A better metric to consider when valuing the energy cost of bitcoin mining.
Detractors of Bitcoin will commonly cite the energy consumption of bitcoin mining as a knock against it, here are a couple recent examples:
And this is not anything new, people have been concern trolling over bitcoin mining’s energy usage as far back as 2010/2011 on the original bitcointalk forum. The main stream media have been quick to latch on to this energy-waste nonsense, here is another one of their shitposts from 2015:
When you dig into these criticisms you will see that they often cite Bitcoin’s transaction throughput limitations (the transaction rate is capped at approximately ~7 transactions per second) and go on to conclude that it is super inefficient due to the energy cost per transaction.
However this is a very narrow-minded view and on it’s face shows a complete disregard for the value proposition it is bringing to its users. Bitcoin has an uncapped value throughput!
I could go into a huge spiel on how the Bitcoin network is actually more likely to fix consumerism (the root cause of global warming and resource waste) but I’ll save that for another day. Instead I just wanted to point out some interesting metrics that counter the FUD:
This chart shows how much value is settled on the bitcoin network for every megawatt-hour of energy consumed by bitcoin miners. The difference between the “adjusted” line according to the source, Karim Helmy from CoinMetrics (who provided the root data), is as follows:
We try to avoid double-counting spends (if a UTXO is spent within 4 hops of creation, it’s only counted once) and don’t count change sent to originating address
So the adjusted metric is conservative and probably more representative of actual trade settled on the network.
Presently, we can see that even the conservative adjusted value settled is approximately $100,000 USD for every megawatt-hour burned by the network.
How efficient is that?
Well, from an energy cost perspective a utility company will generate electricity at an operating cost of approximately $20 / megawatt-hour, give or take. If we stick to the conservative side and say $50, that’s still a measly $50 to settle financial transactions valued at $100,000 dollars!
Interestingly, we observe in bull runs when the bitcoin price is spiking the value settled per megawatt-hour increases, which would be due to the fact that the price and network demand is spiking while mining hashrate and energy usage lags behind. We can see that at the peak of the 2017 bull run the value settled per megawatt-hour reached $250,000 USD!
Historically, Bitcoin has never been less efficient than ~$10,000 / megawatt-hour.
To figure out the energy used by the bitcoin network I assumed that miners were using the best chips at the time, which wouldn’t be the case in reality. Therefore these numbers are a little optimistic.
Considering the network is certainly more efficient on average than the chip efficiency of an Antminer S9, if we want to be super conservative we can divide the numbers in this graph by about 2.25 to create a lower, ultra-pessimistic bound.
So how efficient is Bitcoin compared to other financial settlement layers?
I don’t know, I’ll leave that for you to figure out.
However, considering most other financial settlement layers are never final and can be easily censored by a third party, I don’t think it is even worth comparing. So with that perspective even the low end of $10,000 /MWh is just incredibly cost efficient.
Thanks for reading!
❤ Steve
PS - Tyler Bain already touched on this idea in the past and wrote a great piece that you should check out here:
(Follow him — @blockbain on Twitter)
