Update, April 17, 2019, 3:40 p.m.: References to the Camry Hybrid LE and XLE have been corrected from a previous version, which incorrectly referred to 4- and 6-cylinder versions.
When I started writing about energy storage investments in July 2008, the politically compelling arguments for vehicle electrification included increased national security from energy diversification, reduced CO2 emissions, and sanctuary from the tyranny of imported oil. Today, the US is the world's top oil producer, its energy-related CO2 emissions have fallen by more than 10%, oil prices have fallen by more than 50%, and none of the gains were EV related. These days, the only plausible justification for vehicle electrification is reduced CO2 emissions, but that justification is more smoke and mirrors than substance.
Since investing in smoke and mirrors is rarely successful in the long term, this article highlights the grave intellectual flaws in sustainability myths fabricated by Tesla (TSLA) and explains why its long-range Model 3 has a heavier CO2 footprint than Toyota's (TM) humble but highly efficient Camry Hybrid.
The Model 3 and the Camry Hybrid are both classified as mid-size cars
The long-range Model 3 and the Camry Hybrid are both classified as "mid-size cars" in the US and "D Segment cars" in Europe. My first table summarizes the salient details.
Width (ex. Mirrors)
One can quibble over differences in fit, finish, and zero to 60 acceleration, but the two cars are in the same vehicle class.
Magical thinking dominates electric vehicle discussions
I've always been amazed at the way politicians, ideologues, dreamers, and other electric vehicle hucksters, or EVangelicals, insist on:
- Comparing electric vehicles, or EVs, with fleet-average statistics, or the least efficient new cars, rather than the most efficient new cars
- Ignoring the fact that producing technology metals and manufacturing big lithium-ion batteries is extremely CO2 intensive and the cradle-to-gate emissions from battery manufacturing must be part of an accurate assessment of costs and benefits
- Ignoring the fact that while EVs have no tailpipe emissions, electricity from power plants has a significant CO2 footprint that must be part of an accurate assessment of costs and benefits
- Making absurd assumptions that EVs are charged by renewable electricity from wind and solar, rather than natural gas or other fossil fuels
- Doubling down on the renewable electricity absurdity by suggesting that using green electrons in an EV is more virtuous than using those electrons in a toaster oven.
Calculating the CO2 footprint of a Toyota Camry Hybrid
The CO2 emissions for a Toyota Camry Hybrid are simple to calculate. According to the U.S. Energy Information Agency, burning a gallon of gas produces 8,890 grams of CO2. Since the process of producing, refining and transporting petroleum products is roughly 80% efficient, the well-to-wheels value is 11,113 grams of CO2 per gallon. According to their EPA window stickers:
- Toyota's Camry Hybrid LE uses 1.9 gallons of gasoline per 100 miles
- Toyota's Camry Hybrid XLE uses 2.1 gallons of gasoline per 100 miles
These numbers translate to well-to-wheels CO2 emission footprints of 211.1 grams per mile for the Camry Hybrid LE and 244.5 grams per mile for the Camry Hybrid XLE.
Calculating the CO2 footprint of a Tesla Model 3
The CO2 emissions for a Tesla Model 3 are more challenging to calculate because an accurate assessment of costs and benefits must account for:
- The incremental greenhouse gases associated with producing technology metals and manufacturing enormous lithium-ion battery packs
- The greenhouse gases associated with generating, distributing, and using electricity to charge EV batteries
Unfortunately, the assumptions people make while performing the necessary calculations can and do vary widely. It's not so much a case of "figures don't lie, but liars figure" as a situation where "a man hears what he wants to hear and disregards the rest."
Emissions from battery manufacturing. Developing a reasonable estimate of the cradle-to-gate CO2 emissions from manufacturing a 75-kWh battery pack for a long-range Model 3 is complicated, but not impossible. This graph from the Norwegian University of Science and Technology summarizes the conclusions of sixteen different reports on the cradle-to-gate greenhouse gas emissions from battery pack manufacturing.
Fourteen of the reports were based on assumptions, secondary data, and models crafted by governmental agencies. Two of the reports, however, were bottom-up analyses based on comprehensive bill of materials and primary energy input data from battery cell and pack manufacturers. The 2016 Kim study, which reported cradle-to-gate CO2 emissions of 140 kilograms per kWh for an LMO/NCM cathode chemistry, was based on data from Ford and LG Chem. The 2014 Ellingsen study, which reported cradle-to-gate CO2 emissions of 173 kilograms per kWh for an NCM cathode chemistry, was based on data from Miljøbil Grenland.
Under the circumstances, I view the Kim and Ellingsen studies as highly credible. Since the raw materials for NCA cathodes have a 20% higher CO2 footprint than the raw materials for NMC cathodes and NCA batteries have a 20% higher energy density than NMC batteries, I believe an estimate of 170 kilograms per kWh for NCA batteries is quite reasonable. While I would rather have a detailed analysis from Tesla, squeezing that much information into a 280-character tweet is tough.
For this article, I have assumed that Tesla's batteries will have a calendar life of 12 years or 1.5x Tesla's eight-year warranty period and an average Model 3 will be driven 12,500 miles a year for a total of 150,000 miles before a replacement battery is required. This assumption will be proven wrong when a statistically valid sample of Model 3 batteries have reached end-of-life in the hands of typical Model 3 owners. While I think my assumptions are generous, others may consider them overly conservative. I'm happy to let each reader decide for himself.
When you turn the crank on my assumptions, the amortization required to recover the cradle-to-gate CO2 emissions for the battery in a long-range Tesla Model 3 is 85 grams per mile.
Emissions from producing electricity and charging batteries. One of the most hotly debated topics in the EV space is "how should we account for greenhouse gases associated with producing and distributing the electricity used to charge EVs?" Critics grouse that, "EVs are powered by coal," while EVangelicals proclaim that, "EVs are powered by wind and solar." Academics and bureaucrats frequently use an average emissions intensity for a particular city, county, state or country. Since all these assumptions ignore reality, I believe they're all deeply flawed.
I created the following graphs from electric-power production data for April 15, 2019, that I downloaded from CAISO's Renewables Watch web page. As you study the graphs, the key points to remember are:
- The stacked layer graph on the left shows hour-by-hour contributions of various classes of generating assets to system-wide power supplies in California;
- From bottom to top, the solid segments represent zero emissions power from nuclear, hydro, wind, other renewables, and solar;
- The patterned segments represent imported power, principally fueled by coal and natural gas, and local thermal power, principally fueled by natural gas;
- The black line shows hour to hour variations in relative CO2 intensity where imports and thermal were assigned a value of 100% while zero emissions sources have were assigned a value of 0%.
- The line graph on the right shows how power production from generating asset classes changes from hour-to-hour.
While graphs are great when you want to visualize the big picture, it's hard to beat numbers when you want to grasp the nitty-gritty. So, I've summarized critical hourly data for each class of generating assets in this table which states power in MW and energy in MWh.
The key points I noticed include:
- Power from all non-solar zero emissions sources was pretty stable on April 15th;
- PV- and thermal-solar facilities were quite productive during daylight hours but worthless during the evening demand peak when electricity was needed most;
- While the daily average CO2intensity was 36.2%, the hourly CO2 intensity was much higher during night-time hours when EVs were charging than it was during day-time hours when EVs were on the road.
As you ponder the graphs and the table, bear in mind that the simple act of plugging an EV into a wall socket is a consumption decision. Since the grid can't store power and utilities must precisely match supply and demand from moment to moment, every consumption decision requires the utility to make a corresponding production decision.
When you consider the available generating assets in CAISO's production base, it's crystal clear that imported power and local thermal-power are the only reliable and dispatchable generating assets that can ramp production up or down in response to consumption decisions.
- Wind power is unreliable intermittent baseload because producers cannot control the amount of electricity their wind turbines will generate at any particular time but utilities must take all electricity that wind turbines generate.
- Solar power is unreliable intermittent baseload because producers cannot control the amount of electricity their solar panels will generate at any particular time but utilities must take all electricity that solar facilities generate.
- Nuclear power is reliable stable baseload because once a decision to include nuclear power in the grid is made the plant produces electricity at a stable rate 24/7/365.
- Other renewables, including biomass, geothermal and small hydro are reliable stable baseload because they typically produce electricity at stable rates.
- Hydropower is reliable stable baseload because some producers can tune their facilities to produce less power during peak solar power intervals, but once a tuning decision is made the electricity output generally remains stable for hours.
If you give the graphs and the bullet points a few minutes (or hours, days or weeks) to sink in, the following factual statements become self-evident and incontrovertible.
- At any given moment in time, electricity production from nuclear, hydro, wind, other renewables, and solar is beyond anyone's control
- The only power production assets with the ability to respond to marginal demand are gas-fueled power plants and it doesn't matter whether the marginal demand arises at noon or midnight in San Diego, California or Bangor, Maine.
- The marginal fuel to charge an EV anywhere in the US will always be natural gas; and while increased renewables may improve grid average emissions, they cannot change the marginal fuel source for EV charging.
According to the US Energy Information Agency, natural gas turbines emit 599.8 grams of CO2 per kWh while more efficient combined cycle plants emit 512.4 grams of CO2 per kWh. Since combined cycle plants typically operate as baseload facilities and gas turbines typically fill the gaps between baseload and demand, I believe 600 grams of CO2 per kWh is the best figure to use for EV charging analysis.
The following table summarizes the calculations necessary to roll this number forward from the power plant to the open road in a long-range Tesla Model 3.
I was surprised when this analysis showed that Tesla's long-range Model 3 has a 24% heavier CO2 footprint than the Camry Hybrid LE and a 12% heavier CO2 footprint than Camry Hybrid XLE. I'm certain readers will be shocked because my conclusion is 180 degrees out of synch with Tesla's sustainability mythology. Frankly, there's simply no reason for investors, politicians or governments to support vehicle electrification if EVs offer no appreciable environmental benefit.
The analysis in this article probably won't have an immediate impact on Tesla's stock price because popular myths are macro issues that take time to percolate through the collective consciousness. It may, however, impact politicians who need to decide whether it makes sense to spend taxpayer money to subsidize adult toys that offer no environmental benefits. It may also impact the portfolio management decisions of responsible fund managers who are willing to question their assumptions instead of embracing a three-monkeys approach to investment management.
The mental image of cheap and sustainable electric drive promoted by politicians, ideologues, dreamers, and other electric vehicle hucksters is alluring beyond reckoning. It's also a classic free lunch fairy tale and those never end well for investors because there is no free lunch.
I continue to believe that Tesla's intrinsic investment value is zero and its stockholders will ultimately lose all their money. I cannot predict, however, when a significant portion of market participants will begin to see, hear and speak the unpleasant and inconvenient truths.
Many commenters have criticized this article for failing to plumb the depths of battery manufacturing in a Camry Hybrid. This graph from Page 8 of a 2012 UCLA study that was commissioned by the California Air Resources Board shows why. Once you take the BEV’s battery out of the equation, CO2 emissions from vehicle manufacturing are essentially the same for an ICE, an HEV, and a BEV. In my view, the clarity of the current simplified discussion outweighs any incremental accuracy from discussing everything but the kitchen sink.
Disclosure: I am/we are short TSLA TROUGH LONG DATED PUT OPTIONS. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.