The Home Battery ReportIndependent · No installer money
spec datadatasheetsbackup powerwhole-home backup

How fast do home batteries switch over in an outage? Transfer times by model

Five of the 17 home batteries we track publish a transfer time, from under 16 ms to about 100 ms. Twelve publish the word seamless and no number. What that means for the devices in your house that reboot.

The Home Battery Report8 min read✓ Verified

Every home battery brochure uses the same word for what happens when the grid fails: seamless. It appears on Tesla's datasheet, on Enphase's, on Generac's. It is not a specification, and the products that use it most confidently are frequently the ones that publish no number behind it.

We went through the manufacturer datasheets for all 17 models on our comparison pages, plus the separate datasheets for the gateways, controllers and disconnect switches that actually do the switching, and recorded every published transfer time with its basis. Five products publish a figure. Twelve do not. Every figure below is a published manufacturer specification, and we have not tested any of these products.

The five that publish a number

Note where all five of those numbers live. Not one is on the battery's own spec table. They are on the controller, the disconnect switch, the backup controller, the SmartBox and the inverter block. The battery does not transfer anything; the islanding device does. If an installer quotes you a transfer time, the right follow-up question is which box it belongs to.

The twelve that publish nothing

Four of the twelve have a good reason. The SolarEdge Home Battery 400V, EG4 PowerPro WallMount, Fortress Power eVault Max 18.5 and BYD Battery-Box Premium HVS are DC batteries with no inverter and no grid connection. The transfer happens in a separate inverter or backup interface, which has its own datasheet and is not what you are buying when you buy the battery.

The other eight are complete AC systems that do include an islanding device, and publish no number for it.

We are not saying those eight systems transfer slowly. We have no idea, and neither does anyone quoting a number for them. We are saying the manufacturer chose to publish an adjective instead of a measurement, and that if the speed of the transfer matters to your house, you have nothing to hold the maker to.

A home battery is not a UPS

An uninterruptible power supply keeps a load energised continuously by running it off an inverter fed from a battery that is always in the path, or by switching so fast the load never notices. A home battery system does something different: it detects that the grid has failed, opens a contactor or a solid-state switch to disconnect from it, and then energises the backed-up loads from its own inverter. There is a gap. It is small, and on the fastest published products it is very small, but it exists by design.

You can check this in the compliance tables. The datasheets in our set that publish a compliance block name standards for grid-interactive inverters and energy storage: UL 1741 and its SA, SB and PCS variants, UL 9540 for the system, UL 1973 for the battery, and UL 67 or UL 869A for the panelboard and service equipment functions. Some also reference UL 9540A, which is a test method rather than a listing standard, so a datasheet says a unit was evaluated or tested to it and certified to UL 9540. Not one of them lists UL 1778, the safety standard for uninterruptible power supplies. That is not an oversight. These are not UPS products and they are not certified as such. The plain-English version of both standards is in our UL 9540 and NFPA 855 guide.

So what does the gap actually mean for the things in your house? The most commonly cited reference is the Information Technology Industry Council's ITI (CBEMA) Curve Application Note, which describes an AC input voltage envelope that can typically be tolerated with no interruption in function by most information technology equipment, and states that an interruption within that envelope may last up to 20 milliseconds. It applies to 120V nominal voltages obtained from 120V, 208Y/120V and 120/240V 60 Hz systems, which is the US residential case. Source: ITI (CBEMA) Curve Application Note, listed on the ITI resources page, read August 6, 2026. Read against that reference:

  • FranklinWH at under 16 ms and Anker at under 20 ms fall inside it.
  • Generac at under 50 ms, Savant at under 70 ms and Panasonic at about 100 ms fall outside it.
  • The other twelve cannot be placed either way, because they publish nothing.

Two honest caveats travel with that. The ITI curve is a statement about typical equipment as a class, not a promise about your specific desktop, router, NAS or games console, and equipment manufacturers do not generally publish a ride-through figure for individual products. And a published transfer time is a rated maximum for the switching device, not a measurement of what happens in your house with your loads on a specific fault.

The practical version: if a device must not reboot, put a small dedicated UPS between the outlet and that device, and treat the home battery as what keeps the UPS charged for the following several hours. That is a belt-and-braces arrangement, and it is the arrangement people who care about this actually run. Our work from home outage kit covers what that looks like for a home office.

Where a fast transfer earns its keep, and where it does not

Most of a house does not care. A refrigerator compressor, a furnace fan, lights, a water heater and an EV charger all tolerate a 100 millisecond interruption without any consequence beyond a flicker. For those loads the transfer time is a spec sheet curiosity, and capacity and continuous power are what decide whether the system is any good, which is the argument in our complete backup guide.

The loads that do care are narrow: desktop computers with unsaved work, network equipment and NAS devices that take minutes to come back, some smart home hubs, aquarium and sump controllers with a startup sequence, and certain medical devices. On medical equipment specifically we will not generalise: whether a given device rides through a brief interruption, and what its manufacturer says about backup power, is a question for that manufacturer, and our CPAP battery backup guide covers what those makers actually publish rather than what a battery maker implies.

It is also worth setting expectations against the alternative. A standby generator has to detect the outage, crank an engine, bring it up to speed and stabilise before its transfer switch closes. Every home battery in this category is in a different order of magnitude from that, whether or not it publishes a figure. The full comparison is in battery vs generator.

What to ask before you sign

  • Which box performs the transfer, and what does its own datasheet say? Not the battery's sheet. The gateway, controller, disconnect switch or backup interface.
  • If no figure is published, ask for one in writing. A maker that publishes "seamless" and will not put a number behind it has told you something.
  • Decide which loads genuinely cannot reboot, and plan a small UPS for those specifically rather than choosing an entire storage system on this one spec.
  • Check what is on the backed-up panel at all. A transfer time is irrelevant for a circuit that is not backed up, and which circuits those are is the whole-home versus essential backup decision.

Every figure above is a manufacturer datasheet claim recorded with its published basis, or a recorded absence of one. None of it is a measurement we made, and none of it is a promise about how a specific device in your house will behave when the grid drops.