Does the solar actually charge the EV?

I drive an EV, an electric vehicle: a 2024 Tesla Model Y Long Range. I charge it almost entirely at home on my wall connector, which measures its own draw.

Over the 391 days from 15 October 2025 to 9 November 2026, the connector put 4,175 kWh into it: 19% of the house's use, 26% of what the panels made.
A day the charger wrote nothing counts as no charging, though no other series can confirm it: 9 of these 391 days, the longest run 4 days in a row.
Did the sun supply it? The short answer is no: only 20% of that charging happened while the array was generating.

This chart is an average day, hour by hour. The red bars are my EV charging in each hour; the blue line is what my panels made.
A red bar under the blue line means the EV charged while the sun was up. A bar alone at night drew from the battery or the grid.
Each curve has its own scale, so compare when they happen, not how tall. The slider picks a month; its far right end is the whole record.

Month shown

EV charging solar generation

The EV charges after dark

Each choice is an average over the days with data, so a short month compares with a long one.
A month the record covers only in part says so beside the slider.
Over the whole record the two curves barely overlap.
Charging peaks in the hour from 9 pm. By then the panels have been idle for hours.
75% of it happens between 6pm and 6am.
So the array does not fuel my EV in any literal sense.
Does it pay for it instead? The rest of this page answers that, and the answer is no.
Every credit is already spent on the rest of the house.

My EV is scheduled to precondition its cabin before I leave on a weekday.
While it is plugged in, the wall connector supplies that load. So the load appears in the charging record.
Over the 391 days of the record, the charger draws 1.05 kWh between four and eight in the morning on an average weekday.
On an average weekend day it draws 0.16 kWh in that window. The weekday draw is 7 times as much.
Evenings run the other way: 10.08 kWh a weekend day against 5.47 a weekday.
Monday (0.27) and Friday (0.21) sit nearer the weekend (0.15) than Tuesday (1.97).
That fits a departure schedule that leaves them out.

The coldest band with at least five weekdays has a dawn temperature of 0 to 14°F.
On those weekdays the morning window averaged 5.49 kWh. It drew more than 0.5 kWh on 100% of mornings.
In the warmest band, 75 to 89°F, it averaged 0.08 kWh, on 0% of mornings.
There is no matching rise at high temperatures. So the load is cabin and battery heating, not cooling.
Winter efficiency is another possible cause, which the data rules out: an electric vehicle does use more energy per mile in cold weather.
Total daily charging here averages 16.6 kWh a day from December to February, against 8.3 kWh from June to August.
So a fixed target for the state of charge would also give more morning charging on cold days.
But the 46 weekend mornings below 45°F hold 0.0 kWh between them, against 2.23 kWh a morning on weekdays as cold.
The temperature outside does not vary by day of the week.

generated house load net import

Billing is on a flat rate, so one kilowatt-hour is as good as another. The bill is struck on net import over the period.
The source of any one kilowatt-hour cannot be measured. The monthly chart below estimates it.
The question with a measured answer is what share of total use the array offsets.
The twelve months run from 10 November 2025 to 9 November 2026: 365 days with load and generation.
Over them the array generated 71% of household load.
Over the 338 of those days both grid counters recorded, the house was a net importer of 4,640 kWh.
My EV took 19% of household load on the same 365 days, or 4,035 kWh.
Take out my EV's 3,523 kWh on those 338 days, and the house was still a net importer of about 1,117 kWh.

The 12 bills from 21 October 2025 to 19 October 2026 came to $1,718. Of that, $401 is fixed daily charges: Basic Service and the Customer Benefit Contribution.
$1,299 is the charges per kWh, on the 5,195 kWh the bills charged: an average of $0.250 per kWh. Taxes and adjustments make up the rest.
Less use avoids neither fixed charge. I pay the solar loan separately; it is not in these bills or this rate.
At that rate my EV's 3,960 kWh over the same bill window costs $990.
The gross and net figures are the same, and that is the finding.
Net metering lets a kilowatt-hour exported in June pay for one imported in January. But the house imports more than it exports over the year.
So every credit is already spent on other use. None remains to offset my EV: each kilowatt-hour it draws is one more bought at the retail rate.
The meter's net import was 4,590 kWh on 337 of 364 days both grid counters recorded; with days missing, it does not compare directly with the bills.
Credits bank and carry across the window's edges. So bill over net import is not a price, and this page does not use it.

exported, earning credit imported, spending it

The utility's time-of-day rate has an overnight band, 10pm to 6am, that its own guide calls intended for electric vehicle charging. On its face it suits my house.
My house was on that time-of-use rate, Rate 195, until the billing cycle that started 20 February 2025.
Credits net within a band and not across bands, and my array generates only between sunrise and sunset. So noon exports offset noon use only.
Over the last 430 days it exported 24 kWh into the overnight band, against 6,403 kWh imported from it.
Under that rate, 350 kWh of credit would stay in bands with no use left to offset, while the overnight band was billed in cash.
Whether it would cost more or less in total is open. That turns on prices the rate guide does not publish.
The flat rate has a single band, so every exported kilowatt-hour offsets any imported one. I am billed on it for that reason.

This control zooms the charts that have a date axis. Totals are for the full record.

The solid part of each bar is an estimate of the share from solar. The pale part is the rest.
It is a proportional split, not a measurement.
In each 15-minute bin, my EV's kilowatt-hours are divided in the proportions the house drew from the panels, the batteries and the grid.
Battery energy counts as battery, even when the sun charged it.
From June to August the estimate credits 11% of charging to solar, and from December to February 4%.
In June the sun is still up at eight in the evening, and the batteries start the night with more in them.

Each mark is one session. It sits at the hour the session began, and its size shows how much went in.
A session ends when the charger has been idle for half an hour.