4 ms·

I couldn't find anything but I wonder how global energy consumption changes on track with the predicted bitcoin processing growth? are we talking 1% of global

I couldn't find anything but I wonder how global energy consumption changes on track with the predicted bitcoin processing growth? are we talking 1% of global power consumption in 10 years or 10%?

petertodd 9y ago
It's possible to estimate an upper bound based on the dollar value of the BTC miners receive as revenue from inflation (subsidy) and transaction fees. The inflation rate in Bitcoin drops over time; transaction fees will likely increase, although not in proportion to the total economic value of transactions (and other uses) secured by Bitcoin (bitcoin is not "backed by energy", it is secured by energy expenditure). Currently miner revenue is less than $7 million/day. It's a bit tricky to convert that into, say, CO2 emitted as miners tend to use the cheapest possible sources of energy, which are stranded renewable resources like hydroelectricity dams built in places distant from consumers (apparently China among others have built a lot of uneconomic dams for various political reasons). The PoW security model doesn't require the amount of energy destroyed to increase indefinitely as a percentage of total value secured. Rather, the amount of energy destroyed by PoW simply has to be more than the highest funded attacker can afford. That's easy to do, as attackers who would attack via PoW are few and far between - as your attacker gets more powerful (governments) there are much cheaper attacks that can be done than naively attacking PoW directly. In any case, I wouldn't be at all surprised if the amount of CO2 emitted by Bitcoin mining operations is dwarfed by the amount of CO2 emitted by plane travel associated with people attending Bitcoin conferences...
- danmaz74 9y ago
  > Rather, the amount of energy destroyed by PoW simply has to be more than the highest funded attacker can afford. Miners don't measure or care for how much an attacker can afford. They will simply continue adding mining hardware and increase electricity usage as long as they can turn a profit, which depends mostly on: * the marginal cost of electricity * the amortization cost of the hardware * the price of BTC * the difficulty
elif 9y ago
If you calculate power usage for the global hash rate as if it were produced by, on average, 2-generation old mining hardware @0.51 W/Gh, you end up with 0.017% of average global electricity production (3x more than last year). If you extrapolate an unconstrained tripling of production every year, you will (to my surprise) arrive at your 10% estimate. However, if we assume the last year's additional capacity was on modern hardware (5x more efficient), miners are presently using 0.0005% of global production, and after 10 years of unconstrained annual tripling would only account for 0.295% Further advancements in efficiency or a slowing of the growth rate from 300% annually seem like likely diminishing factors.
- sobani 9y ago
  In the end it doesn't matter how efficient the mining ends up being. All that matters is the price of electricity. Assuming efficiency gains level out and thus the cycle of buying new hardware to keep up approaches infinity length, the only major cost left is electricity use. So in that case the energy used is dependent on (transaction fees * bitcoin price) / (energy price) So to get a % of all energy usage, you need to know the future energy usage, the future energy price, the future bitcoin price and the future amount that will be spend on transaction fees. I'm sure the future energy usage and price have good forecasts available. Then you're left with the bitcoin price and fees (both which are dependent on popularity (which in turn is dependent on price and fees)), but good luck forecasting that.

I couldn't find anything but I wonder how global energy consumption changes on track with the predicted bitcoin processing growth? are we talking 1% of global power consumption in 10 years or 10%?

petertodd 9y ago

It's possible to estimate an upper bound based on the dollar value of the BTC miners receive as revenue from inflation (subsidy) and transaction fees. The inflation rate in Bitcoin drops over time; transaction fees will likely increase, although not in proportion to the total economic value of transactions (and other uses) secured by Bitcoin (bitcoin is not "backed by energy", it is secured by energy expenditure). Currently miner revenue is less than $7 million/day. It's a bit tricky to convert that into, say, CO2 emitted as miners tend to use the cheapest possible sources of energy, which are stranded renewable resources like hydroelectricity dams built in places distant from consumers (apparently China among others have built a lot of uneconomic dams for various political reasons). The PoW security model doesn't require the amount of energy destroyed to increase indefinitely as a percentage of total value secured. Rather, the amount of energy destroyed by PoW simply has to be more than the highest funded attacker can afford. That's easy to do, as attackers who would attack via PoW are few and far between - as your attacker gets more powerful (governments) there are much cheaper attacks that can be done than naively attacking PoW directly. In any case, I wouldn't be at all surprised if the amount of CO2 emitted by Bitcoin mining operations is dwarfed by the amount of CO2 emitted by plane travel associated with people attending Bitcoin conferences...

danmaz74 9y ago
> Rather, the amount of energy destroyed by PoW simply has to be more than the highest funded attacker can afford. Miners don't measure or care for how much an attacker can afford. They will simply continue adding mining hardware and increase electricity usage as long as they can turn a profit, which depends mostly on: * the marginal cost of electricity * the amortization cost of the hardware * the price of BTC * the difficulty

elif 9y ago

If you calculate power usage for the global hash rate as if it were produced by, on average, 2-generation old mining hardware @0.51 W/Gh, you end up with 0.017% of average global electricity production (3x more than last year). If you extrapolate an unconstrained tripling of production every year, you will (to my surprise) arrive at your 10% estimate. However, if we assume the last year's additional capacity was on modern hardware (5x more efficient), miners are presently using 0.0005% of global production, and after 10 years of unconstrained annual tripling would only account for 0.295% Further advancements in efficiency or a slowing of the growth rate from 300% annually seem like likely diminishing factors.

sobani 9y ago
In the end it doesn't matter how efficient the mining ends up being. All that matters is the price of electricity. Assuming efficiency gains level out and thus the cycle of buying new hardware to keep up approaches infinity length, the only major cost left is electricity use. So in that case the energy used is dependent on (transaction fees * bitcoin price) / (energy price) So to get a % of all energy usage, you need to know the future energy usage, the future energy price, the future bitcoin price and the future amount that will be spend on transaction fees. I'm sure the future energy usage and price have good forecasts available. Then you're left with the bitcoin price and fees (both which are dependent on popularity (which in turn is dependent on price and fees)), but good luck forecasting that.