How NOT to report on the carbon footprint of bitcoin and NFTs

Recently, on two separate occasions, I was contacted by journalists asking me for comment on the environmental implications of Proof of Work blockchains. Both news outlets opted not to run the story after I pointed out flaws in the sources they were referring to.

In this article I outline why I think media commentators need to be cautious in their reporting on this issue (expanding on this tweet). In a follow-up post I discuss how we might get better insights into the environmental costs of blockchain by investigating geographical context, technological design, and the practices of miners.

Galaxiss (NFT) by Xenoliss. Licence: CC BY-NC 4.0

News request #1: Bitcoin mining contributes the same greenhouse gas emissions as Italy or the Czech Republic

On 7th of April (AEST), Nature Communications published an article claiming bitcoin mining in China was threatening global sustainability efforts. Hours after Nature’s press release went out, a journalist asked if I had seen it. When I said I had not, she offered me an hour to read the article, at which point she would call me back.

The article states that by 2024, Bitcoin mining will exceed the entire annual greenhouse gas emissions outputs of an entire mid-sized country in Europe. My concern was not so much with the model the authors presented in the article, but the data they had run through it and their subsequent conclusions. Instead of providing the reporter with a neat quote supporting the article, I gave her a list of reasons why their predictions could be wrong:

1. We cannot know the environment impact of bitcoin unless we know what energy sources are being used. The authors of the paper admit that carbon emissions of Bitcoin mining depends on the source which is used for its generation, yet their model calculates energy from regional estimates only. If we assume that bitcoin miners are in it for profit, then there is a strong likelihood that they are using the cheapest energy available at the time. That energy is sometimes surplus energy that gets used at the source. For instance, a Sydney-based company called Cosmos Capital (not the blockchain company) is working to expand its bitcoin mining operations to rural Australia and plans to use latent or excess renewable energy that utility companies need to dispose of to prevent grid surges.

2. We’d need details on where mining is occurring in order to know if renewable energy is being used. Bitcoin miners are notoriously secretive and are likely to use VPNs to obfuscate their location. Cambridge University’s CBECI attempts to map mining activity using data provided by mining pools between Sep 2019 and April 2020. Their estimates are extrapolated from just over a third of the hashrate. As Nic Carter points out, while the CBECI map is incomplete and out of date, it suggests that a great deal of activity may be occurring in provinces that produce excess hydro power.

3. We have anecdotal evidence that bitcoin miners use renewables at least some of the time. Mustafa Yilham from Bixin group, which claims to manage 2.5% of the entire global Bitcoin mining operations, says that miners migrate to different regions based on the season. In summer Bixin uses 100% renewables in the form of hydro. During winter, when there is not enough rain, miners move to regions such as XinJiang and Inner Mongolia where they use power from the grid. Yilham claims that they try to choose areas where the grid itself has high renewable energy use.

Norwegian oil company Aker has started a new company to mine bitcoin exclusively from renewables called Seetee, which they say will help the renewables industry:

The financiers of mining operations will insist on using the cheapest energy and so by definition it will be electricity that has no better economic use. Bitcoin then acts like an economic battery. What otherwise was of little value locally, is turned into an economic asset that can be used globally. Extremely flexible demand from miners can optimise the local supply and demand for electricity, which may accelerate the energy transition by improving the economics for new renewable projects.

There are plenty of other examples out there.

4. Bitcoin mining can help clean up dirty activities. As this video shows, bitcoin is being used by some industrial plants to reduce gas flares (the burning of excess gas for safety or maintenance reasons). Converting excess gas into electricity to mine bitcoin reduces the amount of greenhouse emissions.

While the examples mentioned above don’t give us a complete picture, they should make us pause before asserting that bitcoin will raise the earth’s temperature by 2 degrees. Moreover, the article by Jiang et al. in Nature Communications argues that their calculations provide evidence that China should commit to site regulation of bitcoin operations — in other words, increasing state control of cryptocurrency infrastructures.

Reporter #2: An NFT would drive 500 miles

The second reporter wanted to talk about the environmental cost of NFTs. She had seen an estimate that the environmental cost of minting a single NFT to be the same as driving more than 500 miles in a gasoline powered car. The problem with this calculation is that one NFT doesn’t directly translate to an increase in hash power for the Ethereum blockchain.

The environmental impact of Ethereum depends on the number of miners who are competing to mine blocks, just like bitcoin. We don’t need more miners when transactions increase, or when gas fees are higher. I used the analogy of a Melbourne tram that will run regardless of how many passengers are on it.

There’s an argument to say that congestion will push up gas prices, which might incentivise more miners to get involved (or the same miners to run more hardware to improve their chances of receiving mining rewards in the form of ETH). However, NFTs make up only a small portion of the total Ethereum transactions, so boycotting NFTs is unlikely to lead to a reduction in miners (despite the hype). In addition, some NFT-related platforms such as Illuvium.io are using layer 2 solutions, effectively batching NFT transactions so that they use less gas.

The journalist seemed to be following as I rattled on about trams. However, when I mentioned that Ethereum has begun the transition to Proof of Stake, which will reduce the carbon footprint of the blockchain by 99%, her conviction in the story began to waver. She said it was all very technical and wasn’t sure how she’d edit it into a story for prime-time radio.

End of story?

Negating climate change matters a lot to me and I welcome press coverage of good research in general. While some will argue that any use of fossils fuels in the maintenance of blockchains is unjustified, there’s a chance we might need technologies that allow us to track production and consumption of resources and goods (the so-called circular economy) to combat climate change. Secure public blockchains enable us to do that in ways that cannot be manipulated by corporations and governments. I have more to say about this in Part 2.

Further reading

If these points seem familiar to crypto readers it’s because I’m not the first to make them.

Nic Carter has written increasingly exasperated articles on this for Coindesk here and here.

The non-profit Coincentre has produced an excellent review of research articles on the topic.

A 2019 Coinshares Report estimated the use of renewable energy in bitcoin mining to be 74% with +/- 5% ‘intuitive’ uncertainty. They provide the following caveat: “[W]hile we do our utmost to accurately pinpoint the location of global mining centres, the Bitcoin mining industry remains a highly private and secretive industry. As a result, our estimates may be subject to significant potential uncertainty. We believe that our database of global mining centres has around 70% total visibility on the overall market, and we cross-reference all our model results against this database in order to quality-check our assumptions”. They also note the nomadic nature of some bitcoin mining operations.

China’s renewables industry appears to be producing a lot of curtailment (attractive to bitcoin miners for reasons stated above).

ArtNews published a good article on the carbon footprints of NFTs.

Ethereum.org has a useful plain language summary of the computational work involved in minting an NFT, why securing the blockchain matters, and how replacing mining with staking will make this entire debate redundant.

About me: I am a Professor at RMIT, working across the RMIT Blockchain Innovation Hub and the Digital Ethnography Research Centre. I am also an Associate Investigator (AI 🤖) of the ARC Centre of Excellence for Automated Decision-Making and Society. I acknowledge the support of the Australian Research Council, FT19010372.

I use Medium for work-in-progress communication and may change my views and analysis as the work progresses. I welcome constructive and respectful feedback from blockchain practitioners. DM me on Twitter @elinorrennie or email ellie.rennie@rmit.edu.au

Professor at RMIT University, Melbourne. Australian Research Council Future Fellow 2020–2025: “Cooperation Through Code” (FT190100372) Twitter: @elinorrennie