Gas vs. Electric: The Energy and Carbon Math, Made Visible

I pulled together some numbers for you. These are widely available, but I thought it could be interesting.

How much energy is actually in a gallon of gas?

The EPA's standard conversion is 33.7 kWh per gallon — the "gasoline gallon equivalent" used to compare a gas car's fuel economy against an EV's electric efficiency. On its own, that number doesn't mean much. Here's what it means in units you actually deal with:

33.7 kWh
the energy content, in electrical terms
1.2 daysof the average US home's electricity use
120 milesdriven in a 120-MPGe EV
8.9 kgCO₂ released when it's burned
22 hrsrunning a 1,500W space heater
Home-electricity comparison uses the U.S. average of roughly 10,500 kWh/yr (~29 kWh/day). The CO₂ figure is fixed — burning gasoline releases about 8.887 kg CO₂ per gallon no matter where you are, since that's simple chemistry, not geography.

That last point is worth sitting with: gasoline's carbon footprint doesn't change by location. A gallon burned in Seattle and a gallon burned in Phoenix release the same CO₂. Electricity is completely different — and that's the more interesting half of this post.

Electricity's carbon footprint depends on how it's made

A kilowatt-hour isn't a substance with a fixed carbon content — it's a unit of work that can come from burning coal, burning natural gas, splitting uranium atoms, or catching wind and sunlight, each with a wildly different emissions profile per kWh generated over its full lifecycle:

Lifecycle emissions per kWh generated, compiled from IPCC AR5 figures as presented by the World Nuclear Association. "Lifecycle" includes building and fueling the plant, not just what comes out of the smokestack — which is why even wind and solar aren't exactly zero.

This is the whole reason a grid's carbon intensity varies so much by region: it's really a question of generation mix. A grid leaning on coal and gas looks like the top of that chart; a grid leaning on hydro, nuclear, and renewables looks like the bottom. Nothing about the electrons themselves differs — what differs is what got burned, split, or captured to make them.

Pick your grid: what does this mean for an EV, where you live?

The calculator behind this site uses real EPA eGRID data for 14 US metro areas — each one's own blend of coal, gas, nuclear, hydro, and renewables, averaged into one number. Pick one below and see what it means per mile, next to a typical gas car:

Notice that even on the least-clean grid in that list, the EV still comes out ahead — because EVs convert stored energy into motion so much more efficiently than an internal combustion engine that they start with a real head start. What changes by grid isn't whether the EV wins, it's by how much — anywhere from several times cleaner per mile on a clean grid, to a much narrower margin on a coal-and-gas-heavy one. With a genuinely inefficient EV on a genuinely coal-dependent grid — outside the range of anything in this table — the gap can close further still. "Electric" isn't a single answer; it's a question that depends on the grid underneath it.

The dynamics behind the differences

Why do regions differ so much? Mostly geology and history. The Pacific Northwest sits on hydropower; California and the Southwest lean on natural gas plus a fast-growing share of solar; the Rocky Mountain region and the Midwest still run a lot of coal. None of that is a choice any individual household makes — it's decades of infrastructure and resource endowment, showing up in your outlet.

Why "average" isn't the whole story. The grid-carbon figures here are annual averages for a region — but the electricity supplying your home at any given moment comes from whichever plants are running right then, and that mix shifts throughout the day. Baseload plants (nuclear, some coal and hydro) run nearly constantly; "peaker" plants — often gas, sometimes the dirtiest units on the grid — switch on specifically to meet demand spikes, like a hot afternoon with every air conditioner running. Charge an EV at 6pm on a high-demand day and the marginal generator meeting that extra load may be dirtier than the annual average suggests; charge midday, often when solar output is highest, and it may be cleaner. This is the same logic behind this calculator's own time-of-use and solar-charging settings — when you draw power matters, not just how much.

Why this keeps changing. Grids are getting cleaner most years as coal retires and renewables and storage get built — meaning today's regional numbers are a snapshot, not a fixed fact. The broad pattern (electric beats gas per mile almost everywhere in the US already, and the gap keeps widening) is a moving target in one direction.

If you're modeling a solar-charged EV specifically — where the marginal electron comes from your own roof, not the grid at all — see the companion post on how this calculator values that, and the calculator itself to run your own numbers.