Comment by cgearhart | Hacker News Reader

Comment by cgearhart | Hacker News Reader

Parent Full thread

Napkin math supports your concerns.

- The US used 134.51 billion gallons of gas in 2013.[1] - There are ~33 kWh/gal in gasoline [2]

134.51e9 gal x 33 kWh / gal = 4.44e12 kWh (4.44 trillion kWh / year)

- The US used a total of 3.856 trillion kWh of electricity in 2011 [3]

Ignoring improved efficiency and excluding other factors, converting all gasoline usage to electric would more than double US electricity consumption.

Under reasonable assumptions, PV solar generates about 0.75 kWh/m^2 per day. [4]

4.44 trillion kWh/year / 0.75 kWh/m^2 per day x 365 days/year / 1e6 m^2 per km^2 = 16,219 km^2

Which means you could generate all of the required electricity with about 16,200 km^2 of solar panels - which is just about exactly the total land area of Hawaii [5], or 405 million average home PV systems.

It's doable, but it will be non-trivial.

[1] http://www.eia.gov/tools/faqs/faq.cfm?id=23&t=10 [2] http://en.m.wikipedia.org/wiki/Gasoline_gallon_equivalent [3] http://www.eia.gov/energyexplained/index.cfm?page=electricit... [4] http://physics.ucsd.edu/do-the-math/2011/09/dont-be-a-pv-eff... [5] http://en.m.wikipedia.org/wiki/List_of_U.S._states_and_terri...

Ignoring improved efficiency and excluding other factors, converting all gasoline usage to electric would more than double US electricity consumption.

Not a bad analysis, but there's a lot we're glossing over here.

For one thing, that 33kWh can get you about 100 miles in an average electric vehicle. That's compared to the average new US car which gets 25MPG, and existing cars out there average worse than that. So right away, there's a factor of four that we're leaving out here.

Further (as others have mentioned) it takes a huge amount of electricity to refine oil into gasoline. The electricity savings from not refining a gallon of gas are enough to drive an electric vehicle pretty far on its own.

Also, our grid and generating capacity are built to handle peak load, which (in most parts of the US) will occur on hot days when lots of people are using AC. On the flip side, the majority of electric car charging happens at night, so there's a lot of spare capacity already there.

If everyone switched to EVs, we'd probably need to build out more generating capacity, but probably not as much as your analysis suggests.

"Further (as others have mentioned) it takes a huge amount of electricity to refine oil into gasoline. The electricity savings from not refining a gallon of gas are enough to drive an electric vehicle pretty far on its own."

This isn't as much as you would think. The manufacturing electrical usage of the US (which includes this) is separated out on the EIA site, and you can simply subtract it :)

One fact everyone seems to forget is that it takes approximately 6.5 Kwh of energy to refine one gal of gas. That is JUST refinement and not the energy to pump, store, etc.

My Volt drives around 22 -> 28 miles on that power.

I was going to raise the same point. And, considering that the average new car sold in the US gets 25MPG, we should be just about breaking even.

Breaking even eventually. I doubt they're going to suddenly just stop refining all of that oil. There will have to be a trend before the oil companies start lowering their output.

A given car will either use refined gas or electricity. The point is those both use a similar amount of electricity, either way the impact on the grid is similar. Oil companies are not going to refine more gas than is used. They might refine more than is used on a particular day or even month, but they are not going to outpace demand on any long time scale.

Under those conditions, converting all vehicles to electric would increase net electricity usage by ~20%. (Better than 100%!)

Eliminating gasoline refining would reduce the needed electricity by ~850 billion kWh / year, and the improved efficiency (ignoring charge/discharge cycle efficiency) would reduce the net energy to ~594 billion kWh / year - or 54 million average home PV installations. Quick googling shows that it will be about 10% higher for charging losses.

Keep in mind we'll still be refining oil / petroleum for other uses (lubricants, ink, plastic, fertilizer, diesel, etc). If this is independent of the refining into gas, then all's well with your math.

Do cars typically run on plastic and fertilizer where you're from? If it's true that a Volt can drive 25ish miles on the electricity used to refine a gallon of gas, then substituting electric miles for gas ones is a wash in terms of net electricity usage. Or are you trying to say something more subtle that I'm just missing?

If this is independent of the refining into gas, then all's well with your math.

I think we agree ... that if there's less gas being refined, then there's less energy being used to do that ... if and only if gas refining is independent of refining into all the other products. If, on the other hand, refining into the other products efficiently ends up making gas anyway, then that's different.

This matches the numbers i came up with from an argument on reddit[1].

I subtracted out the electricity spent producing gas in the US, which isn't as much as one would think.

You can entirely ignore it if you want to compare the residential electric grid.

Note that if you stick to residential, it's 6x all current residential usage if everyone drives teslas and charges once a day.

[1] http://www.reddit.com/r/technology/comments/236uzj/toyota_co...

>Ignoring improved efficiency and excluding other factors,

But these factors are absolutely enormous, particularly for the comparison you set up, as electric vehicles are about 3x as efficient. EVs on the mass market get about 3m/kWh, whereas most gas vehicles get about 1m/kWh, if you use your 33kWh/gal.

So instead of doubling US electricity consumption, 100% EV vehicle fleet just raises consumption by 33%, an amount that could nearly be made up for with simple efficiency gains in electricity use.

That's one of the down sides to napkin math. I posted a follow up on another comment. Including some of the efficiencies leads to a net increase of about 20% in electricity usage.

If I read this right this is assuming that all current gasoline usage is converted into electrical usage. It's not like that would happen over night though, or possibly ever. Also, can PV help at night when a majority of charging should occur?

You are correct - it assumes total conversion, but the numbers don't change if you project some ramp for the transition.

A more reasonable assumption for charging is that it will take place wherever electric vehicles are parked, regardless of time of day; in that case, large parking lot installations for PV may be more sensible than individual home systems. In either case, it seems likely that increased adoption of electric vehicles will change what we consider "peak" electricity use hours, since it could consume half or more of total electricity in the US.

Reply on news.ycombinator.com