Currently, many media outlets are complaining about Bitcoin’s high water consumption, while Bitcoiners are dismissing it as another round of misinformation. We look at how the estimate was created and the reasons for criticizing it.
Some media outlets are currently excitedly reporting that a Bitcoin transaction consumes as much water as fits in a swimming pool. So far too much, and that is of course a scandal.
Bitcoiners strongly disagree with this. Such reports are unscientific and defamatory; a misinformation and an attack, as ordered for the rally, because the high institutions are getting fed up.
Before you jump too quickly to whichever side you instinctively feel more comfortable with, let’s look at the sources. What speaks for and what speaks against the water thesis?
2,237 million liters for mining
The background is a paper that Alex de Vries published a few days ago Cell Reports Sustainability published. This paper describes “Bitcoin’s growing water footprint” in relatively dramatic terms.
De Vries, it should be noted, runs the Digiconomist portal, where he has been estimating the energy consumption of Bitcoin for a long time. His numbers are usually higher than others, such as Bitcoin Energy Consumption Index the University of Cambridge. Because de Vries works as a data analyst at the Dutch central bank, he is considered highly biased.
Nevertheless, de Vrie’s study should be viewed independently of his person. That’s what science is all about. So he claims that Bitcoin’s water consumption has “escalated” since 2020: it rose from 591 gigalitres (GL) – i.e. 591 million liters – to 2,237 GL in 2023.
How does he come up with that? That is the crucial question here.
How to determine the minig’s water footprint
First of all, 2.2 billion liters of water doesn’t sound like much. In Germany, 125 liters of water are consumed per capita per day, meaning that private households in Germany alone consume 3.6 trillion liters. In comparison, Bitcoin mining wouldn’t even be a mere crumb.
However, most of the water we use for showering, washing hands and cooking flows back into the tap. So it stays in the ground-level water system. On the other hand, when water evaporates or evaporates, it rises into the atmosphere, disrupting the cycle more. And this type of water consumption is what de Vries is concerned about.
During mining, water evaporates for two reasons: directly, to cool the miners, and indirectly through the electricity consumed. Coal, gas and nuclear power require large amounts of water for cooling.
To measure indirect consumption, de Vries looks at electricity consumption and the geographical distribution of mining. His source for this is the Cambridge Center for Alternative Finance, which has been collecting these values for a long time. He combines this data with the water intensity of electricity generation at the respective location, which is given in liters per kilowatt hour and depends on the local electricity mix.
Since 2020, the water intensity of the miners’ electricity has increased enormously. After the ban in China, many miners moved to Kazakhstan, away from the dams and towards the coal and gas power plants. With this move, not only the CO2 footprint increased, but also the water consumption per kilowatt hour from 8.63 to 15.76 liters at the end of 2021.
After 2021, the Bitcoin hash rate rose – and with it electricity consumption – by an estimated 35 percent. Since the geography of mining has remained largely stable, de Vries assumes that water consumption has increased to the same extent. According to his calculations, in 2022 it would have reached a total of 2,237 gigalitres. Translated into transactions, this amounted to 16,279 liters in 2021 – roughly a garden pool with 16 cubic meters.
To estimate the direct consumption of cooling miners, de Vries consults studies on water consumption by data centers in various climate zones in the United States. He then transfers this to the US miners and then extrapolates the value globally. Using this somewhat vague method, he ends up with a total estimate of 84.9 gigalitres. In the overall picture, this is rather negligible.
Given the threat of a global water shortage, de Vries is concerned about the overall high water consumption of miners. But he also makes suggestions on how miners can reduce it: They can settle in locations where mining requires less water – for example in cold regions with renewable energies – and they can cool with immersion oil instead of water.
In the Bitcoin community, however, you shouldn’t even get to this point. She reacts at an earlier point with a defensive reflex.
“Too high by at least a factor of 1,000”
Daniel Batten, for example, makes a vehement criticism. Batten is, so to speak, the opposite pole to de Vries. Where there is doubt that the ecological damage of mining is too high, Batten is not afraid to reinterpret Bitcoin as a climate-friendly technology.
On Batcoinz, Batten regularly publishes articles that, although they give off a strong scent of “greenwashing,” definitely make a contribution to the matter. He reacts to de Vries’ study with outrage: it is misinformation and an attack; The numbers are too high by a factor of 1,000, if not 100 million. He even writes these numbers in bold letters.
But how does he come up with that?
On the one hand, Batten complains that de Vries uses an inappropriate metric. Attributing the consumption of electricity or water to transactions is actually nonsense whose sole purpose is to create provocative headlines.
Because mining and the number of transactions are inherently independent of each other. It would have made more sense to convert it to a dollar market capitalization because these two values are logically related, which would show that Bitcoin is relatively environmentally friendly compared to stocks or gold.
Above all, Batten complains that the metric ignores off-chain transactions, for example through Lightning. There can potentially be millions of Lightning transactions behind every Bitcoin transaction.
As nonsensical as the “liters per transaction” metric is, its role in de Vries’ paper is unorganized. He mentions them only once, while otherwise focusing on estimating global consumption. A criticism that hinges on this value may sound cheap, but it misses the point.
Batten also mentions that de Vires has a history of incorrect and biased estimates. For example, he predicted that Bitcoin would consume all the world’s electricity by 2020, which obviously didn’t happen and was off by something like a factor of 2,500.
Batten does not provide any further substantive arguments. His accusation that de Vries is too high by a factor of 1,000 to 100 million is apparently based solely on the metric “per transaction” – and even there it is far-fetched. Because according to everything we know, there are currently not 1000 Lightning transactions per onchain transaction, but rather 10 to 100 onchain transactions per Lightning transaction.
Batten therefore fails to attack de Vries’ methodology and its essential results. Instead, he works on his person and a metric that is irrelevant to the matter.
Bitcoin solves the problem of water scarcity – theoretically
In another post, Batten tries to show that Bitcoin not only does not exacerbate the problem of water scarcity – but can actually solve it. Mining helps to expand renewable energies more quickly, which means that more energy-intensive sea desalination plants can be built – a good argument, especially for the Arab and North African countries, which are among the most threatened by water shortages.
In addition, the waste heat generated during mining can be passed on to the desalination plants, making their operation cheaper. Desalination plants, photovoltaic cells and Bitcoin miners on the coasts could work together to produce not only drinking water, but also money.
The argument may be plausible in theory. In practice, however, it has not yet been implemented through individual pilot projects. They do not in any way reduce the water footprint that de Vries estimates.
So it would be better if miners listened to de Vries at this point and took measures to reduce water consumption. Since these are also measures to decarbonize mining, this is necessary anyway.