I Changed My Mind About Valuing EV Charging in Solar Calculations
For most of this calculator's life, if you told it you were charging an EV on solar, it valued every one of those kWh at the price of the gasoline they replaced — not the price of the electricity. A reader sent me a critique of that choice, and after working through it carefully, I think they were right, and I was wrong. This is the story of what changed, and why, because I'd rather show my work than quietly ship a different number.
What I used to do, and why it felt defensible
The old logic was simple: size the EV's yearly charging need from miles and efficiency, price it at whatever gasoline it displaced — usually $0.50–0.70 per kWh, well above a normal electricity rate — and give the car first claim on whatever solar the system self-consumed. My reasoning, which I wrote into this calculator's own guide, was that this was the honest way to represent solar's "highest-value use," and that we're in a genuine transition period where both fuels are live options, so putting a real dollar figure on the swap seemed like useful information, not a trick.
It wasn't a trick. But it was an accounting error, and it took a critique with a sharper vocabulary than mine to show me exactly where.
The reframe that did it: two assets, one interaction term
The critique's central move was to insist on treating solar and an EV as two separate things being evaluated, with at most one legitimate line connecting them. Solar is 25-year infrastructure — you buy a machine that makes kWh, and its return is the value of those kWh over time. An EV is a car whose fuel happens to be cheaper than gasoline; that saving is real, but it's the car's saving, earned by the decision to buy an EV instead of a gas car, not by the decision to buy solar. If you charge that EV from a wall outlet with no panels on your roof at all, you still save the gas-vs-electricity spread. Solar didn't cause that saving. Crediting it to solar's own return was, quietly, counting the same dollar as a reason to buy the car and a reason to buy the panels.
Put as a cash-flow identity: switching to an EV zeros out gasoline and adds charging kWh to your electricity use, whether or not you have solar. Adding solar reduces the electricity you buy from the grid, whether or not you have an EV. The two decisions only create extra, joint value in the narrow wedge where solar specifically lets you avoid paying more for that charging than you otherwise would have — not in the full price of the gasoline neither asset alone was ever going to touch.
The number that made it concrete
I didn't just accept the argument — I ran it against this calculator's own seeded example, the one most first-time visitors see: a 10-panel, 6,423 kWh/yr system, $26,806 installed, a household using 7,000 kWh/yr, and a new EV added on top at 15,000 miles/yr. Under the old model, that scenario's car got credit for 4,213 kWh/yr of solar — its entire annual need — priced at gasoline, worth $2,946/yr, on top of $566/yr of household savings. Total: $3,558/yr, a 7.5-year payback.
Here's what stopped me: that system's physical self-consumption ceiling, given its size and this household's own load shape, is about 5,665 kWh/yr — below the house's own 7,000 kWh/yr demand, before the car is even in the picture. The car was never unlocking new self-consumption. The old model was relabeling the same 5,665 kWh of ordinary household electricity as "car electricity," because the car paid a higher price in the formula. Once I saw that, I couldn't unsee it. It wasn't a rounding error or an edge case — it was the mechanism working exactly as I'd built it, on the calculator's own flagship example.
What changed, mechanically
The fix follows straight from the reframe. Solar's own return — its payback, its equivalent return, the 10-year verdict — now reflects electricity only, never gasoline, in either direction. A kWh that charges an EV is priced exactly like a kWh that runs a refrigerator: at the rate you'd otherwise have paid the grid for it.
The harder part was figuring out how much solar the car should honestly be credited with at all. The old model gave it first claim on self-consumption, full stop. The new model asks a stricter question: how much more self-consumption exists because the car is there, compared to a baseline where the house's own load is served the same way but the car's charging (and its ability to shift into solar hours) is removed? Only that increment counts as the car's — and as the flagship example showed, that increment is often small, sometimes zero, and grows mainly when you can actually shift charging into the hours the panels are producing.
Gasoline never disappears from the picture — it just moves. What an EV saves you versus a gas car, fuel-cost to fuel-cost, is still calculated and still shown, as its own clearly labeled figure. It's just the car's number now, not solar's.
Why this plays out differently depending on your tariff
One thing the critique pushed me to think through more carefully than I had: how much this actually matters depends heavily on how your utility credits exported solar — which is also exactly why this calculator defaults to California rules, since that's the majority of quotes it's built around.
Under a set export rate — net billing, NEM 3, the default here — self-consuming a kWh instead of exporting it is worth a lot, because the export rate is so much lower than what you'd otherwise pay to buy that electricity. Timing an EV's charging to solar hours genuinely helps in this environment, and the calculator will credit it when you say you can do that. Under traditional, full-retail net metering, exporting and self-consuming pay about the same, so there's very little financial reason to prefer one over the other — the old model would have implied a large benefit here that mostly wasn't real. On a system with no export credit at all, anything the car can absorb that would otherwise be curtailed and wasted is close to pure gain — which makes EV timing matter the most, not the least, on exactly the systems where it's easiest to overlook.
What I didn't give up: bundling, on purpose
The one thing I was not willing to lose was a legitimate way to plan solar and a new EV together. If you're sizing a bigger system specifically because a car is coming, and you want the fuel savings counted toward that decision, that's not the mistake this whole post is about — it's a real, common, entirely defensible way to think about the purchase, as long as it's labeled honestly. So the calculator now shows a second, explicitly separate figure: solar's own return, plus the EV's fuel-switching saving, clearly marked as a household bundle rather than solar's own payback. On that same flagship example, the bundled figure comes out to $3,558/yr — almost exactly what the old, mislabeled number used to show. Nothing about that total was numerically wrong. It just needed the right name, and it needed to stop being the number everything else — payback, equivalent return, the verdict — was quietly built on.
That bundle is only offered if you're planning ahead for a car you don't own yet. If you already own and charge the EV, that fuel decision already happened; there's nothing left to bundle it with, and the calculator says so.
Play with it: what's switching from gas to electric actually worth?
Set this aside from solar entirely for a minute — no panels, no roof, just the fuel question this whole post kept insisting was a separate decision. Drag the efficiency slider or tap a real car below, and watch what a gallon of gas's worth of energy actually buys you in each one.
Notice what this widget does not include: no solar, no panels, no roof at all. That's the point — this is the car's own saving, the one this whole post argued shouldn't be counted as solar's return. If you're pairing this with solar, see the companion post on the energy and carbon side of the same comparison, or run your own numbers on the calculator itself.