The Cambridge bitcoin electricity index confirms that it is nigh impossible to get accurate figures.
The Cambridge Centre for Alternative Finance, a research institute at University of Cambridge business school in the UK, today launched the CBECI, an index that attempts to provide a real-time estimate of the total annual electricity usage of the Bitcoin network.
The Bitcoin network's energy consumption has long been a bone of contention among maximalists and detractors. Digiconomist, one of the most cited measurers, says the network is consuming as much energy in a year as Colombia, making it the 42nd biggest energy user in the world, if Bitcoin were a state. A Coinshares report meanwhile, caveats that number by saying the vast majority of that energy is produced by renewables.
Can Cambridge clear up the mess, or has it just made things even worse? Somehow it manages to do neither.
Bringing in the new model
The CBECI says the report is just a “best guess,” acknowledging that it is incredibly difficult to create a reliable measure of how much energy the network uses.
"Reliable estimates of Bitcoin's electricity usage are rare: in most cases, they only provide a one-time snapshot, and the numbers often show substantial discrepancies from one model to another," the report acknowledges.
The problem with trying to guess how much power bitcoin uses is that it has no real data to go on bar the hash rate. Nearly all reports take that number and work backwards to speculate on how much electricity each of the miners are using. A more realistic study would involve just asking each miner for a copy of their electricity bill–good luck finding an email address for all those Mongolian mines.
Sadly, the CCAF model is no different. While it claims its come up with a new way of measuring energy usage, it is largely dependent on the same factors and assumptions that align with previous models. However, by making slightly different assumptions, it produces wildly different estimates of the network’s current usage.
CBECI’s best guess puts the current network usage at 53 Twh, a figure 24 percent lower than Digiconomist’s estimation of 70 Twh. However, Cambridge’s analysis says total power consumption is less a number, more a range, estimating it to be between 21 Twh and as high as 146 Twh—quite the spread. That’s the equivalent of consuming as much power as Poland, the worlds 24th biggest consumer, or as as little as Equador, ranked 70th.
In short, it’s another example of how difficult it is to predict the actual energy usage of the Bitcoin network. Considering the report’s aim was to provide reliable data to help "policymakers, regulators, researchers, the media and others,” make up their minds on whether it’s wasteful or not, it’s done neither.
Instead what we have is: The bitcoin network uses some electricity, and some of it may come from renewable sources. Brilliant.