Bitcoin consumes more power than 159 countries - really?
The headline is of course a self-runner: Bitcoin "eats" more electricity than 159 countries. At almost 30 terawatts, the energy demand of the decentralized currency is about as high as that of Slovakia and only a little lower than that of Denmark. This is of course not quite optimal and makes Bitcoin the climate killer number one among the payment systems.
What's wrong with these messages? They refer to a "study" of the British power comparison portal Power Compare. However, this has taken only an estimate of the crypto news portal Digiconomist and illustrated by infographics. Digiconomist runs a Bitcoin Energy Consumption Index that reflects Bitcoin's hunger for electricity in a chart. Currently, the index is at around 30 terawatts, which is about as much as the state of Oman consumed and put Bitcoin in the power consumption ranking of the States to rank 65th
A critical look at the methodology of the digiconomist is unfortunately not found in any article. So how does the portal get this value?
"Although it's easy to calculate the absolute hash rate of the network, it's impossible to say what that means for energy use," says Digiconomist. The problem is that different mining machines have different levels of power consumption. In order to determine the total power consumption of Bitcoin, it has therefore been estimated in the past which machines are distributed in which ratio in the network. But Digiconomist concludes that such estimates neglect important factors, such as cooling, based on a report on the world's largest bitcoin mine set up by Bitmain in Inner Mongolia, so the results so far were inaccurate.
"The Bitcoin Energy Consumption Index therefore proposes to turn the problem upside down and calculate energy consumption from an economic perspective." Digiconomist assumes that there is a relationship between income and costs of miners. "Since electricity costs are a significant part of the miner's spending, it is logical that the absolute power consumption of the Bitcoin network is related to the revenues of the miners." So the index takes the miners' revenue at the current market price, estimates how much spend them on electricity (60 percent) and how much they pay for electricity (5 cents), and how much electricity Bitcoin should consume.
The bill goes something like this: You take the reward per block (12.5 Bitcoin + fees), multiply it times 144 (that's how many blocks are found around the day), and again with the current market price ($ 9,500). This is then extrapolated to one year to find out what the miners earn each year, takes 60 percent of what the annual electricity expenses can be, and charged with an electricity price of 5 cents per kilowatt hours.
The result is known: 30 terawatt hours a year, or as much as Oman or Slovakia. Roughly speaking, this value should not be implausible. But there are a few reasons to put it into perspective at least.
First, the baseline seems to be too high. Digiconomist assumes electricity costs of 0.284 watts per Gigahash. The current Antminer, S9, however, requires just under a third, namely about 0.088 watts per Gigahash. Even the pre-predecessor, the S7, consumed less than 0.284 watts per gigahash. One can assume that the "additional energy costs" are set a little too high.
Second, a figure of 60 percent for electricity spending seems a bit too high. Take a look at the Antminer S9, for example. It costs 1,415 dollars and consumes about 1.2 kilowatts. If, like the Digiconomist, you start with an electricity price of 5 cents per kilowatt hour, the operation of the S9 costs 7 cents per hour or $ 613 per year. So the device has to run for more than 2 years so that the electricity costs are the targeted 60 percent of the total cost.
Third, the Digiconomist assumes the current price, and assumes that it is immediately implemented in Hashrate at the current state of the art. In fact, the retrofitting of miners in this wild year is almost permanently behind the rise of the price. And even if it comes after, the construction of the Hashrate is usually accompanied by a technical upgrade to more efficient miners. In other words, if miners have invested the network's current revenues in hardware, the rate of gigahash per watt will have continued to improve.
An interesting case study is the Bitmain Mine in Inner Mongolia, described by the Digiconomist himself. It consists of 21,000 miners, which are housed in eight buildings. In early August, when the mine was visited by several media outlets, it allegedly generated nearly 4 percent of the total bitcoin hashrates and consumed 40 megawatts per hour. If you extrapolate that, the network at that time had a total consumption of 1,000 megawatts per hour, resulting in an annual consumption of about 8.7 terawatts. According to the index, electricity consumption has more than tripled since the beginning of August (while the rate of hashrates itself has not even doubled).
The Bitcoin Energy Consumption Index is unable to maintain consistency with its own case study. This shows that one must enjoy the results with caution. As good as 30 Terahash, the network could have a consumption of 10 Terahash - as well as 40 Terahash. One thing is clear - we have an order of magnitude. At best Bitcoin consumes "only" as much energy as a country like Uruguay (rank 90), in the worst comes the cryptocurrency on electricity costs of the format of a state like New Zealand (rank 55).
That sounds dramatic, and it does, to a degree. However, this is put into perspective when you realize that it will probably be the big industrialized nations that are investing excess energy in mining. You could say, for example, that China needs to put in just 0.5 percent of its annual electricity bill to power its entire Bitcoin network, or 0.7 percent of US consumption.
In addition, Digiconomist's statistics stimulate a much more interesting line of thought than sheer outrage: if the average bitcoin mining power cost is about 5 cents, and miners burn 30 terawatts a year, then that means it's in the world 30 terawatts of excess electricity that can not manage to generate more than 5 cents per kilowatt-hour. The bitcoin hashrate, in this sense, becomes an indicator of how lucrative power is otherwise used.
Post a Comment