How much solar should you add for an EV? Kilowatts by car and city

The arithmetic: annual miles times the window-sticker rate

The car's yearly electricity use is the one number to get right, and it needs no estimate if you know your mileage. Every new electric vehicle sold in the United States carries an EPA rating in kilowatt-hours per 100 miles, printed on the window sticker and listed on fueleconomy.gov, the joint Department of Energy and EPA site. Multiply your annual miles by that rating and divide by 100. The result is kilowatt-hours per year, the same unit your electric bill and every production estimate use.

If you do not track your mileage, the national figure is a reasonable starting point. The Federal Highway Administration's Highway Statistics 2023, Table VM-1 (updated March 2025), puts the average distance driven per light-duty vehicle at 11,106 miles a year: 11,026 for short-wheelbase vehicles such as cars, and 11,360 for long-wheelbase vehicles, a class that takes in many pickups and larger SUVs. That works out to about 30 miles a day. Your odometer readings a year apart, or the service records from your current car, will beat the average for your own household every time.

Then divide the annual kilowatt-hours by your roof's expected yield in kilowatt-hours per kilowatt per year, which the free NREL PVWatts calculator gives for any US address. The answer is the array size, in kilowatts DC, that would produce as much electricity in a year as the car uses. Multiply by the system price per watt and you have a first cost line for the EV share of the system.

1. Annual miles: from your odometer, or 11,106 as the FHWA 2023 national average.
2. Rated consumption: kilowatt-hours per 100 miles for your exact model year and trim on fueleconomy.gov.
3. Annual EV load: miles times the rating, divided by 100, in kilowatt-hours a year.
4. Roof yield: PVWatts kilowatt-hours per kilowatt per year for your address, tilt and direction.
5. Array size: annual EV load divided by roof yield, in kilowatts DC.
6. Cost line: array size in watts times the price per watt on your quote, compared with the benchmark.

What five 2025 electric vehicles draw in a year

Ratings differ more than buyers expect, and the gap carries straight through to the array. The table uses the combined rating fueleconomy.gov publishes for each 2025 model, read on October 30, 2026, at the FHWA average of 11,106 miles. The most efficient car here uses a little more than half the electricity of the full-size pickup for the same distance, so a pickup owner needs nearly twice the panels for the same driving.

Trim matters as much as the badge. On the same fueleconomy.gov listing, the 2025 Model 3 Long Range RWD is rated at 25 kWh per 100 miles and the Model 3 Performance AWD at 31, a 24 percent difference inside one model name. Look up your exact version rather than borrowing a figure from a review or a sales sheet. The rated figure is also a test-cycle average; aggressive highway speeds, roof racks, towing and cold weather all push real use above it, which the edge cases further down take up.

The sizing worksheet: four cities, five cars

Roof yield is the second variable, and it moves the answer as much as the choice of car. PVWatts version 8, run on October 30, 2026 for a 1 kW roof-mounted array facing due south at a 20 degree tilt with the calculator's default 14 percent system losses, returned 1,755 kWh a year in Phoenix, 1,308 in Chicago, 1,302 in Boston and 1,088 in Seattle. Divide each car's annual load by those yields and you get the array that would match it over a year.

Read the dollar figures as a comparison line, not a price. The Department of Energy's Solar Photovoltaic System Cost Benchmarks put the modeled market price of an 8 kW residential system at $2.95 per watt DC in Q1 2025, in 2024 dollars, and that system size matters here. A standalone 2.5 kW array would not price at that line, because permitting, design, interconnection and the crew's trip are largely fixed costs and spread over far fewer watts. The realistic way to buy EV-sized solar is as extra kilowatts inside a whole-home design, where the added panels carry mostly hardware and labour. Ask the installer to price the system with and without the EV share, and divide the difference by the added watts; that marginal cost per watt is the number to set against your charging bill.

To turn kilowatts into panels, divide by the module rating on your quote. As an illustration only, with 400 W modules a 2.47 kW share is about six panels and a 4.09 kW share about ten or eleven. Round up to whole panels and check that the roof plane has room after fire setbacks, which the setbacks guide explains.

The panel-count shortcut

Buying panels for a car sized from a panel count rather than a kilowatt-hour figure. A rule of thumb such as six to ten panels can be right for a compact car in Arizona and badly short for a pickup in Seattle, and it says nothing about whether those panels will be credited at retail or at an export rate when the car charges at night.

Do not add charging losses twice

Some energy is lost between the wall and the battery. The Department of Energy's Fact of the Week #1045 (September 3, 2018), built on fueleconomy.gov data, put that charging loss at about 16 percent for the vehicles it studied. Sizing advice often tells buyers to add a margin for it on top of the rated figure, and for the EPA rating that is a mistake.

The federal test procedure already counts it. Under 40 CFR 600.116-12, paragraph (c)(8), the lab measures AC watt-hours throughout recharging and must position the measurement to account for any losses in the charging system. The kilowatt-hours per 100 miles on the sticker is therefore electricity drawn from the outlet, not energy stored in the battery. Use it as printed. Add a margin only if your number came from somewhere else, such as a car's trip computer, which typically reports energy taken out of the battery and so understates what the meter sees.

There is one honest reason to pad. The test charges on a specified setup, and your charger, cable run and climate are not the lab's. If your car or your charger reports energy delivered from the wall, a few months of those readings are better than any margin, and they tell you whether your real draw runs above the label.

When the car charges matters more than how many panels

An array sized to the car does not mean the car runs on sunshine. The Alternative Fuels Data Center says most EV owners do the majority of their charging at home, and that most drivers charge overnight. At night the panels produce nothing, so every overnight kilowatt-hour comes from the grid. The solar connection is an accounting one: the array's daytime surplus goes to the grid, and the utility's export rule decides how much of the overnight charging that surplus pays for.

Overnight Level 2: the common case

According to the AFDC, most residential Level 2 chargers run at up to 30 amps and deliver 7.2 kW, adding about 25 miles of range per hour on 240 volts. For an average 30-mile day that is a little over an hour of charging, usually set for the cheapest overnight hours on a time-of-use plan. Under full retail net metering the timing does not change the solar value, because each exported daytime kilowatt-hour offsets an imported night one at the same price. Under net billing or a time-of-use export schedule, the night import costs the retail rate and the day export earns the lower export rate, so the same array is worth less.

Daytime Level 1: slower, but closer to the roof

A 120 volt Level 1 outlet, which the AFDC models at 1.9 kW, adds about 5 miles of range per hour; eight hours replace about 40 miles for a mid-size EV. That is slow, but its draw sits closer to what a small array actually produces around midday. A car that stays home on weekdays, for a remote worker or a second household car, can soak up the array's output directly at that rate, which turns an export at the low rate into self-consumption worth the full retail price. The AFDC notes Level 1 needs an outlet on a dedicated branch circuit near the parking spot.

Why a big charger outruns a small array

An array can never deliver more than its inverter's AC rating, and on an EV-sized 2.5 kW share that is far below a 7.2 kW charger's draw. Even at noon on a clear day, a car charging at full Level 2 power pulls most of its energy from the grid while the panels supply part. Some chargers and inverters can follow the solar surplus and vary the charging rate, which helps a daytime-parked car use more of its own production. It is worth asking about if the car is home in the day and your export rate is low; it buys nothing for a car that leaves at 7 a.m.

Winter: the months the car and the roof disagree

Annual sizing hides a seasonal mismatch. The car needs roughly the same electricity every month, and more in cold weather; the roof does not. PVWatts puts a 1 kW south-facing array in Boston at 143 kWh in July and 59 kWh in December, and in Seattle at 143 kWh in July against 33 in December. The table follows a Model Y driven an even 268 kWh of charging a month against a 2.47 kW Boston array, the annual match from the worksheet.

Over the year the columns balance, which is what the worksheet sized for. Within the year, about 362 kWh of summer surplus has to carry about 362 kWh of winter shortfall. Under annual netting with rollover credits, that works on paper. Under monthly netting where credits lapse, or net billing where the summer surplus is paid at an export rate, the winter imports are billed at retail while the summer exports earn less. The net metering guide explains which rule you are on, and the true-up guide shows how an annual-netting account settles.

Cold weather widens the gap. fueleconomy.gov, citing a 2019 AAA test, says EV fuel economy can drop roughly 39 percent in mixed driving in cold weather, with about two-thirds of the extra energy going to heat the cabin, and that even with the heater off it is 8 percent lower at 20 degrees F than at 75 degrees F. A northern household that drives the same miles all year can see its December charging load rise just as the array's output reaches its lowest point. Pre-heating the cabin while the car is still plugged in moves some of that heating to the grid connection rather than the battery, but it is still household electricity.

What the export rule does to the value of EV-sized solar

Every panel on this page produces the same kilowatt-hours whatever the tariff; what changes is what each one is worth. Under full retail net metering, the array in the worksheet offsets the car's annual charging one for one, at the retail price. Under net billing, most of the array's output is exported in the middle of the day while the car charges at night, so most of it is credited at the export rate, and the car's overnight charging is billed at the retail or off-peak rate. The value of EV-sized solar can fall by half or more between those two rules, as the worked example below shows with labelled rates.

Two moves recover value under a low export rate, and neither is more panels. The first is daytime charging where the car is home, which turns exports into self-consumption. The second is a time-of-use plan whose off-peak window makes overnight charging cheap anyway, which lowers the bill but also lowers what solar saves on that load. A home battery can shift daytime solar into the night, but at the cost per delivered kilowatt-hour the battery decision guide works through, an EV-sized battery cycle rarely pays on its own; the car's own battery is already doing the storing.

Rate design itself sits outside this site. For what a time-of-use schedule looks like and how off-peak charging windows are set, EnergyBS has a time-of-use rate strategy guide. Bring your tariff's actual export rate and charging window back to the worksheet.

Wiring, panel capacity and the charger itself

Adding a Level 2 charger and solar at the same time puts two demands on the main electrical panel. The charger is a large continuous load; the inverter is a new power source feeding the same busbar. A panel that has room for one may not have room for both, and the fixes for each pull in opposite directions.

• The busbar test for solar. The main panel upgrade guide works the 120 percent rule for common panels. One way to pass it, derating the main breaker, reduces what the house can draw, which is the wrong direction for a home adding a 30 amp charger circuit.
• The load calculation for the charger. The AFDC notes that a qualified electrician can add circuits to accommodate Level 2 charging, that installations must follow local and state codes, and that permits may be required. Ask for the load calculation in writing, with the charger and the solar both on it.
• One design, one permit set where possible. Doing both at once lets the electrician plan panel space, conduit runs and any service upgrade once, instead of reopening the panel a year later.
• Equipment cost is separate from the array. The AFDC cites residential Level 1 costs from $0 to $900 and, from a Rocky Mountain Institute report, Level 2 equipment from $380 to $690, before installation. Keep the charger and its circuit on their own quote lines so the solar cost per watt stays comparable.
• The federal charger credit has ended for new installs. The IRS page for the alternative fuel vehicle refueling property credit (section 30C) covers property placed in service from January 1, 2023, to June 30, 2026, at 30 percent up to $1,000 per item for a main home in an eligible census tract. A charger installed now falls outside that window, and no credit enters any figure on this page.

Where the simple answer breaks

The worksheet assumes one car, steady mileage, a south-facing roof and a utility that credits exports. Each assumption fails for some households, and the failures are predictable.

• Two EVs, or a pickup that tows. Add each car's load separately. Towing raises consumption well above the rated figure; the rating does not include a trailer, so use your own charging records once you have them.
• High mileage or long commutes. A 20,000 mile driver of the Equinox EV needs 6,200 kWh, about 4.8 kW in Boston. At that point the EV share is the size of a small whole-home system, and roof space, not budget, often sets the limit.
• East or west roof planes. The PVWatts yields above are for due south. Rerun PVWatts with your actual direction and tilt; the production estimate guide shows what changes when the plane turns.
• Mostly public or workplace charging. Kilowatt-hours charged away from home never touch your meter. Size only for the share you charge at home.
• Shade on the added panels. EV panels often go on the last roof plane left, which is frequently the worst one. A shaded plane can need noticeably more kilowatts for the same energy; the shade guide prices that loss.
• A grandfathered export rate. If you already have solar on an older net metering rule, adding panels for the car can move the account onto current rules in some tariffs. Check before expanding; the grandfathering guide lists the triggers.
• A car you may replace soon. Panels last decades; a car does not. A future model with a different rating, or a household that moves to two cars or none, changes the load. Size on the household's likely charging over the array's life, not on one vehicle's sticker.