The skyrocketing value of Bitcoin is leading to soaring energy consumption. According to one widely cited website that tracks the subject, the Bitcoin network is consuming power at an annual rate of 32TWh—about as much as Denmark. By the site’s calculations, each Bitcoin transaction consumes 250kWh, enough to power homes for nine days.
Naturally, this is leading to concerns about sustainability. Eric Holthaus, a writer for Grist, projects that, at current growth rates, the Bitcoin network will “use as much electricity as the entire world does today” by early 2020. “This is an unsustainable trajectory,” he writes.
Global energy production obviously can’t double in two years, and it would be an environmental disaster if it did. Fortunately, while the Bitcoin network consumes a ridiculous amount of energy, particularly on a per-transaction basis, the situation isn’t as dire as critics like Holthaus claim.
Bitcoin’s energy consumption won’t necessarily march steadily upward. Indeed, Bitcoin’s energy consumption is designed to fall in the long run. And Bitcoin’s energy consumption isn’t tied to the number of transactions the network handles. That means that increasing use of the network won’t necessarily impose a high environmental cost.
Bitcoin mining—the process that generates new bitcoins while maintaining the network’s shared transaction ledger—is a secretive global industry. No one knows exactly how much energy it consumes.
However, we can make some educated guesses. For starters, we know the industry’s revenue: Bitcoin miners currently generate 75 bitcoins per hour, which, at the current price of around $12,500 per bitcoin, translates to $937,500 per hour, or more than $8 billion per year.
Moreover, the industry is highly competitive, and electricity is one of its biggest costs. So when the price of bitcoins rises, we can expect miners to spend more and more on electricity until electricity costs are roughly on par with revenues.
This is the methodology the Digiconomist website uses to estimate the Bitcoin network’s energy consumption. It assumes that the industry will spend 60 percent of its revenue on electricity and then extrapolates from the current bitcoin price and prevailing electricity prices. It finds that the network is consuming energy at an annual rate of 32TWh.
It also assumes that the network takes time to adjust to big price increases like we’ve seen in recent days. This means that, if Bitcoin stays above $12,000, we can expect this figure to rise further in the coming weeks.
Will the network’s energy consumption continue to rise over the longer run? Under Bitcoin’s current design, this depends entirely on what happens to the price of Bitcoin. If Bitcoin’s price doubles to $25,000, we can expect the Bitcoin network’s energy consumption to roughly double as well. If Bitcoin’s price falls significantly, on the other hand, miners will find their operations unprofitable and will start to switch off their least efficient equipment, causing energy use to decline.
Right now, Digiconomist estimates that Bitcoin is consuming less than 1 percent as much energy as the US economy. This means that, for Bitcoin’s energy consumption to exceed that of the United States, Bitcoin’s price would have to rise by roughly 100-fold to more than $1 million.
Could that happen before 2020? It doesn’t seem likely. Of course, in early 2015, Bitcoin was worth only $200—hardly anyone expected a 50-fold increase over the last two years. But here we are.
There’s a widespread misconception that Bitcoin mining is based on a mathematical process that gets steadily harder as more and more bitcoins are produced.
That’s wrong. The Bitcoin network is designed to automatically adjust the difficulty of mining to ensure that one block is produced every 10 minutes, no matter how much (or how little) computing power there is on the network.
When Bitcoin launched in 2009, each block came with a 50-bitcoin reward for the miner who created it. This figure is scheduled to fall by half every four years. It fell to 25 bitcoins in 2012 and 12.5 bitcoins in 2016. The reward will fall again to 6.25 bitcoins in 2020. When the mining industry’s revenue falls by half, its energy consumption should fall by the same proportion, since, if it didn’t fall, mining would become an unprofitable activity.
The reward halves again in 2024, in 2028, and every four years after that. So, if the price of bitcoins stabilizes, the Bitcoin network’s energy consumption will steadily fall over the coming decades.
Another important point: that fixed 12.5 bitcoin reward doesn’t depend on the number of transactions the Bitcoin network processes. Miners do also collect per-transaction fees from Bitcoin users, but those are currently much smaller than the fixed per-block reward.
This means that the Bitcoin network could easily be upgraded to handle more transactions—potentially a lot more—without significantly changing miner revenues or energy consumption. So it’s not the case that a growing Bitcoin network will necessarily lead to a growing environmental disaster.
On the other hand, growing use of the network could push up Bitcoin’s price, which in turn would increase energy use.
While Bitcoin may not be a total environmental disaster, the Earth would certainly be a greener place if the Bitcoin network didn’t consume so much electricity to process a relatively small number of transactions.
There are basically three ways this could happen. One way, as we’ve already discussed, is for Bitcoin’s price to decline.
A second option would be to shrink the network’s 12.5 bitcoin-per-block reward sooner than the scheduled 2020 reduction. But that’s easier said than done. Bitcoin mining companies are not going to go along with this willingly, and Bitcoin traditionalists are likely to oppose such a move as well.
Governments may also be powerless here. If any one country tries to force a change, mining operations would simply flee to another jurisdiction. Changing Bitcoin by regulatory fiat would require a coordinated global regulatory effort, which doesn’t seem likely to happen any time soon.
A third option would be to change the Bitcoin mining process altogether. Bitcoin’s current mining algorithm is based on computing a supermassive number of cryptographic hash functions. But other cryptocurrencies have been exploring alternatives. Bitcoin Gold is a recently created variant of Bitcoin that uses a “memory-hard” mining algorithm that might prove to be less power hungry—though it would still consume huge amounts of juice. More exotic mining algorithms exist that could dramatically reduce power consumption.
Switching to an alternative mining algorithm would also be controversial among traditionalists and would be strongly opposed by miners. It would wipe out mining companies’ multi-million dollar investments in custom mining hardware. Such a step is not impossible, but it seems unlikely to happen any time soon.
All of which means that Bitcoin’s power-hungry ways are unlikely to change any time soon.