...
According to satoshi and common sense, the cost of mining should approach the price of the coins mined.
This year, approximately 10% of the total Bitcoin in existence has been mined. In other words, if satoshi is correct, the total cost of maintaining Bitcoin network at the present level of security is 10% of the total market cap.
Few will argue that the lion's share of that cost is electricity. There are online mining calculators which will give you a relatively accurate number, the only guesswork on your part would need to be the the breakdown (by brand and model, and, thus, efficiency) of the gear being used.
Right now, Bitcoin's market cap is small enough for this to be almost inconsequential. But if Bitcoin does succeed as the new world currency, this implies that 10% of the world's wealth will be consumed each year. Most of it in electrical costs. That's staggering, and supplying that much energy is likely unfeasible. Certainly not too eco friendly
But that's all irrelevant. We're starting off with the assumption that running a Bitcoin network is the right way to do money. It's not.
Presupposing that blockchain must be maintained is as justified as specifying mice as the prime mover in the next space shuttle design. Sure, it could be done by introducing some truly Goldbergian complications, but ... see where I'm going with this?
...
If I understand something of the macro economics of bitcoin, you can say that the cost of the mining is of the order of the value of the inflation (somewhat less because miners want to make a profit). Now, the cost of mining is in part the hardware, and in part the energy. I don't know the ratio, but let's say half-half.
At the current inflation rate, which is 10%, the energy cost of mining per year would then be of the order of 5% of the market cap.
At a current market cap of $5 billion, the mining energy cost would then be $250 million per year. Let's put the price of a KWhr to $0.1 (in China and USA), then we have 2.5 billion KWhr per year, which comes down (there are 8760 Hrs in a year) to an average power consumption of 300 MW. That is still reasonable. A third of a big power plant for bitcoin to be mined.
However, imagine that bitcoin price goes up with a factor of 10. Then all the mining in the world would go to something like 3 GW - 3 nuclear power plants. If bitcoin takes over the world economy, and the market cap of bitcoin becomes the world fiat market cap, we arrive at 3000 GW. Now that's embarrassing. That's more than the world's electricity production !
First of all, the current 10% inflation is temporary, it will go down gradually over time.
Secondly, the amount of resources and energy spent on mining reflects the demand for competition in control for transactions.
Mining has nothing to do with market cap in the long run, adjustable difficulty reflects the demand for control.
You don't need to mine Bitcoin in order to use it, you can make profit by providing services or doing some useful work.
The beauty of PoW is that money and control are separate. They must be, as they are two different archetypes. Things you cannot buy with money, you get through control. Keeping them separate ensures the competition for both.