An independent study by Greenly, a specialist in carbon footprint measurement and management, has challenged Tesla’s 2023 claim that its vehicles had helped avoid 20 million tonnes in carbon equivalent (CO2e) emissions. Estimating the actual figure at 28-49% lower, Greenly indicates inaccuracies in Tesla’s emissions model and sounds the alarm over a general lack of transparency in emissions reporting within the Electric Vehicle (EV) industry

Tesla’s pioneering role in the rise of EVs over the past decade remains undeniable. Despite mounting competition from the likes of Rivian and BYD in recent years, as well as the progress by legacy automakers such as Ford and BMW, Elon Musk’s flagship company still makes up nearly half of US EV sales. Globally, nearly 1 in 5 EV sales in 2024 were Teslas. In this context, Tesla’s claim of avoiding 20 million tonnes in CO2e emissions over 2023 is significant - avoiding nearly 40% of Greece’s (51.67m tonnes in 2023) total emissions for the year.

However, reapplying Tesla’s own methodology, Greenly has raised serious concerns over the replicability of the findings. Evaluating the avoided emissions at 10.2-14.4 million tonnes, Greenly highlights a divergence based on how one accounts for emissions from vehicles powered by internal combustion engines (ICE). According to Greenly, the gap may be explained by a possible overestimation of ICE emissions. Or, Tesla could have underestimated emissions from producing the electricity needed to charge its own vehicles.

Tesla’s methodology: A seemingly fair premise

The claim of 20 million tCO2e avoided can be verified as follows - calculating the emissions of its entire fleet as if they were diesel vehicles, and subtracting the EVs’ ‘true’ emissions from the result obtained.

To arrive at the size of its fleet, Greenly used Tesla’s historical sales data since 2014, to estimate the number of Tesla cars in each country. This is important because the emissions factor (EF) taken by Tesla for fuel combustion differ slightly for the United States and Europe, consistent with their differing energy mixes. Using historical sales data since 2014, Greenly estimates Tesla’s global fleet at 5.35 million vehicles by the end of 2023.

To calculate the amount of fuel consumed, Tesla uses an average fuel efficiency figure of comparable ICE vehicles, assuming that both diesel cars and Teslas travel 200,000 miles over 17 years. A similar approach is used to calculate ‘combustion emissions’ for EVs. The average emissions of electricity generation in a country are put together with Tesla’s average mileage of 3.8 km per kWh, to calculate emissions per mile.

For emissions during manufacturing, Greenly reuses Tesla’s assumption that ICE vehicles emit 10 metric tons, while EVs emit 20 metric tons. This is consistent with findings that EVs have a larger footprint at the manufacturing stage due to the manufacturing of batteries. These emissions are then spread over the use period of 17 years to arrive at annual emissions.

With this approach, Greenly sought to replicate Tesla’s claims. To calculate fuel combustion emissions, Greenly supposes that an ICE vehicle emits 333 gCO2e per mile (UK GHG Conversion factors 2024, Large Passenger Car, diesel). As a result, an equivalent fleet of ICE vehicles would have emitted an estimated 24.1 million tCO2e. It is important to note that the actual figure may be slightly higher, since the UK conversion factor excludes upstream emissions.

For replicating the emissions of Tesla’s actual fleet, Greenly used the 2024 energy emission factors provided by the International Energy Agency (IEA). This provides a granular view of electric emissions at country level, including upstream emissions. The total emissions of Tesla’s EVs were thus estimated at 13.9 million tCO2e for 2023, a saving of 10.2 million tCO2e, barely half of Tesla’s claims.

Greenly recognises that Tesla’s claim aligns conceptually with their methodology, and the lack of replicability is likely down to specific assumptions within Tesla’s data. In this case, the divergence mostly arises from the way emissions in use are calculated - concretely ICE emissions per mile, and emissions from charging EVs.

Analysing the divergence - A significant overestimate of ICE emissions

Given Tesla’s methodology, the eco-friendliness of their vehicles depends significantly on the ICE vehicles they are compared to - greater the ICE emissions, higher the avoided emissions. Older models or heavier vehicles would have a higher emissions factor, compared to the actual average ICE fleet, which is gaining in efficiency.

For the United States, Tesla assumes that ICE cars emit 445g CO2e per mile (including upstream emissions), drawing from data provided by Consumer Reports. While plausible, it is surprising that Tesla has declined to use data from the Environmental Protection Agency (EPA), usually the reference in the US for climate topics. For Europe, where Tesla assumes an emissions factor of 459g CO2e per mile (including upstream emissions), the figure seems overestimated. In the UK, the corresponding factor for large diesel passenger cars stands at 415g CO2e/mile - 10% lower.

It is worth remembering that the methodology is very sensitive to changes in emissions factors for ICEs - if Greenly were to modify the EF in its own scenario to 400 gCO2e/mile rather than 333 gCO2e as initially used, avoided emissions rise by 40% to 14.4 million tCO2e. Therefore, a higher ICE emissions factor also allows Tesla to claim higher avoided emissions - a lever certainly used in arriving at its claimed 20 million tCO2e.

Analysing the divergence - Underestimated grid emissions

Another variable remains Tesla’s approach to emissions from the electricity grid, underpinning the infrastructure which charges its own EVs. A grid transmitting electricity generated from renewables would have a significantly lower footprint than one dependent on fossil fuels.

Much like Greenly, Tesla has used different emission factors for different countries. However, Tesla’s said value for the US (116 gCO2e/mile) seems far too low. Greenly’s closest estimate of 206 gCO2e/mile, drawn from data from the International Energy Agency, is 78% higher.

Solar charging cannot account for this difference since Tesla already uses a separate EF for solar energy (72 gCO2e/mile). Tesla does not sufficiently detail how it arrived at either figure, and Greenly’s calculations assume that all EVs were charged through the electricity grid. Tesla’s lack of transparency regarding their methodology around grid emissions is unfortunate, making it impossible for Greenly to independently verify the impact of solar charging on emissions.

Analysing the divergence - Fleet Mileage and Manufacturing data

Over time, a number of studies have demonstrated that EVs emit far less in use than their ICE counterparts, with a higher footprint in manufacturing. This allows users to avoid further emissions the longer they drive a single EV. Thus, assumptions on fleet mileage and manufacturing become highly sensitive in the context of such a study.

While the base scenario assumes 200,000 miles driven, dropping it to 150,000 miles would also drop Tesla’s avoided emissions to 6.9 million tCO2e from 10.2 million tCO2e.

In the case of manufacturing, it is unclear if Tesla’s claims include upstream emissions related to battery production, mining and vehicle manufacturing. Excluding these emissions from Greenly’s base scenario, avoided emissions rise from 10.2M to 13.4M tCO2e, thus painting the carmaker in a more positive light.

A far-reaching economic and policy impact

Beyond the possibility of reduced investor confidence or the loss of public trust, such findings point to the need of independently verifiable methodologies to assess the positive impact of EVs, no matter the manufacturer. At a time when firms such as ExxonMobil are attracting accusations of misleading climate disclosures, such findings may seriously shake the EV industry.

Should carmakers be found to be systematically overinflating carbon emissions, not only would it be reminiscent of the Dieselgate scandal, but it could translate to stricter norms on the EV sector at large. Coupled with the Trump administration’s unfriendly view of renewable energies, an erosion of public confidence in EVs would seriously disrupt investment in EV infrastructure at a time when a phaseout of ICE vehicles is desperately needed.

By publishing its own methodology, Greenly demonstrates that accurate and independently verifiable calculations are possible, enabling third-party auditing. Greenly hopes that carmakers follow this example to maintain trust in the industry and ensure that valuable time is not lost. Accurate carbon accounting is a prerequisite to ensure that carbon markets work effectively.
www.licencek.com