And to be complete, there are other ways that PoW blockchains can compete on energy efficiency: 1) Inflation rate 2) PoW function 3) Transaction cost after b

And to be complete, there are other ways that PoW blockchains can compete on energy efficiency: 1) Inflation rate 2) PoW function 3) Transaction cost after blocksize limit reached The inflation rate determines the incentive given to miners. With less incentive, the energy use will go down. This comes at the cost of increasing the ease of a double-spend, but there is already far more than enough difficulty to double-spend at the current mining rates. The PoW function determines the capital cost required to buy the miners themselves. If you increase the capital costs (e.g. with a memory-hard PoW function, or requiring less-efficient GPUs instead of more-efficient ASICs) then miners will be able to spend less of their costs on energy. This is one way in which Eth does well by the above metric, and is a valid way to reduce energy usage in PoW. Of course, transitioning to proof-of-stake (as eth is doing) will eliminate the energy problem entirely. Finally, some blockchains (e.g. BTC, ETH) have reached their capacity of transactions per second, and then users can add a fee to each transaction to incentivize mining pools to include them in a block. These fees do add incentive for miners to dump energy into their blocks. However, they are a much smaller portion of the incentive than you would expect.