Comparison of Energy Efficiency and CO2 of Gasoline and Electric Vehicles

Many studies fail to use the lower heating value of the fuel, or fail to use the correct heating value of the fuel.

Heating value of fuel does not correlate to rotational motion.
Whatever conclusion you are trying to present is flawed, literally from the first sentence.

Willem, you are to be commended on your diligent pursuit of the truth. It’s a lot of hard work, and as you know – the devil is in the details.

But you’re barely scratching the surface. The U.S. DOE has an ongoing, comprehensive analysis of full-spectrum, well-to-wheels emissions for every conceivable fuel and powertrain combination known as “The Greenhouse Gases, Regulated Emissions, and Energy use in Transportation Model” (under the tortured acronym of GREET). The following list describes only updates to the 2016 version of the model:

• Updated water consumption factors for the refining processes of petroleum fuels, thermoelectric and hydroelectric power generation, and hydrogen production processes.
• Updated high-octane-fuel (HOF, with research octane number of 100) pathways in GREET with E10, E25, and E40 ethanol blends.
• Developed new pathways for hydrogen production from dark fermentation of lignocellulosic biomass, high temperature steam electrolysis, and reforming of bio-derived liquids.
• Expanded and updated wastewater sludge-derived fuels and waste-to-energy pathways.
• Updated CCLUB to include N2O emissions associated with LUC GHG emissions for ethanol production scenarios.
• Updated emission factors for agricultural equipment and nonroad emissions.
• Updated DME production from fossil natural gas, natural gas-derived methanol, and renewable feedstocks.
• Updated the fuel economy and material composition of the light duty vehicles.
• Added bio-based ethanol-to-jet pathways from corn and cellulosic biomass, and sugar-to-jet from cellulosic biomass.
• Updated energy intensity and emissions for freight and intercity passenger rail transportation.
• Developed in GREET.net emissions within various regional aggregations of the United States, including regionalization at the state, eGrid subregion, NERC levels.
• Updated methane leakage emissions for natural gas pathways.
• Updated vented, fugitive, and flaring greenhouse gas emissions from crude oil production.
• Updated fugitive CH4 emissions from open channel transportation of vinasse for Brazilian sugarcane ethanol pathway.
• Updated U.S. electricity generation mix based on Energy Information Administration (EIA)’s Annual Energy Outlook (AEO) 2016.
• Updated U.S. crude oil mix and weighted average distance using EIA’s company level imports data and AEO and Canadian Association of Petroleum Producers’ market report.
• Updated farming energy and fertilizer intensity of corn, soybean, willow, poplar, miscanthus, and switchgrass.
• Added LiNCA battery for battery electric vehicles into GREET2, including the NCA cathode production, Aluminum hydroxide and alumina sulfate production.

GREET can be downloaded in Excel format and you can substitute any of its default values for your own. So I put your values into the model, for comparison. Here are the results:

http://www.thorium-now.org/images/greet_ic-vs-ev.jpg

As you can see, the model shows an EV creating 2 grams/mi fewer GHG emissions than E10 gasoline. Not that impressive. But for the electricity generation mix I had entered for this scenario, nearly half came from coal (in California in 2017, less than 8% comes from coal).

Highly recommended, if you use Excel, to download and run it yourself (it gets addictive).

Willem,

I’m suspicious of any model with hidden calculations, as are the scientists at Argonne. That’s why all of the calculations in both the Excel version of GREET and online are in full view.

I don’t know there are errors in your calculations, I suspect they’re 100% accurate. That your model represents “electricity arriving at the meter (after all losses)” is inaccurate – you’re considering a fraction of the criteria necessary to make such a calculation.

I’ve met the team which developed GREET – they are single-minded about the job before them. When I asked a team leader why nuclear plants are being closed, his reply was “we don’t consider politics or policy matters.” That’s socialism at work, and why you’ll never get an honest answer about energy from the private sector.

And why I won’t be surprised if the GREET model disappears during the Trump administration’s tenure or assumptions disappear from view, replaced by “alternative assumptions” based on “alternative facts”.

Seems like this is missing that EVs are the killer app for renewables — they will soak up the excess power during peak wind or sunshine and use it for transportation, and the used batteries will serve as backup supplies in homes, businesses, and utilities.

Willem, right off the bat –

• You have no accounting for regenerative braking, which saves 20% of energy consumption in an EV in urban driving.

http://large.stanford.edu/courses/2016/ph240/brown1/

• You choose the least efficient mode of driving as representative for both gasoline and EVs – 65 mph for one hour. Lithium-Ion batteries are up to 8% less efficient under high load.

https://www.researchgate.net/figure/231169804_fig12_Figure-16-Peukert-co...

• 2 g/mi less than internal combustion when half of an EV’s electricity is generated by coal. ~30% less when generated with an average California mix.

• What’s the source of your figure 4065.97 TWh for electricity generation, or any of the other figures in “The Source-to-Wheel Efficiency of an Electric Vehicle “?

• What’s the source of your estimate that “future E10s” will be capable of a 30% improvement in efficiency?

You’re trying hard to pick the worst assumptions for EVs/best for E10, rather than ones representative of average vehicles and driving situations.

You’re way, way overestimating what EVs can do any time soon.

Take Tesla’s Gigafactory, with a capacity of 50 GWh of battery packs per year.  Average grid load in the USA is about 450 GW.  One Gigafactory-year’s worth of battery packs could support the average US grid load for…

about 6 minutes 40 seconds.  After 9 years of production, you could buffer a whole HOUR! /sarc

It would be a bit more for the unreliables alone.  WInd plus solar came to 263,239 GWh in 2016.  Dividing by 8784 hours, that’s 29.97 GW average.  That would give you a whole 100 minutes, average.

EVs are just not going to be a large-scale energy buffer for the grid for a long time.  They can buffer power on a scale of minutes, but that’s a very different thing.

Nice article. Always a pleasure to see honest and informed numeracy brought to bear on policy issues. Even it there’s room to challenge some of the numbers presented or omitted, at least the numbers are there and the assumptions are spelled out. The quality of the debate is enhanced.

I won’t weigh in about the differences between the GREET model that Bob recommends and the model that Willem put together. But Willem notes that he does not trust “Excel programs of which I have not seen the equations and assumptions,” Fair enough, but why not download the spreadsheet and take a look? The equations should be there, and one can generally figure out the assumptions. Spreadsheets aren’t rocket science.

What I’m more interested in is the issue of applying the current mix of generation resources to calculate emissions for electric vehicles. As it happens, most of the individuals I know of who drive electric cars have solar panels on their roofs, and do most of their vehicle charging at home. Granted, I live in Tesla country, and Tesla owners are a preselected lot. But Hops’ observation about vehicle charging being the “killer app” for renewables still stands. And a rigorous analysis of the effective carbon emissions of electric vehicles needs to consider not the current raw mix of generation sources, but rather the marginal effect of EV sales on the generation mix. That’s assuming that anyone can figure a way to calculate the marginal effect.

I also wonder about Willem’s assertion that “future E10 vehicles likely will become more efficient more quickly, than the US electrical system will reduce its CO2 intensity, lb CO2/kWh.” Could be, but it’s not obvious to me. In fact, I rather expect the opposite. I think we’re approaching an inflection point, where the rate of decarbonization of the electricity supply will accelerate noticeably.

Be that as it may, I think it’s important to keep in mind that what we need to do is not to reduce carbon emissions by some target amount ‘X’, after which we’ll be fine. If ‘X’ is any value less than 100%, then we won’t be fine; we (or our children) will just take a bit longer to hang. So, yes, electric vehicles do have indirect carbon emissions, but as the electricity supply shifts to lower carbon intensity, the carbon intensity of the electric vehicle fleet shifts with it. The same cannot be said for the E10 vehicle fleet.

By then of 2017, Tesla will announce the locations of 4 more gigafactories.This just starting, and will not be static or even linear.

Willem,
You forget to calculate the far more important benefit of electric cars:
The much lower emissions of small particles*) as that kills people.
_______
*) Particulate matter (PM) causes a life shortening of ~2 years for people living in city centers with busy traffic, or living along busy highways (EU study)!
So nowadays older cars and diesel cars are no longer allowed in many cities in Germany and also in some cities of NL, etc.

Tesla will announce the locations of 4 more gigafactories.

Thus extending EPs calculation to 30 minutes from six. Hundreds of giga factories are required to make a grid sized storage play, and then some new facility distribute, and collect, millions of tons of battery. Don’t forget the associated expansion of lithium ore. The price is rising.

Also, if Tesla does build four more, they lock in current battery performance within 5% or so for the following decade, maybe two. The latest, greatest laboratory battery by press release simply won’t matter.

Tell it to German coal and Volkswagon diesel. Don’t bother with the cut n paste on how coal will retire in 2022, 3032, or whatever.

I agree that going to hybrids ASAP will have the most immediate impact on reducing emissions. I’m on record predicting that in the near future (5 – 10 year timespan) virtually ALL new production vehicles will be either pure BEVs or hybrids. They will have electric drive systems, because once the supply chains and production lines are established for it, electric drive delivers better performance and reliability at potentially lower cost than mechanical drive.

In this case, “performance” includes not just torque and acceleration, but precise control of torque to individual wheels for features like traction and skid control and enhanced cornering. Those features have all been implemented for high-end mechanical drives, but there they cost more and are clumsier. With electric drives, they’re just controller firmware. Those benefits are over and above regenerative braking and non-idling that give hybrids superior mileage.

Given the above, the issue becomes how big to make the hybrid drive batteries and whether to include plug-in capability. For me, it’s a no-brainer in favor of plug-in capability. Statistically, it only takes 10 – 15 kWh of battery to enable more than 50% of miles driven in battery-electric mode. 10 – 15 kWh isn’t exactly trivial, but it’s a hell of a lot less than the 100 kWh that pure BEVs are trending toward. With 10 – 15 kWh of storage, the per-mile driving cost plunges and the effective long term mileage gets into 100 mpg range.

Also, with that much electric capacity, emission reductions for transportation would benefit from de-carbonization of the grid.

Particulate matter (PM) emissions by power plants are very insignificant compared to that of traffic. Nowadays power plants here are obliged to filter all PM out of their exhaust gasses. They use a.o. cycloned, electrostatic filters, etc.

Dutch life expectancy is almost two years higher than that of USA…
Furthermore USA spends substantial bigger part of its GDP for health care than Netherlands (in 2010: USA, 16.2%; NL, 10.8%).
So relevant facts indicate the opposite of your assumption:
“NL being too crowded to remain healthy.”

Though it’s also clear that US citizens get less value for their health care money (also corrected for GDP/PP). Probably because US health care is reigned by huge oligarchical private enterprises which make good profits.

Can confirm.  Data point:  my car maxes out at mid-high 20 miles all-electric range in optimal weather, but I’m achieving well over 100 MPG average.  (The car reports 131.3 MPG lifetime as of getting home this evening.)  It does this with less than 8 kWh of battery.

Down at the actual hybrid end, we have a lot more options than just batteries.  Ultracaps are great for high power/mass and help relieve compromises that engines have to make to “driveability” by sacrificing efficiency.  If you’ve got a fat energy buffer that gives instant response, your engine can lag like crazy and the driver doesn’t care.  If the designer can go all-in for efficiency and emissions they can both be quite a bit better.

Is that truly how the EPA calculates mileage for plug-in hybrids?

I know that the mileage that is displayed on the car console (for the Prius, anyway) is a straightforward “miles driven / gallons consumed” since reset. For a plug-in, that number would obviously vary a lot depending on driving pattern. (It would go down substantially, e.g., if the car were driven on long trips, with little opportunity for charging, and would go up if it were driven on a lot of short trips with recharging in between most trips.)

My assumption — and that’s all it is — has been that the EPA rating was calculated on the basis of a standardized driving cycle that accounted for length of trips and opportunities for recharging. That’s what would make sense, and it’s consistent with the standardized driving cycles used to calculate figures for “city driving” and “highway driving” for non plug-in vehicles. But I’ve never looked into it.

I’ll raise a separate flag about trying to compare technologies on the basis of “primary energy” consumed. A megajoule is precisely defined in physics, but “primary energy” is a fuzzy concept. What’s the “primary energy” consumption for hot water from a solar thermal collector? And does it really make sense to equate a unit of thermal energy from coal with a one from natural gas or one from cow dung?

What really matters in terms of transportation efficiency and global warming is GHG emissions per distance travelled. The significant fact about E10 is that for a given vehicle make and model, that number will be relatively constant over the life of the vehicle — and will never be zero. For a BEV or plug-in hybrid, the number will depend on the carbon intensity of the charging source. It will decline as the carbon intensity of the grid in its region declines, or go to zero if the vehicle is charged exclusively from solar panels.

The latter isn’t improbable — at least in California — since in addition to rooftop solar on vehicle owners homes, there’s a trend for businesses to install solar panels in their parking lots. Free vehicle charging for employees during work hours is a nice perk that appeals to workers of the type that many companies here want to attract. (Not to mention being good for corporate image in California’s culture.)

How many kWh from the meter for 25 miles of travel?

I put it on the Kill-A-Watt once and it took about 7.5 kWh from the wall.  (Okay, the post in the garage.)  IOW it’s about 300 Wh/mile depending how I drive it.

What this incurs in emissions depends on the generation mix of the moment.  If the marginal watt is coming from the 3400 MW of coal-fired plants in Monroe, it’s one thing.  If it’s coming from the Donald C. Cook nuclear plant or the wind farms in Berrien county, it’s close to zero.  If it’s coming from the Midland Cogeneration Venture it’s intermediate.  If it’s coming from the pumped storage station at Ludington, it’s perhaps a 25% multiplier over whatever was used to fill it (probably wind or nuclear power these days, so also close to zero).

If we figure the emissions of a CCGT plant running in load-following mode at 400 gCO2/kWh, that plant is generating the marginal watt, and 300 Wh/mile, net emissions for the vehicle come to 120 gCO2/mile or about 75 gCO2/km.  If we allow 30% non-emitting generation and credit the vehicle with that, it falls to 84 gCO2/mile.  Pegging the emissions of E10 at 17.68 lbs (8.018 kg) per gallon, this is equivalent to an E10 vehicle achieving 66.8/95.5 MPG respectively.

Thanks Willem. Here’s one I will hand to you: the inefficiency of EVs in cold weather.

I think we’d both conclude that EVs can produce significant benefits in the right circumstances, and that gasoline/E10 vehicles are less dependent on circumstance.

A study came out a few years ago by a national environmental org (forget which one) which showed EVs were significantly worse for emissions in select parts of the country (basically, the Southeast – coal country). There was a flurry of denial in the EV community, but try as I might I could not fault the conclusions they drew therein.

I would like to think in the bigger picture – where government has more say in the collective emissions we generate, as it should – generating our energy collectively from point sources like power plants enables us to improve everyone’s emissions en masse, and not rely on individual incentive. When it comes to finding solutions to societal problems, I don’t believe we can count on the judgment of individuals to carry the day.

When it comes to finding solutions to societal problems, I don’t believe we can count on the judgment of individuals to carry the day.

Hoo boy! As it happens, I agree with you on that. But you’re poking at a hornets’ nest. It’s certain to raise the hackles of libertarians and small-government conservatives. As perhaps it should.

I’m not unsympathetic with the libertarians and small-government conservatives. My dad was a consulting engineer; he constantly railed against the idiocy of petty bureaucrats and the way blind code compliance inflated costs for projects in situations where compliance wasn’t functionally necessary. He hated what we now call the “nanny state”. His views rubbed off on me. For much of my life, I identified as libertarian.

I haven’t exactly changed my mind about those views. Now, however, I try to see things from a wider perspective. I look at society and the world in terms of system dynamics and game theory, rejecting ideologies as “what one falls back on in the absence of science and honest attempts to understand the issues”. It’s a pragmatic POV that starts with values and focuses on what will actually work to advance those values.

The tension between the interests of the individual vs. those of the group is incredibly fundamental. It has literally shaped the evolution of life on the planet. When you look deeply, the “prisoner’s dilemma” turns up everywhere. Evolution is just as much about cooperation and enabling win-win strategies as it is about competition and being a more successful predator.

So what works for solving societal problems when the interests of society collide with the short-term interests of individuals? Regulations are necessary. However, in formulating the regulations, one needs to be mindful of how they can be gamed. Engineering’s KISS principle (“keep it simple, stupid”) becomes supremely important. As a policy-maker, resist the temptation to specify solutions. Focus on incentive structures and alignment of incentives. People are very good at grasping incentive structures, and figuring out how to benefit from them.

Saw an interview recently with a Tesla executive talking about the cost and capacity of batteries from the Gigafactory for Model 3. He said that they would be 30% cheaper (per kWh) with 30% higher energy density. IOW, the same cost per kg, but 30% fewer kg needed to do the same job.

His explanation should be a cautionary note for those counting on continued exponential improvement in the economics of battery storage. It was simple: once you’ve gotten to high production volumes with factory equipment designed and optimized for the production task, then cost of product becomes proportional to mass, regardless of complexity. (More properly, cost of product becomes dominated by the cost of raw source materials. A kilo of platinum is always going to cost far more than a kilo of iron — or of dirt.)

He didn’t say it, but the message I heard was that with the gigafactory, Tesla is now down near the limits imposed by raw materials and battery chemistry. Battery lifetimes will likely improve, and tweaks to the electrolyte and electrode structures will likely yield further gains in energy density. But the gains won’t be dramatic.

Hmmm. Does that shed a new light on Musk’s latest venture? Is his “Boring company” perhaps a covert sally into mining technology? It does kinda fit with his interest in Mars colonies.

Battery lifetimes will likely improve.

Good prediction.

Recent article

Interesting and fun story that shows how important measurements are as a first step in almost any improvement.

Willem, very well written and detailed article as usual. The following are a few issues that could impact and increasingly support your conclusions: 1) my detailed analysis of the GREET model well-to-wheel (WTW) calculations (re. my past TEC article: “Is Ethanol a Cost Effective Solution to Climate Change?” http://www.theenergycollective.com/jemiller_ep/172526/ethanol-cost-effec... ) which clearly showed that the Argonne folks relied on less-than-fully accurate international data bases, and overestimated WTW fossil fuels consumption (and associated carbon equivalent emissions) of petroleum fuels by up to about 9%., 2) what is often overlooked in evaluating EV energy usage and efficiency is the significant impact of ambient (battery) temperatures; which deteriorate quite significantly during cold winter environments (100 degree F). This factor also affects batteries’ capacities and lifespans. And, 3) the negative impacts of higher charging voltages, capacities and cycles (longer term charging frequencies) on battery charging efficiencies and full-available charge capacities. Higher voltage (220V+) definitely reduces charging times, but unfortunately, negatively impacts overall charging efficiencies and wear on batteries’ capacities. Those who perceived that EV batteries can be effectively used as residential/commercial backup battery capacities in order to directionally increase the capacity factors of intermediate solar and wind power, are going to be in for a big surprise as to the efficiency and wear on their EV battery packs; and shortened life’s/charging capacities’ impacts. Under these circumstances, EV owners will be very surprised as to the actual efficiencies and added cost of resultant shorten battery life’s.

Roger, agree the KISS principle can be equally effective at constructing regulations as it is in engineering, if not more so.

That’s why reknowned climatologist James Hansen has been an advocate (some say founder) of the revenue-neutral carbon tax, aka “fee and dividend”. In Hansen’s version, a tax directly proportional to its molar carbon is added to any quantity of fuel extracted in or imported to a jurisdiction. Once a month, after deducting administrative fees, revenues are divided and distributed equally to all citizens. If your consumption is less than average, you make money. If more, your lose (taxes you pay at the pump will be more than your refund check).

KISS – it doesn’t get much simpler than that. And it works – British Columbia’s RNT has reduced gasoline consumption by 17%. Then why doesn’t the U.S. have a revenue-neutral carbon tax?

It would work too well – it would limit the profit potential of selling gasoline. There is your head-to-head collision of the interests of society with the short term interests of individuals, or corporations, or any special interest. If we permit self-interest to rule the day, we might as well give up on climate change – it is, quite literally, hopeless.

Simply-worded regulations are the worst of all for corporations, because they introduce the wild card of human judgment – there’s a judge somewhere who will interpret them, and quite possibly not in the corporation’s best interest. So today teams of lawyers construct reams of law, with carefully-designed loopholes to maximize profit.

In sum, these laws categorically defy public interest. My favorite example is the Energy Act of 2005, which repealed the Public Utility Holding Company Act of 1935 (PUHCA). EnACT2005 is a monstrous, 551-page behemoth which very specifically eliminated public protections inherent in the 59-page PUHCA. To restore some sanity to this situation, a start would be assigning word and amendment limits to any law before Congress.

And of course, Citizens United has got to go. Representative democracy in the U.S. has always been a tenuous balance of capitalism (self interest) and socialism (public interest). Citizen’s United has upset that balance by rendering public interest moot – it’s become a battle between exclusively private interests, one in which laws are shaped by campaign contributions instead of popular vote.

Interestingly, the 1802 standoff between Hamiltonian Federalists and Jeffersonian Republicans was identical, in both ideology and acrimony, to today’s Democrat – Republican split. We can hope for a stalemate, but never the twain shall agree.

That’s why I specified “marginal watt”, Willem.  That’s the average of whatever sources are ramping up and down as demand varies in whatever locality, and of course it varies by region.

Willem, you and EP are comparing two different quantities: energy and power.

Strictly speaking energy on a grid does not “travel” at the speed of light. Changes in power – the rate of transfer of energy – do, and it’s very possible to track them. SCADA networks at independent system operators record power changes at grid junctions every second of every day and, with some accounting, can determine with a fair degree of accuracy the sources of energy on a grid. That’s how generators get paid.

But I’m not an engineer, and I love to be corrected.

Willem, I’m confused. The reasons you give for why it makes sense to equate units of thermal energy from sources with very different characteristics are exactly the reasons I’d use to argue that they shouldn’t be equated.

Did you perhaps take “make sense to equate” as “make sense to weigh”? I.e., “make sense to consider the differences”?

Willem, you’re still confusing energy with power.

If you know the sources of energy fed into a grid, you know what comes out. There’s is no net surplus, otherwise by the first law of thermodynamics lines would melt from their poles. The mix going into everyone’s meter is exactly what is going into the grid at that time – nothing “generic” about it.

Energy is fed into the grid over time, not instantaneously. If one kW is added for one hour, the total energy transferred is one kWh. Transferring energy through a grid instantaneously would require an infinite amount of power:

P * t = E
For the case t = 0:
P * 0 = E
P = E / 0 (undefined)

A point that I’ve made elsewhere is that a significant amount of EV charging is done from solar energy that doesn’t get to the grid. I don’t know how much; I’ve never seen specific statistics. But on the basis of very rough and unscientific sampling from California EV owners I happen to know, it could easily be half.

Even the indirect carbon emissions from EVs / PHEVs charged from grid will differ depending on when the charging is done. If it’s done while regional load is dropping, it will reduce the drop, and sometimes avoid the need for cycling of a dispatchable unit. Since nearly all dispatchable resources (other than storage) are inefficient during startup and shutdown, avoiding the need to cycle is good. It reduces operating costs and lowers emissions per kWh.

To reap those benefits, however, signaling between the system operator and the charging units is needed. A form of “smart grid”.

What is a marginal watt?

It’s the next watt of generation that will be added if demand increases.  It’s what determines the incremental change in emissions from adding load.  This works the other way also; the Joe Wheatley analysis of wind power in Ireland showed that the incremental decrease in emissions from adding wind was far less than 1:1.

How is this marginal watt so special, that it does not need to be treated as having to obey the laws of physics.

About those laws of physics…

Transmission lines have impedance.  The higher the impedance, the less power flows through them for a given set of conditions between the ends.  Ceteris paribus, longer lines have higher impedance.  This means that distant generators will contribute less to a load than a local generator.  Physics says that Iowa wind power isn’t going to contribute significantly to New York consumption given an AC grid in between (point-to-point HVDC is another matter).

This means that the emissions from charging an EV depend substantially on where it’s charged.  Also when, because in most places that aren’t using e.g. 100% hydropower the marginal watt comes from different generators at different times of day.