2018-01-15

On Bitcoin power consumption

A few weeks ago, there were some articles making the rounds claiming that "in 2020 Bitcoin will consume more power than the world does today", that it creates a lot of pollution, etc. It seems that Bitcoin will be blamed for cheap, dirty power plants opening up again and so on. Well, let's break it down and see where this whole situation might be really headed...

Rate of growth


It seems that the origin of the claim that Bitcoin will consume all of world's electricity came from this site. The misconception probably stemmed from these two bullet points:

"
  • In the past month alone, Bitcoin mining electricity consumption is estimated to have increased by 29.98%
  • If it keeps increasing at this rate, Bitcoin mining will consume all the world’s electricity by February 2020.
"

As usual, there is a relevant XKCD comic on this:


The obvious crux here is the phrase "if it keeps increasing at this rate" - which can be very unlikely to happen. It's basically exponential growth, which can be really hard to match on scale - you would have to produce more mining hardware, which would also require you to at some point start creating more electronics factories and more chip fabrication facilities and so on. All of those can take a lot of time to get going.

Even if you could fabricate all of those chips, at some point you will run into some limits imposed by the economics of Bitcoin mining.

Economics of Bitcoin mining


The Proof of Work algorithm Bitcoin uses is self-balancing - it doesn't really care how much or how little mining power there is in the network, it will still try to maintain the 1-block-per-10-minutes mining schedule. With the block reward being inflexible and the transaction fees being finite, the only factor that can have a large impact on the Bitcoin mining rate is the price of a single bitcoin. If the price doubles, you can expect the difficulty to double (with some time delay) due to more miners coming onto the network. If the price goes down, some miners will be priced out of the market and the difficulty will go down.

In an ideal world, the cost of mining 1 bitcoin would be about 1 bitcoin. In real world, you have some inefficiencies and opportunity cost, so you should see some small profit margin at the least.

To figure out how much electricity Bitcoin will consume in the future, you have to break things down into basics.

First - what is a miner really? It's a computer that converts electricity into Bitcoin and produces some heat. In a perfect scenario, every dollar of electricity you would pump into the machine would give you about a dollar worth of coins.

The total amount of power produced in the world in 2013 was 23'322TWh. The average electricity price ranges from 0.08 to 0.42 USD / kWh. Taking the cheapest cost, we know that in 2013 we generated about 1.865T USD.

Every year we have about 52'596 Bitcoin blocks generated. Assuming 12.5 BTC block reward and 2.5 BTC in transaction fees, we should expect to give the miners a net gain of 788'940 BTC in a year.

Assuming the miners converted 100% of the value of electricity into mining BTC, we would expect 1 BTC to be worth about 2.36M USD. At the time of writing, one bitcoin is "only" worth 13.5k USD, or about 200 times less. For some perspective - 1/200 of the current price would be 67.5 USD / BTC, a price range from about early 2013 - 5 years ago.

Since the Bitcoin mining rewards halve every 4 years, the price per BTC would have to increase by further factors to keep up with the global power generation revenue.

Taking a more realistic look


A lot of the numbers here assumed perfect scenarios - no money being lost while converting from electricity to bitcoins, no cost of manufacturing and maintaining the miners, etc. In reality, you can expect noticeable losses from those factors, as well as things like taxes, transfer fees, internet cost, etc.

It's very unlikely Bitcoin will ever become a large enough factor in the global energy production to have an impact on the electricity prices. It's more likely to be used close to some renewable power plants that have excess of electricity to get rid of from time to time, possibly some places that can re-use the heat generated from the miners that can't rely on more efficient heat pumps, or simply pooling around places that have really cheap electricity.

At the very least, Bitcoin mining is a universal "overflow valve" - it's a simple place to push excess electricity production capacity (and to some extend - electronics production as well) and turn it into tangible money. The mining process will gladly take any amount of excess and give some returns. If the returns are better than the savings from spinning the operation down and then back up again - it's worthwhile.

Conclusions


Bitcoin is very unlikely to be a major consumer of the global energy production. On the other hand, it's a good sink for excess power generated from renewable sources, always allowing one to convert spare electricity into cash.

No comments:

Post a comment