How Much Solar Does a Home Actually Use? Measured in 117 Los Angeles Homes

These are all 117 homes in our Enphase monitoring fleet with a consumption meter and a clean twelve-month record, 22 of them with a battery. A consumption meter records what the house uses and what crosses the utility meter in each direction, not only what the panels make, so the whole energy balance can be measured. Each home is measured over its latest 12 complete months; for 88 of them that is October 2025 through September 2026.

The two shares answer different questions. Self-consumption asks how much of the solar stayed in the house; self-sufficiency asks how much of the house ran on solar. A typical home without a battery kept 41% of its solar and still bought 62% of its electricity, because the hours do not line up: panels peak around noon, and most households use the most power in the evening.

Homes differ far more than groups do. Without a battery, the middle half of homes used between 32% and 50% of their solar, and 20 homes used less than 30%. Bigger households used more of it: LADWP homes without a battery used a median 38 kWh a day against 25 kWh in SCE territory, and kept 44% of their solar against 38%. More daytime use simply absorbs more of the midday output.

As production grows relative to use, nearly all of the extra goes to the grid. Homes whose panels made under 50% of their yearly use exported 32% of their solar and covered 21% of their use; homes making 130% or more exported 78% and covered 42%. Without storage, coverage levels off at around 40% because panels can serve only the hours when the sun is up, and those hours are already covered.

So a system sized to 100% of yearly use does not run the house on solar; it balances the year on paper, with the grid as storage. How much that is worth depends on what each exported kilowatt-hour earns, which is where NEM 2.0, net billing and LADWP’s program differ. See NEM 3.0 export credits explained and LADWP net metering.

Spring is the export season. In April the typical home without a battery exported 73% of its solar: the sun is strong and the house needs little, 24.5 kWh a day. By August home use had risen to 37.7 kWh a day, most likely for cooling, and exports fell to 50%. In 82 of 117 homes the highest-use month was August, while production peaked in June or July in 91: the heaviest demand arrives after the longest days.

Winter is the weak side. In December the panels made a median 16.2 kWh a day against 40.9 in July, and own solar covered 28% of home use, the lowest of the year. More on the winter drop in how much solar production drops in winter.

Counted day by day, the typical home sent more to the grid than it took on 51% of days. But even on days when its panels made more than the house used, a home without a battery still bought a median 8.9 kWh from the grid, 56% of that day’s use: the surplus went out at midday, and the house bought power back after dark.

Days with no meaningful grid purchase at all (under half a kilowatt-hour) were essentially absent without a battery. With one they were still rare: the median battery home had 0.4% such days, and only the top tenth reached 16% or more. A typical home battery shifts the evening; it does not take a house off the grid.

In the 22 battery homes, the typical battery took 24% of the solar production and delivered 20% of the home’s electricity, about 2,441 kWh a year, and it was active on 85% of days. That is the main reason battery homes kept 58% of their solar against 41% without one. They also tend to be larger homes with larger systems: a median 41 kWh of use a day against 31.

The clearest single example is a SolarEdge home. In 2022 this home’s battery barely moved energy, and the home exported 71% of its solar while buying 71% of its electricity. From 2023 the battery delivered 3.9–5.0 MWh a year; exports fell to 22–39% and the grid share of home use to 35–39%. Same roof, same panels: the battery moved midday solar into the evening.

In six SolarEdge homes, the two LADWP batteries barely moved energy and behaved like outage backup, while the SCE batteries cycled daily. In the larger Enphase set, though, LADWP batteries worked as hard as SCE ones: 27% of production in, 22% of home use out, active on 84% of days, against 23%, 18% and 86% in SCE territory. So the utility alone does not decide it. The likelier difference is how each battery is set to run — for self-consumption, time-of-use savings or backup only — which is chosen home by home. How the modes work: how a home battery works.

Net billing has not changed how much a home without a battery exports, at least in our SCE homes. Those that began operating on or after April 15, 2023, when net billing replaced NEM 2.0 for new SCE applications, exported a median 61% of their solar (30 homes), against 63% for earlier systems (14 homes). The start date is only a proxy, since some later systems applied earlier and kept NEM 2.0. Under net billing each exported kilowatt-hour earns much less, and that export share is exactly what a battery is meant to shrink; see do I need a battery with NEM 3.0.

These six homes add a longer view: 26 complete home-years with the same energy balance. They span a similar range to the larger set, from 15% of solar used on site to 70%, and none of them came close to running without the grid: even the most self-sufficient bought 23% of its electricity.