How not to address the EV emission question

*UPDATE 8/22/22 - I published a new article where I basically dismantle my own post below. And the majority of my statements below I now believe to be incorrect.

Original Post:

I recently came across this post from Kelly Senecal about "zero emission vehicles" and how zero emissions vehicles actually produce a non-zero amount of emissions, which we all know; however, what got me was this quote: "All cars produce emissions from driving. And often the so-called zero emission vehicles produce more of them."

So I commented: "Can you share an example of a “so called “ zero emission vehicle producing more emissions than a fossil fueled one? I’m very curious to see this"

and Peter replied: "Please see our paper: (He linked to SAE hosted version but I found this link worked better)"

I was worried it wouldn't have numbers and rigorous calculation, and I was wrong...but...that doesn't mean there aren't a lot of problems IMO. The main problem is what is stated in the Abstract: "We find that currently there is no evidence to support the idea that BEVs lead to a uniform reduction in vehicle emission rates in comparison to HEVs and in many scenarios have higher GHG emissions." But the issue I found is I'm not sure that the paper proves that there is "no evidence".

First: I'm a mechanical engineer. I graduated with high honors from one of the best mechanical engineering schools in the world ( Georgia Institute of Technology ). I've spent about 15 years analyzing papers on emissions and mechanical and electrical systems. I say all this to demonstrate that I'm not a dummy on this topic...but I had trouble deciphering this paper. It's overly complicated and I can't tell if it's intentional (ie: to obfuscate) or if it's just poor communication.

An example: The paper highlights the important metric of CO2eq/kWh and says: "This yields a weighted average for the average mix of 0.41 kg CO2eq/ kWh. This value, or one close to it, is typically used for U.S.- based LCA calculations when considering the GHG contribution of electric vehicles (EVs)". But then a few paragraphs laters says: "It should be noted that this paper used a 2008 U.S. average grid intensity of 0.67 kg CO2/kWh, which is significantly higher than the 2020 value."? So why use a value of .67 if you previously say that .41 is the value? A higher grid value results in worse performance of BEV emissions which would help a paper that is trying to show that BEVs are dirtier than they are?

Second: A lot of the paper's argument hinges on average vs. marginal emission rates. It states: "The argument for all-electric vehicle fleets is that the GHGs produced in the production, operation, and end-of-life phases by these vehicles are less than the corresponding GHG emissions from other powertrain technologies. This argument is more compelling when the GHG contributions from the electricity production sector are calculated with average mix emission rates that are inconsistent with the U.S. electricity supply process and associated grid data. In this work, we apply physically representative modeling to reveal the source of the error associated with the average mix emission rate and develop a methodology for calculating the more representative marginal emission rates using available grid data."

The paper does a good job of explaining marginal mix: "Another electricity mix to consider is the marginal mix. In the marginal mix, the sources of electricity that are brought online as demand increases (or taken offline as demand decreases) are considered to determine the appropriate emissions intensity for loads on the grid. The grid power demand is usually met using the lowest cost mix possible [15]. In many cases the lowest cost mix is the lowest carbon intensity mix as non-emitting sources have low fuel costs, although coal has fallen out of favor in many U.S. regions recently due to the associated emissions. As more power is needed, this merit order dictates that capacity is taken from the next cheapest source, as illustrated in Figure 1, for a grid where coal and gas are the primary marginal sources at a particular time and oil is only utilized at the highest demand levels. Other factors impacting the marginal mix are availability, startup costs, ramping rate constraints, and minimum up-time."

But here's where I started to get lost. Because after that description, the next 20 or so pages are just explaining grid emissions rates and average mix vs. marginal mix and bla bla bla. And this is all important stuff, but it's all stuff that should be in the appendix.

Why? Because the paper isn't about the grid makeup, but rather BEV emissions vs. non-BEV emissions.

Furthermore, if the paper's argument mainly hinges on marginal rates vs. average rates, it does a poor job explaining if BEV's actually force a utility or grid operator to use marginal fuel sources or average fuel sources. For instance, most BEVs I know of charge at night! In that case, the power being supplied is unlikely to come from marginal sources because during the night, grid demand is low and supply is coming from base load power (ie a lot of nuclear and wind). I don't think the paper really digs into this.

All the paper needs is a simple table that shows something like this (supported by the appendix work):

Once you've laid out your grid emission rates, you then simply multiply these rates by the efficiency of the BEV vs. non-BEV (factoring some multiplier for transmission line losses). I couldn't find the equivalent CO2/kWh of gasoline in the paper, and I don't know if it's not in the paper (which it absolutely should be) or if I just didnt see it, so let's do our own research!

Engineering Toolbox lists the kgCO2/Gallon of gasoline at 9.2 kgCO2/Gallon and this link from EPA has it at 8.9 kgCO2/Gallon. Once you know the emission rates of different fuels you can then compare apples to apples (comparing cars in the same class). The below Table 2 compares a Tesla 3 vs. BMW 328 as they are similar types of vehicles and I found the different efficiencies online (miles/(unit energy storage)):

And what does this very simple and easily communicated analysis show? That the ICE vehicle has 3 times more CO2 emissions than the BEV using US Avg Grid emission rates! Now of course the authors of the paper could show how marginal rates differ from average rates and whatnot, but in terms of analysis, the correct way is to compare vehicles in the same vehicle class using their average efficiencies (or whatever as long as it's like to like, ie City Mileage or Highway Mileage) and just show the different comparison tables. THAT'S IT! But the paper doesn't do that. The paper isn't simple, and I'm not sure if it's on purpose (to obfuscate) or if it's just poor communication.

Finally, now that I've gotten the basic stuff out of the way, the biggest issue with the paper is THAT THE ENTIRE PREMISE IS INCORRECT. The paper approaches the problem as if the goal is to immediately convert all vehicles to BEV today using today's grid emissions. And we know that is not reality. Converting the fleet to BEV is a DECADES LONG PROCESS. Decades. Yes, there are states that say by 2030 or 2035 no new ICE vehicles can be sold, but if an average ICE lifespan is 25 years, if the vehicles stop being sold by 2035 they will still be on the road in 2060. 2060!!!!

You see, the goal we are all working on in the electric car space is not the IMMEDIATE conversion of all ICE to BEV. That's impossible and a dumb thing to even talk about for a multitude of reasons. The goal we're working towards is simply starting the work to show that BEVs are a better driving experience and better for the environment so the fleet of vehicles begins to transition from fossil fuels to electric. (I'm not going to even address security of energy supply in this paper but Russia's war in Ukraine has again demonstrated reliance on despotic regimes for fossil fuel supply).

And I'm not even done yet.

The paper tries to make the point that BEVs arent always cleaner than ICE because of grid makeup as if the grid will always be dirty. BEVs are one side of the equation, cleaning up the grid is the other side of the equation and both need to be worked on in parallel! You see, utilities in the United States are regulated monopolies with atrocious levels of R&D and no huge incentives for innovation. You can let John Oliver from Last Week Tonight educate you on the topic if you really want to be depressed. The onus to clean up the grid is not on adopters of BEVs but rather on utility regulators! The populace shouldn't have to try and decipher a difficult paper that is clear as mud to determine if BEVs are cleaner than ICE (they are) but rather, the populace needs to demand that regulators mandate a cleaner grid from the utilities they regulate!!! And you can always power a BEV with rooftop solar on your home...you can't get a garage gas tank to fuel your ICE vehicle.

I didn't even mention local air pollution and smog and all that!

Ok, I think I've made my point...let the debate begin, but please, for Heaven's sake, keep the data analysis simple (see above tables). All you need is data for Table 1 (emission rates (would be good to see different emission rate projections for the future decades), and maybe another column for line losses?) and Table 2 (comparison of BEV vs. ICE in same vehicle class). That's it.

Other reading on the topic:

Yale - Indirect and Direct BEV emissions: https://environment.yale.edu/news/article/yse-study-finds-electric-vehicles-provide-lower-carbon-emissions-through-additional

https://electrek.co/2022/03/04/light-duty-evs-have-64-lower-life-cycle-emissions-than-ice-vehicles-ford-study/