Hacker News Reader

Comment by bsdetector | Hacker News Reader

This video is actually a great example of how smart people can fool themselves.

The sources he uses for breakeven time (DOE) assumes 24.3 mpg [1] for the average regular car, but people don't drive average cars they drive specific cars.

If you take the DOE assumptions and approach this from the other end, CO2 pollution by mpg, then you find an ICE car getting approx 30 mpg or more is less polluting per mile driven than an EV powered entirely by coal, and one getting >45 mpg is less polluting than an EV powered by natural gas. This does not include the battery at all, which makes sense as it is an independent consideration.

This is a more honest way to present the information so individuals can make informed decisions for themselves among potential car purchases and so that people aren't fooled into thinking that all EVs are less polluting in all circumstances.

1 - https://afdc.energy.gov/vehicles/electric_emissions_sources....

View on HN

How is doing a calculation based on assuming electricity sourced 100% from coal or NG help develop an informed decision? US electricity is roughly 40% carbon free, 40% natural gas and 20% coal - wouldn’t it be more honest to use this mix in your calculation? And that’s ignoring particulates and NOx.

Coal is used to establish a lower bound.

That makes little sense to me.

Your analysis would result in the conclusion that charging an EV in France (90% carbon free electricity) causes the same amount of CO2 emissions as using one in South Africa (90% fossil fuel electricity).

The expansion of EVs has coincided with a massive drop in coal electricity share and a massive increase in solar PV and wind share, globally. Assuming that expanding EVs are being powered by coal is odd to say the least. Especially given that (at least where I live) 70%+ of EV charging is done at home, mostly using off-peak electricity, which tends to have a much higher renewable share.

And maybe as the type of generation used for the marginal difference in energy required. If your country generates 90% of its electricity from hydroelectric and 10% from coal, you might still use coal for the comparison because hydroelectric power would be slow to add and the extra electricity for EVs would have to provided by coal

Assuming that it's more likely that marginal demand will be met by coal than by renewables makes no sense to me.

Coal is rapidly been retired as a source of electricity generation - 85% of generation capacity being retired in 2022 will be coal[1] - and there hasn't been a new large coal plant built in the US for over a decade[2]. In Europe the amount of electricity being generated by coal is half of what it was at its peak.

Meanwhile, renewables now make up about a 87% share of new electricity generation capacity in the the US[3]. Europe is similar.

Pricing the CO2 emissions cost of an EV using the dirtiest and fastest declining electricity source (coal) does not seem fair or honest to me.

[1] https://www.eia.gov/todayinenergy/detail.php?id=50838

[2] https://www.scientificamerican.com/article/will-the-u-s-ever...

[3] https://www.renewableenergyworld.com/solar/renewables-are-87...

Still... if you plan to add 100MW of renewable this year and retire 100MW of coal, then demand goes up by 100MW, do you now add 200MW of renewable instead, or do you delay the retiring of the 100MW of coal?

This is an important point. I live in Illinois which is over 60% carbonless mainly through aging nuclear investments. It would be naively optimistic to suggest that marginal demand will be only 40% fossil fuel since the nuclear will not be replaced.

There are many, many assumptions to such calculations. Here's PG&E's best guess at g CO2 / kWh based on time of day/year in California (page 2):

https://ww2.arb.ca.gov/sites/default/files/classic/fuels/lcf...

If you only charge from noon to two in Q1, then use have zero marginal emissions per mile driven (this is the best case). If you only charge from 7-8pm you emit 142.06 gCO2e/MJ.

1 megajoule = 0.2777777778 kilowatt-hour

Converting gives us ~512 g CO2e per kWh.

Car efficiency is measured in miles per kWh. 4 miles per kWh is doable (some cars are better, some worse). So, that's 128g CO2e per mile.

50 miles per gallon would be great. A gallon of gas emits 20 lbs CO2, or 0.4lbs per mile, which is 181 grams per mile.

So, if you only charge at the worst possible time in California (basically forcing PG&E to use natural gas plants to charge your car), then a 50 mpg car is still 50% worse than a slightly larger EV.

In practice, you charge during off peak hours. At night, emissions are about half worst case. During the day, they're 10% of worst case.

If you just apply CA's 49% natural gas (according to DOE) to the 45 mpg for 100% natural gas you come up with a gas-burning car needing >90 mpg to be cleaner -- well higher than the 70 mpg you calculated vs charging on natural gas -- but easily achievable with hybrids.

I feel like there are many people who see 30 coal / 45 nat gas and take offense that it doesn't make EVs sound particularly good. This is why I called it a more honest presentation because EVs in terms of CO2 are fairly bad in WV and marginal in DE or MI or many other places compared to a high-mpg regular car. Instead of taking offense and setting out to correct the record, understand that not sounding particularly good is an accurate view.

Even in CA, the mpg needed to be cleaner than EV formulation makes it clear that driving a 100 mpg hybrid for instance is so good that any further gains are of little consequence.

In the same vein as my sibling comment, the average is an average because not everyone drives a prius; tons of the US are dominated by trucks and SUVs. A more informed metric would be comparing EV vs ICE in cities based on those cities' percentage of electricity generation and average MPG based on the car population.

Reply on news.ycombinator.com