In September, the Ethereum cryptocurrency network managed to pull off what it had been promising to do for years: switch to a model of mining that is far less emissions-intensive than the standard, a move known as the Merge. Did the switch do anything for the currency’s emissions? Could it be a good solution to the huge climate issues plaguing crypto?
The preliminary answer, according to a peer-reviewed opinion paper published Tuesday in the journal Patterns, is, hesitantly, yes. The total energy demand involved with the Ethereum network, the paper estimates, went down by more than 99% after the Merge, and could represent a reduction in energy demands as big as some medium-size countries — an enormous figure.
A brief refresher, first, on what exactly these terms mean. Traditional crypto mining operates on what’s known as a proof-of-work model: constantly running machines in order to race to solve an equation. In this model, the more computing power a miner has, the better chance they have of solving the equation and earning the reward, which incentivizes mining operations to run as many machines as possible 24/7 — and creates an enormous amount of associated carbon emissions. By contrast, proof-of-stake, which Ethereum adopted in September, operates more like a lottery system, meaning miners don’t need all those energy-guzzling machines.
The new paper, authored by Alex de Vries of the cryptocurrency blog Digiconomist, contains “the best and worst case scenarios before and after the Merge,” he told Earther in an email. Estimating the exact energy use of any cryptocurrency network is not an exact science, but it’s possible to estimate a range of potential energy uses, based on the number of devices that are connected to the network.
“The best and worst case scenarios before and after the Merge…can actually be estimated quite easily with a very high degree of certainty,” de Vries said. A best-case scenario before the Merge would be one where the machines on the network are running as efficiently as possible, resulting in lower energy use; meanwhile, a worst-case scenario would involve those machines being run to “the maximum amount miners could possibly afford if they were spending all their money on electricity.” De Vries used similar calculations for after the Merge.
The Ethereum network “achieved a tremendous reduction in electricity use; the big challenge is trying to determine whether this is also reflected from a global point of view,” de Vries said. Looking at the range of possibilities contained in these different scenarios, it’s clear that the Merge had a significant impact on the Ethereum network’s energy use either way. Even when going from the total best-case scenario before the Merge — assuming all the devices on the network were running as efficiently as possible — to the worst case after, the energy reduction is still 99.84%, the paper found. Meanwhile, going from the worst case to best case represents a more than 99.99% reduction. In real-world terms, that means that the Merge could possibly have reduced the Ethereum network’s power use by as much as the electrical demand of the entirety of Austria. (This best-case scenario, de Vries said, “seems unlikely,” but we simply can’t know for sure.)
There are a couple important caveats to consider when thinking about this potential reduction and the maths involved. First off, it’s not possible to know what happens to all those energy-sucking machines after the Merge — just that they’re not connected to the Ethereum network any more. The fact that the network has shifted to a process that requires significantly fewer machines doesn’t mean that miners who have invested money in countless machines are simply going to throw them in the trash and end their energy use.
So what are these machines being used for? Ethereum has already split off into two different forms of Ethereum-based currencies, Ethereum Classic and Ethereum POW, which, while not as profitable as Ethereum to mine, are still proof-of-work based and could theoretically provide ways for miners to keep using their expensive equipment. Graphics cards previously used to mine Ethereum could also be used to mine smaller cryptocurrencies or be repurposed outside of crypto mining, for processes like gaming or cloud computing. While these uses represent significantly less energy use than running mining equipment 24/7, they would still impact the overall global emissions reductions of the Merge.
More importantly in terms of energy use, while Ethereum miners would not be able to use their devices to mine Bitcoin — some mining machines can only be used to mine Ethereum, while others couldn’t compete at a profitable level with Bitcoin — suddenly nuking a handful of machines from service frees up server space that can be taken up with other devices miners may have been keeping offline, meaning that more Bitcoin machines can get started on mining dirty crypto.
“What likely did happen is that a bunch of previously unused Bitcoin mining devices have taken the data centre space Ethereum miners were previously using (rack space was in short supply),” de Vries said. “This probably explains the jump in the Bitcoin network’s computational power after the Merge too.”
Despite these caveats, it looks from this first analysis that the Merge was a success in significantly reducing the emissions associated with the Ethereum network. The real challenge will be finding other ways to encourage other cryptocurrency networks to follow suit — especially Bitcoin, which is far and away the largest source of emissions from crypto mining. As we covered when the Merge happened, there are lots of factors that are getting in the way of the Bitcoin network undergoing a similar shift to a proof-of-stake model. But that doesn’t mean it’s not worth it to try — particularly if governments want to step in and give regulation a shot.
“What we can say is that if Bitcoin was to make a similar change this would certainly be more noticeable from a global perspective,” de Vries said. “With the Bitcoin network still responsible for as much electricity consumption of half of all global data centres combined (also as much electricity consumption as my home country the Netherlands), this is absolutely worth pursuing.”