Energy Consumption

A lot of people criticize Bitcoin for its environmental impact. Is that merited? And why are some blockchains “greener” than others?

An article published in Nature Climate Change 2018 claimed that if Bitcoin was adopted on a large scale, the emissions caused by it alone could lead to global warming of more than 2◦C in the next three decades. Notwithstanding the questionable assumptions the article might have made, it aligns with the seemingly endless criticism that particularly Bitcoin faces from environmentalists. Nevertheless, the outsized energy consumption of Bitcoin and similar blockchains cannot be denied. Therefore, this lesson will cover:

• Why Bitcoin and blockchains like it use so much energy
• Why this energy consumption poses a problem
• Which type of blockchain uses vastly less energy and why
• What we can do about the energy consumption of Bitcoin

Why Bitcoin is so energy-hungry

Let's start with some numbers first: Bitcoin consumes anywhere between 60 terawatt-hours (TWh) and 500 terawatt-hours per year, depending on the estimate you use. To put this into context: that could be the yearly electricity consumption of a country like Sri Lanka (lower end) or twice the consumption of the UK (upper end).

The conundrum of estimating Bitcoin's electricity usage is that there is no definite method to do so. There is a wide range of estimates across different studies; however, the most reliable and repeating figures put the 2021 electricity consumption of Bitcoin at around 120-130 TWh. That is roughly the yearly consumption of a country like Sweden or Argentina.

To understand why Bitcoin requires so much electricity in the first place, recall that new coins need to be mined. That happens by special computer chips solving a complex mathematical puzzle called a hash function. This process requires a lot of computational power and can only be solved by brute force, which requires operating entire halls full of these computer chips. Solving such a puzzle generates new coins, and since this process is easy to verify but hard to conduct, it is called proof-of-work (because other parties can easily verify the computational work that had to be done).

This proof-of-work mechanism is the root cause for Bitcoin's (and to a lesser degree Ethereum's) row with environmentalists. Not only Bitcoin consumes copious amounts of electricity, though. According to estimates, Ethereum and other, less popular, proof-of-work blockchains add another 50% to Bitcoin's electricity bill. However, electricity consumption in itself is not problematic. It's the energy sources this electricity is generated from where Bitcoin gets into trouble.

Why Bitcoin's energy consumption is a problem

Carbon footprint

It's estimated that, although three out of four Bitcoin miners use renewable sources of energy, about 61% of the energy used to mine Bitcoin comes from non-renewables. Bitcoin miners are perfectly rational actors, so they will set up their mining rigs wherever electricity is cheapest. And what is the cheapest source of energy you can generally find? Energy made from fossil fuels.

That's why more than half of all Bitcoin mining measured by computational power takes place in China, where electricity is cheap. Chinese Bitcoin miners use two energy sources to generate electricity: hydro-energy in the summer and coal in the winter. That helps explain why Bitcoin has a carbon footprint of about 70 to 90 million metric tons of CO² - as much as Israel or Belgium.

E-waste

Another problem is electronic waste Bitcoin, and other proof-of-work blockchains generate. Because mining is competitive and miners benefit from economies of scale, improving hardware is necessary to operate profitably. As with all computer processors, the distinct processors used for Bitcoin mining have an economic lifetime of about 18 months before being replaced by significantly faster versions. Replacing processors every 18 months generates an enormous amount of electronic waste, to be precise nearly as much as the country of Luxemburg does every year (6.76 kilotonnes).

Economic incentives

The problem gets compounded by the particular economic incentives miners have. Let's assume Bitcoin miners are perfectly rational and will mine Bitcoin as long as the price of one coin covers their costs. If the electricity price drops below this equilibrium, new miners will join the game. If it rises above it, unprofitable miners will exit it. Since Bitcoin mining is competitive and open, anyone can join, and the output is, by and large, dependent on the price of electricity and the price of Bitcoin. If electricity is cheaper, miners will join, as they will if the price of Bitcoin rises (and vice versa).

Therefore, a higher Bitcoin price drives more mining, which, in turn, increases the economic cost. Since Bitcoin's scalability is limited, how much the network is actually used is irrelevant for mining. Scaling Bitcoin would only increase the amount of data generated but not impact electricity consumption because coins would still need to be mined.

Arguably, even compared to mining gold, Bitcoin's energy record looks poor. While you can stop mining gold and continue using circulating supply, you couldn't do the same with Bitcoin because an end to mining would also mean an end to validating transactions and, thus, the network as a whole.

Hence, the only possible change to limit the environmental impact would be a switch from proof-of-work to another mechanism of validating transactions.

Why other blockchains use much less energy

The alternative to proof-of-work would be using proof-of-stake. The validation of transactions would be linked to staked capital on the blockchain, which serves as a security to ensure the correct behavior of the validator. The only energy requirement, in this case, would be running a commercially available computer. Since proof-of-stake blockchains are easier to scale, Ethereum plans to switch to this mechanism. Consequently, transaction numbers would rise, but the amount of energy used to run the blockchain would plummet. Ethereum visualized this nicely in a blog post on their website:

In other words, proof-of-stake reduces electricity consumption by 99.95%, or from the size of a country to that of a house.

So, why don't all blockchains use proof-of-stake?

Firstly, Bitcoin was not designed to be a massively popular payment solution. Remember that price drives energy consumption, so in a world where Bitcoin costs less, its environmental impact is also negligible. Ethereum, on the other hand, was always designed with an eventual switch to proof-of-stake in mind (why it wasn't launched as such is another issue, though).

Second, almost all "new" blockchains use some form of proof-of-stake but simply haven't been able to catch up with the popularity of Bitcoin and Ethereum. Blockchain purists still doubt the security of proof-of-stake when used at scale, as well as the fact that not everyone can be a validator due to minimum staking thresholds. On the other hand, economies of scale and electricity costs lead to geographical and computational centralization in proof-of-work blockchains, meaning they're far from perfect either, apart from their energy inefficiency.

Solutions to "dirty" blockchains

Even blockchains that do not mine coins require energy for validating blocks and updating databases. This kind of node maintenance is inherent to a blockchain's decentralized structure. The more nodes a blockchain has, the more decentralized it is but also, the more redundant operations it needs to perform. This redundancy is significant in comparison to centralized structures:

Traditional payment systems like credit cards are more energy-efficient than even the most energy-efficient current proof-of-stake blockchains. This is not necessarily a bug, but a feature since a blockchain does away with a central point of failure at the expense of using more electricity.

Scaling solutions

Scaling solutions should greatly help reducing blockchains' electricity bills. If not all nodes need to validate all transactions, this reduces redundant transactions. Less on-chain workflows do so as well.

For instance, imagine a scaling solution that limits the transactions performed on the blockchain to receipts instead of all transactions. Instead of Alice sending Bob ten times 1 ETH and recording it all on-chain, the blockchain would only record the final receipt of 10 ETH transferred. Another blockchain would do the rest. Studies estimate that such scaling solutions can reduce electricity costs on proof-of-stake blockchains by 98.5%.

Policies

Technological solutions are one piece of the puzzle, but proof-of-work blockchains will likely continue to meet with resistance from policymakers and environmental protection groups. Measures like partial or complete mining bans are unlikely to have a lasting impact. After China cracked down on miners to work towards its goal of zero net carbon emissions, mining rigs started moving to countries like Kazakhstan that offer cheap fossil fuels. Barring a concentrated global effort, bans will only push miners to other locations.

Incentivizing the use of renewable energies would be a solution, and countries like Iceland have successfully attracted miners that way. However, depending on the energy source, this can create new problems. In the case of Iceland, miners' energy demand may soon outstrip the country's supply. Miners in China that use hydroelectricity are dependent on the cyclical nature of their energy source, whereas mining requires a steady stream of electricity that renewables often cannot guarantee.

Carbon taxes and forcing investors to carry Bitcoin's carbon footprint costs may be another vector of attack. Still, it is difficult to see how imposing regulation on a deregulated and decentralized industry is supposed to work on a global scale.

Possibly the most realistic options, barring a full-on ban of proof-of-work blockchains, are a massive collapse of energy costs due to a technological breakthrough (unlikely in the short term), and/or constant improvements in the efficiency of processors and electricity allocation. A globally more equal distribution of miners due to economic incentives could at least slow the growth in electricity consumption. Also, a proof-of-stake blockchain successfully scaling and being mass-adopted could inflict lasting damage on the Bitcoin price.

Conclusion

Bitcoin, and Ethereum in its current form, are undoubtedly economically highly inefficient versions of new technology. Although the doomsday prophecies of their environmental impact may be exaggerated in the long term, it's hard to see them survive in their current form.