18
Jan
2018
Does Bitcoin Use Too Much Electricity?
My sources say no. My latest at IER. I like this opening:
An optimist says the glass is half full. A pessimist says the glass if half empty. And a Vox writer says if you drink 60 glasses of that stuff in the next hour, it’ll kill you.
A case in point is the recent Vox column by Umair Irfan, warning that the Bitcoin network has caused a huge surge in energy consumption. And yet, Irfan’s own article admits that even the largest estimate—which could be double the actual figure—suggests Bitcoin only uses about 0.14 percent of global electricity. It seems somewhat unfair to single out Bitcoin and ignore the other 99.86 percent of the activities that use electricity.
Socialism: Does Bitcoin Use Too Much Electricity?
Capitalism: That’s impossible.
Well, the question is not framed correctly. Obviously the question must be: Are energy costs making BTC unattractive if you want to transfer wealth or do transactions as compared to other tradional means.
Whether or not it is too much, to me that a very surprising amount of electricity.
As you say, whether it is too much depends on externalities. If all the costs are internalised the “right” amount will be used.
To put it in perspective I would very much like to see the energy used in mining gold. So I had a look and found this very interesting analysis
https://srsroccoreport.com/bitcoin-vs-gold-which-ones-a-bubble-how-much-energy-do-they-really-consume/
Some highlights:
“While it only takes 1.4 barrels of oil equivalent to produce an ounce of gold, it takes 10.1 barrels of oil equivalent to produce one Bitcoin. Thus, Bitcoin consumes seven times more energy to produce each digital coin than it does for each gold oz. Even though it takes a lot more energy to produce each Bitcoin, the Gold Mining Industry consumes one hell of a lot more energy overall…
“Which means, the primary gold mining industry is currently producing approximately 135 times more gold than the Bitcoin mining industry.”
They estimate the mining costs at $1500 per bitcoin.
“Currently, Bitcoin is fetching an estimated $4,700 profit per coin versus $136 for gold. Here’s how I arrived at those figures:”
Fascinating stuff, but even compared to gold mining, which I would have considered pretty energy intensive, bitcoin is far more profligate.
.14 percent sounds insanely high to me, and makes me actually more concerned than before…
What about 0.07%? Because in the Vox article, I think they asked another expert who thought the one guy’s numbers were probably double the true value. (Pretty sure that’s right.)
The other expert did think the true value was half the original, but it was the lower one that was 0.15% of global power.
“Marc Bevand, an investor and entrepreneur, was skeptical of de Vries’ tally of Bitcoin’s energy use and argued that the real global energy footprint of mining was likely closer to 15 terawatt-hours, which is still a huge amount of electricity, but half of the estimate on Digiconomist….
“Bevand estimated that data centers account for 1 percent of power demand around the world, while Bitcoin mining uses 0.15 percent of global power.”
It seems that the 0.15% is the lower estimate (from Bevand, not Digiconomist).
The article I linked to estimated 11.7 TWh/yr, close to Bevands estimate of 15 TWh/yr, and that was last November. They estimate a 45% increase in 3 months so they are in good agreement.
Using this lower estimate it still takes 7 times as much oil to mine one bitcoin than an oz of gold, which was very surprising to me.
as I understand it, the problem is the rate of growth of el. consumption
>https://digiconomist.net/bitcoin-energy-consumption
with the current growth rate, BTC should consume some 20k TWh by 2020 which is roughly the World consumption.
For some time that was my main concern about the long term prospect of cryptos. A friend just bought cca 1500 graphics processors and is setting up a small power station(gas) to support it. Even though, I thought, there is no way the miners could at least double the world supply in less than 2 year – just not enough time.
Then, I saw the beast nvidia is releasing sometime later this year – the Pegasus chip – for level 5 autonomous driving, which marks a 10x fold increase in raw power (320 trillion operation/sec.), and stopped worrying about electricity consumption.(https://www.engadget.com/2017/10/10/nvidia-introduces-a-computer-for-level-5-autonomous-cars/)
Combined with the software tweaks to the ledgers, I think they will be able to smooth out the curve and continue growth. I would be more conservative with the time frames tho