Energy experts slam 'deeply flawed' Harvard-led study on Bitcoin mining's environmental impact

A recent widely reported Harvard-led study into air pollution associated with Bitcoin mining facilities in the United States is the latest academic paper to come under the ire of energy experts focused on the digital assets industry.

The study, published in Nature Communications, argues that Bitcoin mining in the U.S. significantly contributes to harmful air pollution, with 1.9 million Americans exposed to increased levels of fine particulate matter between mid-2022 and mid-2023. The researchers tracked 34 of the largest U.S. Bitcoin mines, claiming they consumed 32.3 TWh of electricity — 33% more than the city of Los Angeles — 85% of which from fossil fuels.

"Bitcoin mines, which are largely unregulated in the U.S., are an emerging and significant challenge to U.S. environmental health and air pollution regulation," the research team, co-led by Dr. Francesca Dominici and Gianluca Guidi of Harvard University's T.H. Chan School of Public Health, said. The authors cited Bitcoin mining's "enormous" energy demand and cross-state pollution limiting local regulatory power as reasons for federal intervention.

The paper identified major pollution hotspots — including New York City, the Houston/Austin corridor, Northeast Texas and the Illinois/Kentucky border — where it claims mining-attributable air pollutant concentrations were the highest, posing serious health risks.

The researchers warned that Bitcoin mining is a rapidly expanding, largely unregulated source of pollution with growing public health consequences, especially for downwind communities — risks that could intensify as demand for both crypto and AI data centers rises

'Deeply flawed'

In response, Daniel Batten, recognized as an expert in energy and environmental sustainability, particularly in the context of Bitcoin mining, told The Block the study is "deeply flawed," suggesting that it "looks like the conclusion was 'Bitcoin mining must look bad' then went looking to find data and methodologies that supported that."

Batten is a climate tech investor, a co-founder and Managing Partner at CH4 Capital and an advisory board member at the largest public Bitcoin mining firm MARA — which has a $4.4 billion market cap, according to The Block's MARA price page .

More specifically, Batten said the authors used marginal instead of standard emissions accounting, without disclosing its limitations.

The WattTime model used to estimate the marginal emissions of electricity use is a proprietary system developed by the nonprofit environmental tech group of the same name. It identifies which power plants ramp up or down in response to changes in electricity demand — like when a Bitcoin mine starts operating — so that emissions can be attributed more accurately to specific activities. WattTime's model is widely used, but its application to annualized, long-term emissions is contested — with Batten making the case that its proprietary nature meant it does not allow for academic scrutiny.

He also criticized the study's reference list as being comprised mostly of news stories and very few actual peer-reviewed papers, as well as "the fact they cherry pick 34 sites, while ignoring Bitcoin mining operations that don't support their thesis."

Furthermore, Batten alleged the study looked like the controversial New York Times article on Bitcoin mining from 2023, which was criticized by the industry over cherry-picking data, exaggerating fossil fuel usage and employing flawed methodologies. He suggested the article may be "trying to find a second life as an academic study," pointing out that NYT Deputy Investigations Editor Gabriel Dance, who wrote that piece, is also one of the study's co-authors .

"This article is a throwback to an early generation of academic work on Bitcoin that uses flawed methodologies and cherry-picked data — an approach that was debunked by Sai Vranken in 2023," Batten said. "It should not be taken seriously by policymakers and regulators."

The Digital Assets Research Institute (DARI) also published a formal rebuttal of the Harvard-led paper, finding similar issues to Batten.

DARI is an independent research organization established in 2022 that focuses on quantifying the environmental and social impacts of Bitcoin and other digital assets using data-driven methodologies. DARI's team comprises individuals with diverse backgrounds in academic research, policy analysis and business.

DARI criticized the study for its "selective use of data, a flawed method of attributing emissions, the inappropriate application of marginal emissions calculations and an over-reliance on news media and a small number of selectively chosen peer-reviewed studies that support a specific narrative" — exaggerating the impact of Bitcoin mining on air pollution, in the organization's view. "This is unfortunate, given the growing body of other, more balanced, research on the environmental costs and benefits of Bitcoin mining," DARI said.

Contrasting academic papers on Bitcoin mining

Batten argues there is not just an issue with this Harvard-led paper, referencing five other negative academic studies on Bitcoin mining since 2022, often receiving formal rebuttals over similar issues such as the use of non-contemporary datasets, discredited sources, citing news reports over academic articles and selection bias.

In contrast, he highlighted 20 positive peer-reviewed papers during the same period that received no rebuttals, criticizing Dominici and Guidi's paper for not building on such prior literature, arguing they ignored scientific consensus showing that Bitcoin has numerous environmental benefits. Those positive papers argue Bitcoin mining can support climate goals, bolster renewable energy and grid stability and help decarbonize power systems through demand response while demonstrating fundamental flaws in the earlier negative research.

"Bitcoin mining in recent years has become increasingly recognized as mostly powered by sustainable energy sources, and recognized in peer reviewed research (using recognized emissions accounting techniques) as accelerating the renewable energy transition (Lal et al, 2023) obviating the need for gas peaker plants (Bruno et al, 2023) stabilizing the grid (Rudd et al, 2023) halving the payback time for solar farms (Hakimi et al, 2024) accelerating renewable microgrid development (Moghimi et al 2024) reducing methane emissions from landfills (Sechrest et al, 2024) averting the need for expensive grid upgrades (Norris et al, 2025)," Batten said.

Batten also noted a line in the paper suggesting that most Bitcoin mining machines could be used for other compute purposes, arguing that it demonstrated a lack of awareness of prior scholarship. "Rhodes et al showed in 2021 that 'some other compute purposes' (normally HPC, AI data centers) cause more fossil fuel power on the grid, whereas Bitcoin mining data centers help decarbonize the grid," he said.

The paper and its rebuttals come amid a broader political moment, as President Trump pushes to turn the U.S. into a " Bitcoin mining powerhouse ." He has also recently signed an executive order to boost domestic coal production — an issue directly tied to the concerns raised by the Harvard-led study.

Earlier this month, the Trump administration froze over $2.2 billion in federal funding to Harvard and threatened to revoke its tax-exempt status after the university declined demands to limit campus activism, dismantle diversity programs and alter admissions and hiring practices. Harvard President Alan Garber asserted the university's commitment to academic freedom and constitutional rights.

The Block reached out to Dominici for comment.