Over a full year, a typical 8 kW rooftop array can power everything in an average American house. On a mid-range US planning assumption of 1,400 kWh per installed kW, which is worth checking for your own location in PVWatts, that array makes roughly 11,200 kWh a year, and the average US residential customer bought 10,791 kWh in 2022 (EIA). The honest complication is timing. Panels produce between sunrise and sunset, and your dryer doesn't care what time it is.
TL;DR: An 8 kW array produces around 11,200 kWh a year, slightly more than the 10,791 kWh an average US home buys. Annually it covers the lot. Instantaneously it covers whatever is running while the sun is up, which is why air conditioning pairs beautifully with solar and evening cooking doesn't. Without a battery, a grid-tied system also shuts off during an outage.
I think the question people are really asking is not what solar can run, but what it can run at 7pm in January. Those are two different questions, and mixing them up is why so many homeowners feel misled after their first winter.
What Does a Year of Production Actually Cover?
US homes spend their electricity in fairly predictable proportions. The three biggest end uses in 2020 were air conditioning at 19 percent, space heating at 12 percent and water heating at 12 percent of residential site consumption (EIA), with lighting and refrigerators next.
Applying those shares to the 10,791 kWh average gives a rough picture of what each load needs from your roof:
| End use | Share of home electricity | Approximate annual kWh | Panels at 400 W |
|---|---|---|---|
| Air conditioning | 19% | 2,050 | 3 to 4 |
| Space heating | 12% | 1,295 | 2 to 3 |
| Water heating | 12% | 1,295 | 2 to 3 |
| Everything else | 57% | 6,150 | 10 to 11 |
Those kWh figures are the EIA shares applied to the EIA average, so treat them as a national sketch rather than your bill. A house in Louisiana bought 14,774 kWh in 2022 while one in Hawaii bought 6,178, and the same array covers wildly different fractions of each.
Why Timing Matters More Than Total Energy
Here's the part that trips people up. Solar output and household demand peak at different hours. Production runs from mid-morning to late afternoon and peaks near midday. Household demand peaks in the early evening, when everyone's home, the oven is on and the sun has gone.
When I compare quotes, the export rate is the first line I look for, ahead of panel model or inverter brand. Grid-tied systems handle the mismatch with net metering: surplus midday production flows out to the grid, and you draw back in the evening. Your meter does the bookkeeping. It works well where the utility credits exports fairly, and much less well where it doesn't. That single policy detail changes what your panels are worth more than any hardware choice.
Which loads line up naturally with production?
- Air conditioning. Runs hardest on hot sunny afternoons, the best match there is.
- Pool pumps. Fully schedulable, so run them at midday.
- EV charging. Excellent if the car is home during the day, poor if it charges at 11pm.
- Dishwashers and laundry. Schedulable with a delay timer, and most people never bother.
- Water heating. Can be shifted with a timer or a heat pump water heater's own scheduler.
Our Tesla charging analysis works through the EV case specifically, where an average commute needs several thousand kWh a year on its own.
What Solar Cannot Do Without a Battery
Two hard limits, and both surprise new owners.
The first is night. No sunlight, no production, so evening and overnight loads come from the grid unless you've stored something.
The second is outages. A standard grid-tied inverter must disconnect when the grid fails. That's a safety requirement, not a product flaw: an array feeding a dead line endangers anyone working on it. So a house with 30 panels and no battery sits dark in a blackout alongside its neighbours, which feels absurd until you've seen what backfeeding does.
Adding storage changes both. A battery moves midday surplus into the evening and, if specified for backup, keeps selected circuits alive during an outage. It won't power a whole house indefinitely, and sizing it well is the difference between useful and expensive, which our battery sizing guide covers properly.
So Can It Run the Whole House?
Annually, yes, on most roofs with decent exposure. Continuously and independently, no, not without significant storage. And at $20,000 to $25,000 installed for the typical 8 kW system, the sensible target for most homeowners is offsetting the annual bill rather than cutting the wire.
That bill keeps getting bigger, which helps the case. US residential electricity averaged 18.16 cents per kWh through the first half of 2026, up from 17.30 cents across 2025.
Summary
An 8 kW array makes roughly 11,200 kWh a year and the average US home buys 10,791 kWh, so on an annual basis solar can power the whole house. In real time it powers whatever runs in daylight, which suits air conditioning, pool pumps and daytime EV charging. Evening loads and blackout resilience need a battery. Size the array to your own consumption, not to the national average.