A recent New York Times article argues that Bitcoin mining hurts people and the planet. Here's what they miss.
The most recent New York Times bitcoin exposé—“The Real World Costs of the Digital Race for Bitcoin,”—is a morality tale. Its main character? Bitcoin mining. Its lesson? No good can come from a malign and irredeemable presence like bitcoin mining.
We set the stage with a murder scene: a rare winter storm in Texas has triggered massive power outages and grid failure, leading to scores of deaths. Next, we introduce our misfeasor: bitcoin mining data centers, those energy guzzling computers racing to score more bitcoin in an endlessly repeating lottery. The bitcoin miners whine with robotic indifference while human beings literally freeze to death for lack of electricity, until at last the grid operators force them to power down. The implication, though not explicitly stated: this industry kills people.
The remaining tale unfolds with an impressive paper trail of evidence to round out the character sketch. Bitcoin mining is portrayed as a grift, exploiting grid programs for obscene profits while using precious resources and raising power prices for everyone else. It is shown to be an epic polluter that hides and downplays its staggering carbon impacts. The paper trail is laden with data and scientific-sounding claims. How could one take issue with The Science™?
But alas, the facts do not add up. The opening scene is a misleading setup. The character sketch is partial, and highly skewed. The paper trail is devised with unpublished proprietary methodologies. The metrics are generally accepted, but wrongly applied to a narrow context. In other words, the scientific appearance lacks the rigor of actual science. The truth is more complex, and far more positive.
In reality, we know what happened during Winter Storm Uri. It wasn’t that bitcoin miners were stealing power from residential communities. The cold weather froze water in gas and oil wells, which in turn caused gas production to plummet. 38 gas plants shut down or reduced production. Many gas plants, plus coal, nuclear, and some wind power plants, were affected. In response to the 40-degree below average weather, power demand surged.
According to EnvironmentAmerica, a computer glitch led to a massive price spike as a result of the grid’s desperate attempt to entice generation to come back online. Meanwhile, bitcoin miners operating in Texas’ demand response programs, were told to curtail their power, and did so on cue.
Demand response isn’t unique to bitcoin miners. Though, after reading the New York Times piece, one might think that was the case. Any kind of large load can participate in demand response and get paid for doing so. For example, universities sign up for such programs with the expectation that they might have to reduce consumption anywhere from zero to two times a year. Providing 5 megawatts of load as grid insurance could earn a participant an annual profit of $185,000 according to Enel X’s profit calculator.
Indeed, in the most recent report of demand response, the Electric Reliability Council of Texas (ERCOT) reported a 2021 average Responsive Reserve Service (RRS) payment of $208.28 per megawatt per hour, the same year that Winter Storm Uri impacted the Texan grid. Ultimately, ERCOT made nearly 2.6 billion dollars worth of payments that year. Payments made to Bitdeer, the mining company that was highlighted in the New York Times' opening salvo, accounted for 0.69% of total RRS payments that year.
The New York Times report does not make any mention of how demand response works in Texas. Instead, a strange twist, the report appears to demonize flexible load, despite a well understood need for more of it to meet the unique requirements of an electrical grid with high levels of solar and wind generation. The IEA forecasts that by 2025, renewable capacity will overtake gas and coal and that by 2030, all this additional variable generation will require more than 12 times as much demand response from 2020 levels.
According to the New York Times, it’s just not fair that miners can power up and down almost instantly, making them so valuable to the grid that ERCOT is willing to pay them millions of dollars each year just to make their flexible load available for emergencies like Winter Storm Uri. It’s not fair that a hospital can’t benefit from the same program. The truth is, of course, that a flexible load should shut down in a crisis so that hospitals don’t have to.
What about the data? The evidence seems irrefutable. Two studies were commissioned, one from Wood MacKenzie, an energy research and consulting company, and another from WattTime, a non-profit tech company. The results claim to show that bitcoin’s grid demand is damaging to electricity ratepayers and the planet. The Times reported increased electricity prices across the state and claimed that the largest mining companies are running on 78-99% fossil fuels, producing millions of tons of carbon dioxide as a result.
It is difficult to evaluate these claims because the Times does not explain the methodology behind these studies. We have no idea how bitcoin miners were modeled beyond assuming a 95% uptime. One thing we do know is that WattTime uses marginal emissions accounting to measure bitcoin miners’ greenhouse gas emissions footprint.
Marginal emissions are used to evaluate short-term effects on grid emissions when a change in demand occurs. Marginal emissions accounting tells you what kind of generation is available to meet that additional demand. It doesn’t matter if the change in demand is a result of electric vehicles, batteries, heat pumps, or bitcoin miners. One would apply the evaluation in the exact same way.
Marginal emissions should not be confused with average grid emissions. Marginal emissions count the emissions of the generator(s) that would need to come online to meet the increased demand. Average grid emissions are based on the average grid generation mix which assumes that the electricity generated to meet demand will be a combination of all available generation.
In a grid where a response to a change in demand will likely be met with fossil fuels like natural gas (which is typical of most electrical grids in the United States), one would intuitively expect marginal emissions to be higher than average grid emissions. In fact, when studying the effect of energy storage on short-term grid emissions, multiple research groups have found that when storage relies solely on price arbitrage, they increase marginal emissions rather than reduce them. Another study found that adding a significant load of heat pumps required more natural gas in the first few years to meet this demand increase. Both energy storage and heat pumps are part of the plan to achieve decarbonization goals and reduce long-term emissions. Yet, they have the same kind of high marginal impacts on carbon emissions as bitcoin miners.
Looking at the broader context, it should be clear that an electrical grid that has yet to add constraints on carbon intensity of generation will have high emission factors that affect all kinds of loads that create a change in demand. In the long term, as countries work toward meeting emissions reductions, we expect that marginal emissions will go down. In the meantime, bitcoin miners, batteries, heat pumps, electric cars, and other new demand, will all have high marginal emissions when price is the only incentive.
Therefore, claiming that bitcoin mining is bad for the planet on marginal emissions alone shows that The Times does not understand how to apply greenhouse gas emissions accounting. While it is true that it is preferable that new demand bring its own renewable generation with it, this is only available to the largest and wealthiest of corporations like Google and Meta. If we think that this standard is the only way to achieve emissions reductions, then the ultimate conclusion is that there should be no increase in demand ever again unless one can afford their own solar farm. This is not the world we live in.
The New York Times does not add any of these contexts to its report, but instead, uses carbon accounting metrics to create a false sense of certainty that there can be no debate about it: Bitcoin miners are destroying the planet. The science may be settled on climate change and the risks from failing to meet this challenge, but it is definitely not settled on the effects of bitcoin miners as highly flexible loads that provide demand response.
To answer these questions, we need bitcoin mining companies to step up. We need miners to be transparent. Bitcoin miners voluntarily reporting their emissions, uptime, demand response participation–whether in official markets or as price response–will go a long way to dispel the misconceptions and fear that cloud the industry’s reputation.
Despite the lack of available data, a closer look at the industry reveals a more nuanced story than what the Times reported. An independent researcher recently identified over 30 United States-based sustainable bitcoin mining companies, accounting for over 14% of global bitcoin mining operations. When it comes to methane mitigation, bitcoin miners are now looking toward capturing landfill gas and burning it which reduces near-term warming. There are now three companies developing products to incentivize sustainable bitcoin mining. Finally, the energy industry is beginning to see the value of bitcoin mining as a floor for revenue and we are now seeing the co-location of miners with wind and solar power generation.
The New York Times has revealed itself to be more concerned with spinning morality tales disguised as investigative reporting than fairly representing the truth. The reality is that Bitcoin miners play a valuable role in improving grid resilience and reducing carbon emissions. Rather than demonize the mining industry, we should work to understand how this unique consumer of power can be leveraged in the fight against climate change. To win that fight, we must keep an open mind toward all available solutions, including bitcoin.