Why Does My Home Battery Lose Charge Overnight Even When Nothing Is Running?

Put a number on the overnight loss

Percentages hide the scale. Convert the drop into energy: a battery with 13.5 kWh of usable capacity that falls from 90% to 70% has supplied about 2.7 kWh. Over a ten-hour night that is an average of about 270 W, which is an entirely ordinary base load for a house with a refrigerator, a freezer, networking gear, a few chargers and a gas furnace's blower control or a heat pump's crankcase heater. The same arithmetic in reverse tells you what to expect: base load in watts multiplied by hours of darkness, plus the system's own use, is roughly what the battery should lose.

So the question is not "why did it drop?" but "does the drop match the load?" If the monitoring shows 250 W of house load all night and the battery lost 2.5 kWh, there is no fault, just a bigger base load than you thought. If the monitoring shows 50 W and the battery still lost 2.5 kWh, the energy went somewhere the monitor isn't seeing, and that is worth chasing.

Symptoms and likely causes

Always-on loads add up

"Nothing is running" usually means nothing you can see. A modern home keeps a surprising number of devices awake: the router and mesh nodes, smart speakers, game consoles in instant-on mode, cable or satellite boxes, TV standby, security cameras and their recorder, doorbell transformers, irrigation controllers, smoke and CO alarms on hardwired circuits, and the control boards of the washer, dryer, range and microwave. Individually they are a few watts each; together they are often 50 to 150 W before the fridge compressor cycles.

The highlighted item catches people out. Many heat pumps and air conditioners heat their compressor crankcase in cold weather to keep refrigerant from settling in the oil, and that draw continues all night even when no heating is called for. Our heat pump cold-weather guide explains the other winter loads to look for.

Inverter and battery self-consumption

The equipment that runs the battery uses power too. A hybrid inverter that stays on overnight to cover house loads draws its own idle power, the battery management system and communications run continuously, and some batteries heat or cool their cells when the temperature leaves a comfortable band. Converting stored DC back to AC also loses a few percent of whatever passes through. Makers publish idle or standby consumption in their datasheets, often in the tens of watts; read your system's figure rather than assuming one.

This self-use is often invisible in the household load graph, because the CT clamps that measure house load sit downstream of the inverter. You see it only as a gap between what the battery reports supplying and what the house reports using. A gap of a few tens of watts is normal. A gap of several hundred watts deserves an installer visit.

Settings that empty the battery on purpose

Many overnight losses are the system doing exactly what it was told. Most systems let you choose a mode that prioritizes self-consumption, backup or time-of-use savings, set a backup reserve, and allow or block grid export from the battery. A self-consumption mode will happily run the house from the battery all night, down to the reserve. A time-of-use or virtual power plant program may export to the grid in the evening peak, which shows up as a steep early-evening drop that the house load can't explain.

Read the current mode, reserve and export settings in the app, and check whether you have enrolled in a utility or aggregator program that can dispatch the battery. If settings changed after a firmware update, note the date for the installer. Our guide to storage systems that stop discharging at peak times covers these modes from the other direction.

When the monitoring is wrong

The system learns about house load and grid flow from current transformers, the CT clamps an installer puts around conductors in the panel. A clamp installed facing the wrong way, placed on the wrong conductor, or not fully closed can make the app report import as export, show negative house load, or miss a whole circuit. The battery then responds to a picture of the house that isn't true, for example discharging to "cover" load that is really solar going out to the grid.

A daytime sanity check catches most of these. On a sunny afternoon with the house quiet, solar production should roughly equal house load plus battery charging plus export. If the numbers don't add up, or the utility meter disagrees with what the app says you imported overnight, photograph the screens and send them to the installer. Re-seating or reorienting CT clamps happens inside the panel, so it is their job.

1How to measure your baseline

• Read the utility meter. Note the import register at bedtime and at dawn on a night when the battery is set to backup-only. That night's grid import is your true base load without the battery in the picture.
• Survey plug-in loads. A plug-in energy meter between the outlet and each always-on device gives watts and accumulated kWh. Leave it on the fridge for a full day to get an honest average.
• Clamp cord-connected appliances. Through a line splitter, an AC clamp meter reads the current of a single appliance cord without opening anything. A clamp that reads watts saves you the voltage-times-current conversion and accounts for power factor.
• Switch circuits off one at a time. Using breaker handles only, turn off one non-essential circuit for twenty minutes and watch the base load in the app. Skip the refrigerator, freezer, medical equipment and anything that resets awkwardly.

Use a CAT III rated meter and safety glasses for any measurement on a live outlet, and see how to measure AC current with a clamp meter for why you clamp one conductor, not the whole cord.