constructed to show the diagnostic order rather than taken from a specific site.
A house with an all-in-one unit reports that the system goes down whenever the air conditioning starts, and comes back by itself within a second. The homeowner's proposal is a larger battery. Three readings separate the candidates before anything is opened. The state of charge at the time of the event is well above the reserve setting, which removes the energy explanation and the reserve setting together. Neither the inverter nor the battery has an alarm within the minute that contains the event, which argues against both the output limit and the discharge limit. The state of charge afterwards is unchanged to within a few percent, which argues against a discharge event having taken place at all.
That combination points at a load step that landed close to the transfer rather than at a limit inside the system, and the published switching time for that model carries the qualifier that leaves the case open. The capacity proposal would not have touched any of the three readings.
The correct statement
A shutdown on appliance start is a protection or transfer event with an owner, and the owner is identified by two pieces of evidence: which device logged the event, and how the system came back.
If the system returns by itself within the same second, the event is a transfer or a device outside the system, because a latched protection does not clear itself. If an inverter alarm appears, the AC side limit is the owner. If the battery logs the event, the DC side limit is the owner. If the shutdown lasts about as long as the appliance start lasted, the published surge window is the owner, which is the one case where the duration attached to the surge figure decides the outcome.
Consequence
The consequence for the purchase decision is that capacity and load-step capability are separate purchases. Upsizing the battery addresses an energy complaint and leaves every limit in the table above where it was: the surge window is set by the inverter model, the discharge limit by the battery model, and the transfer time by the inverter page that publishes it.
Two questions belong in writing before the order. Ask for the duration attached to the published surge figure, since four pages in this range state one, two state only a magnitude, and two state neither. Ask which loads the published switching time is seamless for, because one page in this range attaches that condition and the other pages do not.
Two items cannot be corrected after delivery. The surge window and the switching time are properties of the inverter model ordered rather than settings, so neither changes with a firmware update or with more battery capacity.
At handover, record the coincident load the system was commissioned against, the surge duration published for the installed inverter, and the switching time with its qualifier. Those three lines are what a later load-step complaint is measured against, and the first of them is the one that is usually missing.
FAQs
1. Does a home battery shut down on appliance start because it is too small?
Usually not. An energy shortage appears as time, with the state of charge falling before the house stops. A shutdown inside the first second of a load step points to a limit or a transfer event, and the state of charge at that moment is the reading that separates the two explanations.
2. Which limits can cause it?
The inverter's output allowance, the battery's discharge limit, and the backup transfer. The published figures differ by model: the standalone inverters publish twice rated power for ten seconds, the all-in-one units publish 9 kW and 15 kW of surge without a duration, and two models publish no surge figure at all.
3. How is the owner of the event identified?
By which device logged it and how the system recovered. A latched protection does not clear itself, so a system that returns within the same second points to a transfer or to a device outside the system. An inverter alarm points to the AC side limit, a battery event to the DC side limit.
4. Can a larger battery fix it?
It addresses an energy complaint rather than a load-step event. The surge window is set by the inverter model, the discharge limit by the battery model, and the switching time by the inverter, so more capacity leaves all three where they were.
5. What should be confirmed before ordering?
The duration attached to the published surge figure, since four pages in this range state one, two state only a magnitude, and two state neither. Also ask which loads the published switching time is seamless for, because only one page in this range attaches that qualifier.
