How Much Electricity Do Solar Homes Still Buy From the Grid? Measured Month by Month in 117 LA Homes
Solar homes keep buying electricity. Across 117 Los Angeles-area homes with full energy metering, the typical home bought 6,581 kWh from the grid over its latest 12 months — about 18 kWh a day and 61% of everything it used — while sending 5,453 kWh back. Only 53 of the 117 sent back more than they bought. Purchases follow the clock and the calendar: a home without a battery bought the least in April, about 10 kWh a day, and the most in August, about 19.
These are the same 117 homes, 22 of them with a battery, measured in how much solar a home actually uses. That page follows the solar; this one follows the electricity coming in from the grid: how much, in which months and at what point in the year the balance turns. Each home is measured over its latest 12 complete months, for most of them October 2025 through September 2026.
How much solar homes buy in a year
Median home in each group, over its latest 12 complete months.
| Group | Homes | Bought from the grid | Sent back | Bought, % of home use | Sent back minus bought | Homes that sent back more |
|---|---|---|---|---|---|---|
| All homes | 117 | 6,581 kWh | 5,453 kWh | 61% | −495 kWh | 53 of 117 |
| No battery | 95 | 6,534 kWh | 5,647 kWh | 62% | −567 kWh | 42 of 95 |
| With battery | 22 | 8,261 kWh | 4,740 kWh | 44% | +209 kWh | 11 of 22 |
| LADWP, no battery | 48 | 8,290 kWh | 5,543 kWh | 64% | −2,043 kWh | 16 of 48 |
| SCE, no battery | 44 | 5,421 kWh | 5,411 kWh | 61% | +346 kWh | 24 of 44 |
| Each column is a separate median, so “sent back minus bought” is the median of each home’s own difference, not the difference of the two columns. Middle half of homes bought 3,859–9,484 kWh. | ||||||
https://cali-energy.com/research/solar-home-grid-purchases-los-angeles. Aggregated data are free to use under CC BY 4.0 with a link to this page.The typical home ended its 12 months 495 kWh behind: a little more bought than sent back. The utility split is wider. LADWP homes without a battery used a median 38 kWh a day against 25 in SCE territory, and their panels made 87% of their yearly use against 106%. So the median LADWP home without a battery bought 2,043 kWh more than it sent back, while the median SCE home sent back 346 kWh more than it bought.
Battery homes bought a smaller share of their electricity, 44% against 62% without one, yet more kilowatt-hours, 8,261 a year: they are bigger households, at a median 41 kWh a day. What the battery itself moves is measured in how much solar a home actually uses.
A bigger system changes the year, not the night
Homes without a battery, grouped by yearly solar production as a share of yearly use.
| Solar production, % of yearly use | Homes | Bought a year | Sent back a year | Homes that sent back more |
|---|---|---|---|---|
| under 50% | 10 | 11,769 kWh | 1,145 kWh | 0 of 10 |
| 50–80% | 23 | 7,489 kWh | 4,280 kWh | 0 of 23 |
| 80–100% | 19 | 7,737 kWh | 5,786 kWh | 0 of 19 |
| 100–130% | 23 | 5,407 kWh | 6,525 kWh | 22 of 23 |
| 130% or more | 20 | 2,629 kWh | 7,195 kWh | 20 of 20 |
Whether a home ends the year ahead is decided by size. All but one of the 43 homes whose panels made more than they used sent back more than they bought; none of the 52 homes making less did. How much a home buys is decided by timing. Even the homes making 130% or more of their use bought a median 2,629 kWh a year, because the panels cannot serve the evening, the night or a short winter day. More panels add to the midday surplus; they do not reach the hours after sunset.
Month by month
Homes without a battery, every complete home-month in 2023–2025.
| Month | Bought a day | Sent back a day | Sent back minus bought, a day | Homes buying more than they sent back |
|---|---|---|---|---|
| January | 14.9 kWh | 11.2 kWh | −4.7 kWh | 66% |
| February | 13.2 kWh | 14.1 kWh | −0.4 kWh | 51% |
| March | 11.7 kWh | 17.9 kWh | +5.9 kWh | 26% |
| April | 10.0 kWh | 21.8 kWh | +11.2 kWh | 15% |
| May | 10.8 kWh | 21.8 kWh | +9.3 kWh | 19% |
| June | 12.1 kWh | 21.5 kWh | +7.5 kWh | 22% |
| July | 16.5 kWh | 20.1 kWh | +3.7 kWh | 40% |
| August | 19.3 kWh | 16.9 kWh | −3.1 kWh | 58% |
| September | 17.5 kWh | 14.1 kWh | −4.8 kWh | 63% |
| October | 13.6 kWh | 15.1 kWh | −0.1 kWh | 51% |
| November | 13.8 kWh | 10.8 kWh | −3.3 kWh | 67% |
| December | 16.0 kWh | 8.5 kWh | −8.4 kWh | 76% |
| Each column is a separate median across home-months; “sent back minus bought” is the median of each home-month’s own difference. | ||||
From March through July the typical home without a battery sent back more than it bought. From August through January it bought more, with October and February roughly even, and December was the deepest month: −8.4 kWh a day, with 76% of homes buying more than they sent back. April was the opposite: +11.2 kWh a day, with 15%.
Two different things drive the two buying seasons. In late summer the homes simply used more: a median 37.7 kWh a day in August against 24.5 in April, most likely for cooling, and in 82 of 117 homes August was the highest-use month, after production had already peaked in June or July. In winter use is ordinary but the sun is short: in December the panels made a median 16.2 kWh a day against 40.9 in July.
The late-summer peak is not a solar-home quirk. SCE’s own figures for all of its residential customers show the same shape: about 337 kWh per account in April and 702 in September on average in 2024–2025, the highest month in its reports. SCE reports billed kilowatt-hours by report month, and for its solar customers those are already net of their own solar.
What a sunny day still needs from the grid
| Season | Homes | Bought that day | Share of that day’s use |
|---|---|---|---|
| Summer (June–September) | 81 | 11.1 kWh | 50% |
| Winter (November–February) | 73 | 8.5 kWh | 63% |
| Median of each home’s median surplus day; homes with at least 10 surplus days in the season. | |||
Daily totals still show the night clearly. On a day when the panels made more than the house used, any electricity the home bought had to come in at hours when the panels were not keeping up: early morning, evening and night. On such summer days the typical home without a battery bought 11.1 kWh, 50% of that day’s use. On surplus winter days it bought less, 8.5 kWh, but a larger share, 63%, because winter evenings are longer. Panels make nothing after dark; see do solar panels work at night.
The running balance: SCE true-up and LADWP credits
How these kilowatt-hours turn into a bill depends on the utility. SCE bills monthly and settles the year with a true-up in the month the system started service. LADWP nets energy over each billing period, usually two months; a surplus becomes a credit that carries forward with no time limit but is never paid out. Details: SCE solar billing explained, the NEM true-up bill and LADWP net metering.
In kilowatt-hours, the typical SCE home without a battery came out slightly ahead over its 12 months, +346 kWh, and 24 of 44 sent back more than they bought. The typical LADWP home without a battery came out 2,043 kWh behind, and 16 of 48 came out ahead.
The month the year starts changes the path, not the total. Starting in March, at least half of the homes never went below zero: the spring surplus came first and carried them through late summer and winter. Starting in July, the same homes were a median 957 kWh behind at the low point before spring caught up. A running balance deep in the red in December is normal for a system whose year starts in summer; it says little about where the true-up will land.
These balances are energy, not money. SCE’s NEM 2.0 time-of-use plans net the bill in dollars at different prices by hour, and under net billing each exported kilowatt-hour is credited at an hourly export price that is usually far below the retail price, so a kilowatt-hour balance near zero can still leave a true-up bill. See NEM 3.0 export credits explained.
Why solar homes keep buying electricity
Night
Even on days the panels out-produced the house, the typical home bought 11 kWh in summer: early morning, evening and night.
Winter
December production fell to 16 kWh a day, and 76% of homes bought more than they sent back that month.
Late summer
Use peaks in August, after the longest days; purchases reached 19 kWh a day.
Size
Homes whose panels make less than they use buy the difference. See still have a bill after solar.
Methodology
- Homes
- All homes in our Enphase monitoring fleet with a consumption meter and 12 complete consecutive months: 117 homes, 22 with a battery. The data checks are described in the methodology of the self-consumption study.
- Bought and sent back
- Electricity bought from the grid and sent back to it, as measured by the home’s consumption meter at the service entrance. These are not utility billing data; we checked them for internal consistency, not against bills.
- Twelve months
- Each home’s latest 12 consecutive complete months, mostly October 2025 – September 2026. “Sent back more” means the 12-month export total exceeded the 12-month purchase total.
- Months
- Every complete home-month of homes without a battery in 2023–2025: 2,335 home-months. Each column is a separate median.
- Surplus days
- Days when the production meter recorded more than the home used. Purchases on those days are the minimum the home needed from the grid outside the solar hours; with daily data we cannot see the exact hours.
- Running balance
- Cumulative sent back minus bought, month by month, over each home’s same 12 months, replayed from each start month. It ignores tariffs, prices, fixed charges and the difference between NEM 2.0, net billing and LADWP’s rules.
- Utility area
- From California Energy Commission utility service-area maps, used only where the match is certain or highly likely.
- SCE context
- Average kWh per residential account by month, all SCE ZIP codes, 2024–2025, from SCE’s quarterly energy data reports published under CPUC Decision 14-05-016.
- Limits
- Not a random sample: these are systems we monitor, weighted toward the San Fernando Valley.
Sources: Cali Energy monitoring fleet (Enphase consumption and production meters); Southern California Edison, energy data reports.
Frequently asked questions
Do you still have an electric bill with solar panels?
Almost always. In the 117 LA homes we measured, homes bought electricity from the grid on almost every day of the year, including the 53 that sent back more than they bought over the year, and utility bills also carry charges that solar credits do not cover. Why the bill does not go to zero: still have a bill after solar.
How much electricity does a home with solar buy from the grid?
The typical home we measured bought 6,581 kWh a year, about 18 kWh a day and 61% of its use; the middle half of homes bought 3,859–9,484 kWh.
In which months do solar homes buy the most electricity?
In August and December. Homes without a battery bought a median 19 kWh a day in August and 16 in December, against 10 in April.
Do solar panels make electricity at night?
No. Even on days when the panels made more than the house used, a home without a battery still bought a median 11.1 kWh in summer and 8.5 kWh in winter for the hours without sun.
Why is my SCE balance negative in winter?
Usually because your 12-month year started in summer or fall, so late-summer and winter purchases come before the spring surplus. In our SCE homes, a year starting in July went a median 957 kWh below zero before recovering, while a year starting in March mostly stayed above it. The kilowatt-hour total at the end of the year is the same either way; the dollar result also depends on your rate plan.
Does a battery stop me buying from the grid?
It cuts it. Battery homes bought a median 44% of their electricity from the grid against 62% without one, but days with no purchase at all were rare.
More research
Next steps
Buying more from the grid than you expected?
Prepared by Cali Energy, October 15, 2026. Figures describe systems in our monitoring fleet and are not a guarantee of what any specific roof will produce; orientation, tilt, shade, equipment and weather change results. Not engineering, financial or utility advice. See our Content Disclaimer.