Electric Cars Pollute More Than Gas Cars - National Bureau of ER Says

[hiroshiy]

I've purchased and read the original paper, and it tries to impress public with lies, by carefully manipulating and pseudo-simplifying numbers and statistics. Why gas cars pollute differently in various cities? What kind of skewed parameters are used for EVs? They used MPG numbers for gas cars, which is ok, but used MPGe numbers for EVs, which makes EV numbers worse. There seems to be a lot of traps in there.

To counter this type of propaganda is not easy I think. I would create a very simple model (as opposed to this flawed, deceptively complex model) that everyone can easily understand, and show that EVs are cleaner even with 100% Coal generated electricity, even in winter.

 

[Alexander]

I agree on all fronts; I was just picking a nit about saying it takes more electricity​. When rebutting fallacies, it's always better to state the case accurately ourselves.

I agree, we have to fight fallacies with facts, otherwise we lose credibility. I didn't know the energy used was from steam, I thought it was actual consumption.

I've purchased and read the original paper, and it tries to impress public with lies, by carefully manipulating and pseudo-simplifying numbers and statistics. Why gas cars pollute differently in various cities? What kind of skewed parameters are used for EVs? They used MPG numbers for gas cars, which is ok, but used MPGe numbers for EVs, which makes EV numbers worse. There seems to be a lot of traps in there.

To counter this type of propaganda is not easy I think. I would create a very simple model (as opposed to this flawed, deceptively complex model) that everyone can easily understand, and show that EVs are cleaner even with 100% Coal generated electricity, even in winter.

I wonder if we could crowd fund our own study that debunks this. Personally, I'm tired of explaining it and would donate money just to have a study of my own to push back in their faces

 

[brucet999]

The refinement process alone already makes gasoline cars more polluting then EV's, and that's just the pollution generated from the refining process. You also have to include the pollution created from:

- The drilling of the oil
- The pumping of the oil
- The storage of the oil
- The transportation of the oil (in ships or trucks which also burn gasoline) - The pumping of the oil into the refineries
- The refining process itself (which we've already talked about)
- The transportation of the gasoline (in ships or trucks which also burn gasoline)
- The pumping of the gasoline into gas stations
- The pumping of the gas into your car
- Then finally the burring of the gas in your car

When you add up this entire process it comes out to 450 grams of CO2 per kilometer of driving (+/- 50 grams). Even if you charge an EV from the most polluting source of electricity (coal), its only 60 grams of CO2 per kilometer. Most of the time its much less then that. We use hydroelectric here in California so its 0 grams for us.

So if you're going to use the Well-to-Wheel's argument and account for the pollution that's created from the generation of electricity, then you also have account for the pollution that's created from the production of gasoline. When you do that EV's are the clear winner."

Your argument suffers from many errors, I'm sorry to say, that will dilute its strength.

- The drilling of the oil Drilling occurs at the beginning of production and a well may produce tens of thousands of barrels over perhaps 40 years after drilling.
- The pumping of the oil Sometimes powered by electricity, sometimes diesel, sometimes natural gas.
- The storage of the oil There is no energy consumed by storage tanks
- The transportation of the oil (in ships or trucks which also burn gasoline) Trucks burn diesel. Ships burn bunker C fuel Most crude oil is shipped by rail that is very efficient at about 480 ton miles per gallon of fuel or by even more efficient pipelines
- The pumping of the oil into the refineries Pumps powered by electricity, as clean as the electricity used by EVs.
- The refining process itself (which we've already talked about)
- The transportation of the gasoline (in ships or trucks which also burn gasoline) Once again diesel or bunker C, but mostly by highly efficient rail and pipeline or by river barges
- The pumping of the gasoline into gas stations Gasoline drops by gravity from tank trucks into underground tanks; no pumping involved.
- The pumping of the gas into your car By efficient electric pumps
- Then finally the burring of the gas in your car

When you add up this entire process it comes out to 450 grams of CO2 per kilometer of driving (+/- 50 grams). At 24 mpg, a gasoline car produces 254 g CO2 per kilometer. If your number were correct, there must be 0.8 units of fuel consumed per unit sold at the gas station. That cannot be correct.
Even if you charge an EV from the most polluting source of electricity (coal), its only 60 grams of CO2 per kilometer. Most of the time its much less then that. We use hydroelectric here in California so its 0 grams for us. California power is not all hydro. SoCal Edison gets 22% of its power from renewables, 4% from hydro, 21% natural gas, 7% coal, 7% nuclear, and 39% "other", meaning purchased from the Grid from unknown origins. SDG&E and PG&E sources are likely about the same. According to my calculations, a Model S 85 powered by SoCal Edison at 85% efficiency AC to DC conversion produces about 100 g CO2 per kilometer. That is still only about 40% of a typical ICE, however, without counting whatever emissions come from well to tank.

 

[Jeff N]

We've been over this before -- it take 4 to 7.5 kWhs of energy but most of that is process steam, which is effectively a by-product of the refining process. Use of electricity is not negligible but much less.

Refineries sometimes purchase or use steam derived from non-refining sources but it's a relatively small source of the energy consumed during the refining process. Most of the energy used comes from natural gas, leftover unmarketable fractions from earlier refining called "still gas", along with some grid electricity (about 0.2-0.3 kWh per gallon of gasoline refined).

Here's a useful and brief overview from the GREET folks in 2011. See Table 1 for the various energy sources consumed during refining. Table 3 breaks down the different input energies by percentage. Electricity was 5.2% (other sources have reported as low as 3.5%).

https://greet.es.anl.gov/files/petroleum

 

[Robert.Boston]

No one ever likes it when I say so, but I think that we should use the marginal fuel being used for electric generation in a region, not the average. If I make the decision whether to take the BMW or the Tesla, the extra power I'd use with the Tesla requires the grid operator to ramp up some incremental generation source. Nuclear, wind, solar, and our hydro here in New England (which is all run-of-river or has minimal pondage) are already running as much as they can, so incrementally it's going to come from biomass or natural gas. In other parts of the country, the marginal fuel is coal.

 

[dsm363]

No one ever likes it when I say so, but I think that we should use the marginal fuel being used for electric generation in a region, not the average. If I make the decision whether to take the BMW or the Tesla, the extra power I'd use with the Tesla requires the grid operator to ramp up some incremental generation source. Nuclear, wind, solar, and our hydro here in New England (which is all run-of-river or has minimal pondage) are already running as much as they can, so incrementally it's going to come from biomass or natural gas. In other parts of the country, the marginal fuel is coal.

Seems fair to me.

Is there actually a good study around that deals with the issue of EV pollution?

 

[RDoc]

No one ever likes it when I say so, but I think that we should use the marginal fuel being used for electric generation in a region, not the average. If I make the decision whether to take the BMW or the Tesla, the extra power I'd use with the Tesla requires the grid operator to ramp up some incremental generation source. Nuclear, wind, solar, and our hydro here in New England (which is all run-of-river or has minimal pondage) are already running as much as they can, so incrementally it's going to come from biomass or natural gas. In other parts of the country, the marginal fuel is coal.

But generally EV's are charged at night during off-peak hours so while it's true that if one person charges an EV once, that is incremental. However, if 1000 people charge their EV's every night on average, doesn't that just add to base load at the best time for generation?

 

[jerry33]

But generally EV's are charged at night during off-peak hours so while it's true that if one person charges an EV once, that is incremental. However, if 1000 people charge their EV's every night on average, doesn't that just add to base load at the best time for generation?

Ideally, electrical generation should be 100% base load. Otherwise they have to dump power, store it (and live with the associated losses), or use more peakers. EVs allow greater base load which reduces costs for everyone.

 

[Robert.Boston]

Ideally, electrical generation should be 100% base load. Otherwise they have to dump power, store it (and live with the associated losses), or use more peakers. EVs allow greater base load which reduces costs for everyone.

Err, no. All baseload and mid-merit fossil-fueled plants have a significant operating range, usually at least half of the unit's nameplate. So, for example, a 500 MW natural-gas combined-cycled unit may generate anywhere from 250 MW to 500 MW, as required to meet system demands. It's not "dumping" power when it operates below 500 MW; it's using less gas and generating less power. There's not a big penalty for these plants operating below maximum, that is, the input/output ratio (MMBtu/MWh) doesn't change a great deal within the normal operating range.

But generally EV's are charged at night during off-peak hours so while it's true that if one person charges an EV once, that is incremental. However, if 1000 people charge their EV's every night on average, doesn't that just add to base load at the best time for generation?

Not unless the extra load changes the fundamental unit commitment (which generating units are turned on). Regardless, I'd still prefer to use the marginal numbers: they are the worst case and yet will still show big EV benefits.

 

[GregRF]

No one ever likes it when I say so, but I think that we should use the marginal fuel being used for electric generation in a region, not the average. If I make the decision whether to take the BMW or the Tesla, the extra power I'd use with the Tesla requires the grid operator to ramp up some incremental generation source. Nuclear, wind, solar, and our hydro here in New England (which is all run-of-river or has minimal pondage) are already running as much as they can, so incrementally it's going to come from biomass or natural gas. In other parts of the country, the marginal fuel is coal.

That study does use "marginal" sources. But as I read it, they tried to average the outcome as each plant across the US adding that marginal amount, not specifically the plants that the distributor would necessarily choose. The electrical values they used were also about 5 years old.

The other thread has a link to the full paper as well as some comments from one of the contributing authors.

 

[AudubonB]

This entire thread should be merged with the one specific to the NBER article, which appeared here right around 1 July. One of the study's authors (from one of the two NCarolina schools) appeared in the thread to discuss it...although I recall he did not respond to my criticism, which is that their treatment of point-source, ex-urban pollution and that of street-level, inter-urban pollution was identical, which it absolutely is not from any cost analysis.

 

[dhrivnak]

I have done an analysis using our local power which unfortunately is 80+% coal right now. Dropbox - GAS_VS_Electric.xlsx
Considering just the burning and refining the gasoline if you compare a Leaf versus a Malibu the Leaf is 50% less even on coal based power. Here a Prius, and ONLY a Prius at 50 mpg will beat the Leaf. The break even is 44 mpg and there is exactly one gas car that the EPA rates over 40 mpg. So I think it is safe to say an EV is better than gasoline in most all cases if you add in the EPA analysis for refining found at http://www.epa.gov/otaq/climate/documents/420f11041.pdf

Then when you add it that many, like us, have added rooftop solar Driving on Sunshine and one can drive emission free there is no contest.

 

[Robert.Boston]

This entire thread should be merged with the one specific to the NBER article, which appeared here right around 1 July. One of the study's authors (from one of the two NCarolina schools) appeared in the thread to discuss it...although I recall he did not respond to my criticism, which is that their treatment of point-source, ex-urban pollution and that of street-level, inter-urban pollution was identical, which it absolutely is not from any cost analysis.

That one was over in the Model S thread, I believe. I'll see if we can get a merge done.

 

[Jeff N]

I have done an analysis using our local power which unfortunately is 80+% coal right now. Dropbox - GAS_VS_Electric.xlsx

Those numbers seem odd. How can 80% coal generation result in 1.3 pounds of CO2 per kWh? For one thing, the 134 million tons of CO2 emissions from the annual report are metric tons so the conversion to pounds should be 2,204.6 pounds per ton rather than 2,000. Even with that adjustment the CO2 per kWh still seems oddly low.

The URL to the utility report listed in the spreadsheet no longer works. It should be:
http://www.aepsustainability.com/docs/2011_AEP_CAReport.pdf

 

[stopcrazypp]

Those numbers seem odd. How can 80% coal generation result in 1.3 pounds of CO2 per kWh? For one thing, the 134 million tons of CO2 emissions from the annual report are metric tons so the conversion to pounds should be 2,204.6 pounds per ton rather than 2,000. Even with that adjustment the CO2 per kWh still seems oddly low.

The URL to the utility report listed in the spreadsheet no longer works. It should be:
http://www.aepsustainability.com/docs/2011_AEP_CAReport.pdf

EPA says 2.07 lbs CO2 per kWh in best case, so 80% coal would be 1.656 lbs CO2 per kWh assuming the other 20% has no pollution.
http://www.eia.gov/tools/faqs/faq.cfm?id=74&t=11

- - - Updated - - -

Not unless the extra load changes the fundamental unit commitment (which generating units are turned on). Regardless, I'd still prefer to use the marginal numbers: they are the worst case and yet will still show big EV benefits.

Marginal numbers usually show EVs are worse than gasoline powered cars and are used by naysayers because it represents the worse case scenario (the other favorite is to use 100% coal, which it is argued that marginal load sometimes produces).

A lot of arguments over marginal load were already made here, so I don't want to rehash.
http://www.teslamotorsclub.com/showthread.php/8067-Marginal-power

Long story short, I don't consider assigning EVs to marginal loads in comparison to gasoline cars is a valid comparison. I laid out a few points why: we don't assign new gasoline cars to "marginal" gasoline when comparing, there is no guarantee that an EVSE will always be the "marginal load" during a given day, and grid operators can and will factor in the growth of plug-ins into their grid design (such that it can be assigned to "baseload" no matter how you want to define marginal).

 

[3mp_kwh]

- The drilling of the oil Drilling occurs at the beginning of production and a well may produce tens of thousands of barrels over perhaps 40 years after drilling.
- The pumping of the oil Sometimes powered by electricity, sometimes diesel, sometimes natural gas.
- The storage of the oil There is no energy consumed by storage tanks
- The transportation of the oil (in ships or trucks which also burn gasoline) Trucks burn diesel. Ships burn bunker C fuel Most crude oil is shipped by rail that is very efficient at about 480 ton miles per gallon of fuel or by even more efficient pipelines
- The pumping of the oil into the refineries Pumps powered by electricity, as clean as the electricity used by EVs.
- The refining process itself (which we've already talked about)
- The transportation of the gasoline (in ships or trucks which also burn gasoline) Once again diesel or bunker C, but mostly by highly efficient rail and pipeline or by river barges
- The pumping of the gasoline into gas stations Gasoline drops by gravity from tank trucks into underground tanks; no pumping involved.
- The pumping of the gas into your car By efficient electric pumps
- Then finally the burring of the gas [Synth Diesel] in your car

That last part is the funniest one. We hack the hydrogen, vs. BEV, argument down so many different ways. Here's a popular hydrogen flow chart, for which the addition of a third branch could show Audi's re-introduction of CO2 into the atmosphere:

Why a hydrogen economy doesn't make sense

As the article dates to 2006, Toyota could get credit for better transport efficiency, than just 23kwh, of 100kwh. Still, they are a long way from yielding 69kwh, of 100kwh....CLEANLY. Lastly, natural gas is used in liquefaction of U.S. hydrogen, not electricity.

...enough barrel fishing.

 

[3mp_kwh]

EPA says 2.07 lbs CO2 per kWh in best case, so 80% coal would be 1.656 lbs CO2 per kWh assuming the other 20% has no pollution.
http://www.eia.gov/tools/faqs/faq.cfm?id=74&t=11

EPA does not say "2.07 lbs CO2 per kwh in best case". Thanks for the link. They claim "average" case. Following utils, I see 1,600-2,400 lb/MWh values for (100%) coal, and recently heard many of the NG CCGT units going in are dropping from 7500, to 6500 heat rates. None of this jives w/EPA. The point is that there is big dispersion of emissions in a U.S. power fleet whose in-service dates begin as early as the 40's.

I bet U.S. coal actually is closer to about the 1.8lb/kwh level, today, since about ~40GW, of ~320GW, of the smaller, worst offenders closed because of mercury regs. It's the highest concentration of any electric source, but getting better.

 

[dhrivnak]

Thanks for finding and point out the broken link. It appears they moved it to a history folder. And yes a big mistake on the long ton/short ton. Many of the figures are in US tons but the CO2 is in metric tons. I have corrected the spreadsheet. But I still come up with 1.31 lbs-CO2/KWh.

In 2014 AEP with 73% of the power generated by coal, produced 122,700,000 metric Tons/CO2 for 207,211,000,000 KWh. That comes out to 1.31 lbs/KWh. But even with that number, which most say is low, I calculate that a Leaf is still cleaner than a Prius. Granted it is a modest 4% but that is still cleaner. And it is a whopping 50% cleaner than a Malibu which is a more typical car. And AEP is reducing their CO2 at a rate of 3%/year. Not bad for a company that does not want to move away from coal.