What is a likely solution to Bitcoin's energy consumption problem?

If the price of Bitcoin continues to climb, the total energy consumption of the system will increase proportionally at least until 2020 when block reward halves. It is not unlikely that at this point the price of Bitcoin will have increased so much that energy consumption will be unsustainable in a way that Bitcoin will be opposed by a large proportion of global society (we are still below 1% of total electricity consumption but a 10-fold price increase could change that quickly).

I see the same options that are laid out in this article:

Bitcoin will be abandoned and people just use something else
the protocol is changed to reduce the block reward earlier than 2020
the protocol is changed to exchange Proof-of-Work for something else (maybe Delegated Proof-of-Stake?)

since 2 and 3 will almost certainly result in a fork, they are related to 1. I feel that a solution will have to be found, I'm just wondering if I missed another option and which of the options is most likely.

energy-consumption

lex82

Posted 2017-12-17T18:09:12.250

Reputation: 149

I just found this statement by A. Antonopoulos: https://www.youtube.com/watch?v=2T0OUIW89II However, he does not really convince me. He does not address the problem that the block reward decreases very slowly.

– lex82 – 2018-01-30T07:54:01.560

First why would an increase in price cause energy consumption to go up? That's not immediately clear and probably worth more justification than some of the other nonsense your on. As far as Bitcoin is concerned it does not need a specific energy consumption level. That is not in the standard. That energy is spent on mining Bitcoin simply because it is available. And if at some point it isn't, the Bitcoin will adjust. – marshal craft – 2018-05-13T04:09:28.100

Also the price of bitcoin, to dollars is probably more related to the ratio of liquid Bitcoins to u.s. m2 or m3. So actually one could argue given fixed u.s. supply (which isn't fixed) more mining would reduce the u.s. dollar price of a Bitcoin. Assuming all btc are nice and liquid. Either way completely irrelevant to Bitcoin protocol. – marshal craft – 2018-05-13T04:14:32.003

And as for society's view on it, when was last time society complained at Facebook for storing huge amounts of user data? Are making sure they areent "waisting energy"? Did you drive a car today instead of a bicycle or take a bus? If society cares then the first and only place they need to look is at themselves. – marshal craft – 2018-05-13T04:30:45.533

Also if you actually look at the mining dynamics itself, distributed mining software that says barrows computation time on a web browser, if a relatively small number of people support it, can out pace any single mining group. Nothing you said has to happen and is simply fud. – marshal craft – 2018-05-13T09:27:02.873

Also I wonder how many of these mining pools ultimately end in failure in the long run. Bitcoin simply makes use of the computational power available to miners. It doesn't by any means require it and that seems to be biggest point everyone is missing. Unless I'm missing something? But nobody has voiced that point so. – marshal craft – 2018-05-13T09:32:49.940

Theoretically (not familiar or too lazy to loo) , iff all the computers disappeared but somebody managed to print out block chain, you could still follow the protocol and perform the hash on paper. You could then break it up and hire people to improve hash performance. It would still be hard to alter the chain and again if say a computer did come back it would be better off mining. This Bitcoin would still work completely. Hope this helps to explain why there is no tera flop requirement or min wattage requirement for Bitcoin. – marshal craft – 2018-05-13T09:37:32.220

"some of the other nonsense your on" is not the appropriate tone for stackexchange, especially when you are totally missing the point. – lex82 – 2018-05-13T14:52:38.507

@marshalcraft I don't claim Bitcoin needs a certain amount of energy. You are right that the protocol could -- in theory -- work on paper. But you also point out that you have an advantage when you use a machine. The increased energy usage in case of a rising price is simple economics. – lex82 – 2018-05-13T15:02:19.560

You have an advantage over other miners if you buy more hardware and use more electricity. At some point this becomes unsustaible because you spend more on mining than you earn in Bitcoins. However, when the price doubles, you can (and many people will) spend twice as much on mining resources as before, because it would still be profitable (given it has been profitable before) and you fall behind if others do it and you don't. – lex82 – 2018-05-13T15:02:26.157

Another similar explanation can be found here: https://www.quora.com/Is-there-a-correlation-between-bitcoins-hashrate-and-price

– lex82 – 2018-05-13T15:03:53.540

Would it be bill gates most productive strategy to hold so much wealth? What about Rockefeller? A lot of times these situations 90% of the wealth is left in used. This could be similar to using 1% of world energy to sustain accentually a Fiat currency. I don't these economic choices are simple or fully understood either. – marshal craft – 2018-05-14T10:44:01.323

My opinion is that these aren't clear set of possible outcomes. And if you really think they are that is fine, maybe you are right. But I would need some justification before specifically picking one of them. – marshal craft – 2018-05-14T10:46:53.027

Let us continue this discussion in chat.

– marshal craft – 2018-05-14T10:48:39.613

One thing is Bitcoin earns about 7mil from btc a day. How much does so much energy cost per day to mine that? Probably a lot more than 7mil. So I'm betting on its just a lot of failing get rich quick ventures. – marshal craft – 2018-05-14T11:02:46.203

Answers

I feel that a solution will have to be found

I'm not sure I agree there is a 'problem' that needs to be explicitly fixed. The article you linked explains the relationship between price and electricity usage, now remember that price is a measure of the demand (since the supply is fixed and known).

So the market acts as a naturally regulating force: as people pay more and more to buy and use bitcoin, the network's power consumption will increase (and vice versa). So to say "bitcoin uses too much energy!" is to say "bitcoin is worth less than the current market value!". The article linked supports this point, but you kind of have to 'read between the lines' to realize it is true.

Consider gold mining: if the price of gold were to rise, then it would become profitable to start mining gold in places where a more expensive processes will be needed to extract the gold from the earth. Is it a problem that companies will spend a lot of energy to get this extra supply that is now worth while?

Interestingly, consider what happens if the demand for bitcoin continues to increase, to the point that the network is in direct competition with all other use cases for power. In that case, for the btc price to increase, electricity prices would have to be increased, effectively spreading the cost of mining across all of the world's energy usage (because if the demand for bitcoin did not exist, everyone would be paying less to power their home appliances, etc).

I think it is easy to say "this is a problem!!", but the less obvious realization is "bitcoin is using this much energy because the market considers it useful enough to be worth the cost"

I'm just wondering if I missed another option and which of the options is most likely.

I don't think any of the three options you listed are likely to happen.

Another 'option' to consider is this: once mining chips become commodity (more easily and readily available from multiple suppliers), I think the nature of the mining competition will change. If everyone has access to good ASIC chips, then there will be immense competition to develop and utilize the cheapest sources of power available. What if bitcoin miners must help fund the development of ultra-efficient solar panels in order to maintain a competitive advantage? This is how ASIC miners were developed, why can't the same happen for another variable in the mining industry?

If this occurs, rather than the energy consumption of the network being a problem, it becomes an incentive for finding a solution to the larger issue of 'needing to develop green sources of power'.

chytrik

Posted 2017-12-17T18:09:12.250

Reputation: 10 276

Thank you for this answer. I'll have to think about it a bit. Just one quick note. You write: "So the market acts as a naturally regulating force". Markets are good at allocating resources with regard to certain incentives and costs. However, they can be self-destructing too, if not regulated (e.g. tragedy of the commons). – lex82 – 2017-12-18T12:37:42.143

This is true, so perhaps a good question to ask is "will the advent of bitcoin lead to a net positive for humanity?". This is tough to answer definitively, but I think that environmental damage caused by the creation of a politically neutral system of value is better than (say) environmental damage cause by the proliferation of fossil fuels. I'm not sure "hey, at least its better than X" is a great argument in itself, but ultimately any system we create will have some impact on the environment, so it becomes a matter of "how much, how bad, and for what benefit?". – chytrik – 2017-12-18T22:46:07.250

Resource consumption is ok, to secure the network. I just worry it could get really get out of hand. The electricity consumption seems to be sufficient to secure the network already. The market is not demanding more mining but the market wants Bitcoins. That both is coupled is an arbitrary decision by the protocol designer. Block reward could be much lower and the system would still work well. With 4% inflation, miners could spend electricity worth 4% of Bitcoin market cap per year (mining fees on top). I just have the feeling that this is not sustainable. However, I don't see how to avoid it. – lex82 – 2017-12-18T23:14:03.357

The market will never get more Bitcoins, since mining is adaptive. Insofar it is a bit different from gold or oil where more (per year) is extracted when price increases. So it's a bit weird. The demand for Bitcoin increases but what people get is more wasted energy. Looks like the market cannot easily regulate that. But I'll think about it a bit more. – lex82 – 2017-12-18T23:18:51.293

FWIW, the network is currently at ~4% inflation alreaddy. I'm not sure inflation matters anyways, bitcoin holders are essentially subsidizing the block reward already, by letting their bitcoins to lose value to inflation in order to pay miners to secure the network. As the inflation level drops, the miner's reward will come more and more from tx fees. I think that where the miner's reward comes from does not really matter though, what matters is how much of an incentive they have to mine. That is determined by what people will pay, which is not capped by the explicit supply/inflation of btc. – chytrik – 2017-12-18T23:42:56.893

In the long term you're right. But currently inflation equals roughly what miners get (12.5 BTC block reward vs. ~1.5 BTC fees). It doesn't matter if it is inflation or something else. What matters is the 4% of total market cap. If Bitcoin price increases by 500% per year, market cap could be as high as 30 trillion in 2020 already. 4% of this is more than 1 trillion dollars. This would have a huge effect on electricity prices. – lex82 – 2017-12-19T07:05:35.303

Right, but if the supply of new coins were reduced, then the price of the bitcoins already in existence would increase to account for this reduction. The variable that matters is "how much money is trying to enter the market". More people buying = more money entering the market = higher prices = more energy spent on mining. The number of coins in supply / inflation dictates the price per coin, but as you pointed out, it is the total market cap that that more directly influences the energy usage of the system. – chytrik – 2017-12-19T09:05:27.540

Let us continue this discussion in chat.

– lex82 – 2017-12-19T09:18:41.183

Another solution would be to work on the bitcoin scalability problem.

Technological solutions like the lightning network or Schnorr signatures will hopefully make it cheaper to transact in bitcoin (and thus, reduce the reward and by extension also reduce the amount of computational power competing for the block), in some cases making it even somewhat secondary (as with the Lightning Network).

ranchalp

Posted 2017-12-17T18:09:12.250

Reputation: 449

Currently the basic block reward is 12.5 BTC + about 1.5 in fees, so fees play only a minor role. With lightning network enabled the value of BTC will very likely rise a lot, so the problem will actually worsen since miners are competing for the much more valuable block reward. – lex82 – 2017-12-17T21:35:51.430

I strongly disagree with your views, as you said miners will have to compete, which discourages miners from getting in the network. I have the feeling you are thinking about miners and not trying to answer the question. – ranchalp – 2017-12-18T10:39:07.740

I asked the question, so I am interested in an answer. However, technological advances will not change to value of the resources spent to mine the Bitcoins. A reward of one billion will lead more people to compete for it than a reward of one million. Thus, more resources will be spent to get it. If everybody has better and cheaper hardware, then more of it will be bought. – lex82 – 2017-12-18T12:33:49.093

-1

Bitcoin uses Proof of work to validate,secure and record transactions. This is done through mining and as long as mining continues, electricity will be used. Following are ways to avert it: 1). For micro transaction, use offchain technology like Lightning network. 2). Pack more transactions in a block, either by increasing block size or by removing signature from block(segwit) or a combination of both. 3).Use alternate algorithm (other than POW). This is not possible for BTC.

user2203937

Posted 2017-12-17T18:09:12.250

Reputation: 89

surprised to see folks abusing their power and downvoting without any reason. – user2203937 – 2018-05-12T13:25:21.893

Mining will not stop when 21M coins are mined. Miners still receive the fees. – lex82 – 2018-05-12T14:53:46.963

@user2203937 your answer contains inaccuracies, for example, putting more txs in a block will not necessarily reduce the power spent on mining, in fact if the utility of the network increases (more tx/s) one may expect the price (and thus mining energy consumption) to increase as well. Also, mining will not stop once all of the coins are minted. The end of the supply inflation era may not have much of an effect on the mining electricity consumption, as miners will have become dependant on tx fees instead by that time, so it won't be a shock to their income. – chytrik – 2018-05-13T08:38:52.970

@chytrik It seems that you did not understand mining and blockchain. It is block which is validated during mining so if we include more transactions in block, overall less numbers of block will be mined for any given number of transactions. It's simple arithmetic. Also, do not twist my words when I say ,mining will stop it means no new coins will be generated. Tx fee will skyrocket then. If it is you who is downvoting my answer without even understanding the context, read more and stop abusing your little power here. Just because you have been here a little more doesn't make you smart. – user2203937 – 2018-05-13T09:03:11.847

1@user2203937 I'm sorry, but your understanding of mining is off a little. The number txs that fit into a block has nothing to do with how many blocks will be mined in a given time period. I don't think I twisted any words, you literally wrote that mining would stop, not that coin minting would end. In any case, please don't take a downvote to be an insult, the majority of users on here seem to be quite pleasant and willing to help. – chytrik – 2018-05-13T09:16:07.860

@chytrik, If you downvote for some valid reason, I will take it in right stride but if you do it just that you didn't understand one's argument then there is problem within you. You need to open up. Tell me,let's say 1000 transaction are generated, 1 block can keep 10 transaction - so total 100 blocks will be mined, what if one block keeps 200 transaction how many will be mined. The answer is 5. So won't that take less power overall. – user2203937 – 2018-05-13T09:19:56.573

@Chytrik, use power with responsibility. I have heard many instances on stack overflow where people abuse it. Please don't do that. – user2203937 – 2018-05-13T09:23:03.733

@user2203937, I would also downvote your answer if I could (need 125 reputation), just because it is wrong. Bitcoin miners generate one block every 10minutes regardless of the number of transactions. For this block they receive currently 12.5 BTC block reward. It's that simple. – lex82 – 2018-05-13T15:20:33.290

@lex82, block generation time is 10 mins....I know that. What has it got to do with over all electricity consumption. If you pack more transactions in block, overall fewer blocks will go through validation and pow, hence will need less electricity. I also would love to downvote both your ill researched comments – user2203937 – 2018-05-13T15:39:17.487

The number of Tx and the block time has nothing to do with the electricity consumption. That's the point. Currently there is a mining reward of around 80 BTC per hour and people will spend a lot of resources on mining to get them. If the BTC are worth 10x as much as they are today, around 10x the resources will be spent. – lex82 – 2018-05-13T18:21:46.003

@lex82, I have given you the reasoning. You said it has nothing to do with transactions in block without proper reasoning and then attributed it to cost. I would have really considered your logic if you talked in terms of watts or kilowatts but seems you do not know difference between cost and watts So I drop my argument here. – user2203937 – 2018-05-13T18:37:36.147

1@user2203937 Your math about fitting 1000 txs in either 100 blocks or 5 blocks is right, but that example tells us about the miners ‘electricity cost per confirmed transaction’, whereas OP’s Q is about ‘electricity cost per time’. A block is generated every ten minutes, whether it is full of txs, or empty. Miners are always mining (spending $/sec), no matter how many txs are waiting to be confirmed. So having the ability to add more txs to a block will not decrease the $ spent per 10 minutes, only the $ spent per confirmed transaction. – chytrik – 2018-05-14T05:05:28.530

-2

Lightning network could be answer but if is not we can always do fork on bitcoin to upgrade to different system, so we can keep Bitcoin name, users and amount of each address. I also see no reason why in the future they could not make one joint fork on the top 5 cryptocurrencies with migration to one "multicurrency" coin and keep it old max supply of each coin (This is just a suggestion that could be developed). Only consensus is needed.

Noel G

Posted 2017-12-17T18:09:12.250

Reputation: 1

None of these things are in any way related to energy consumption of PoW. – Pieter Wuille – 2018-05-14T05:37:27.903