Texas paid a Bitcoin miner to use less electricity during hottest month(texasmonthly.com)
texasmonthly.com
Texas paid a Bitcoin miner to use less electricity during hottest month
https://www.texasmonthly.com/news-politics/texas-bitcoin-miner-riot-31-million-energy-credits/
43 comments
Thanks! Macroexpanded:
Texas paid Bitcoin miner more than $31M to cut energy usage - https://news.ycombinator.com/item?id=37428605 - Sept 2023 (130 comments)
Texas paid Bitcoin miner Riot $31.7M to shut down in August - https://news.ycombinator.com/item?id=37418866 - Sept 2023 (55 comments)
Texas paid Bitcoin miner more than $31M to cut energy usage - https://news.ycombinator.com/item?id=37428605 - Sept 2023 (130 comments)
Texas paid Bitcoin miner Riot $31.7M to shut down in August - https://news.ycombinator.com/item?id=37418866 - Sept 2023 (55 comments)
Grid stability is not free. This concept is not new. Historically things like aluminum smelting had very similar contracts to ensure grid stability.
Key difference being that aluminum smelting is a useful thing for society to have. Bitcoin, not so much, so it's easy to see why this sticks in the craw of someone.
Presumably the bitcoin miners were the cheapest way the grid operator could buy this stability. If the grid operator chose a different solution to achieve grid stability then power may cost more for all consumers. It’s a trade off.
Wait till everyone being negative about bitcoin miners getting paid to not mine learn that sometimes the grid operator is going to pay the bitcoin miners TO mine. Effectively the bitcoin miners will sometimes not only get free electricity but be paid to consume it.
Again, grid stability at the lowest cost is the name of the game.
Again, grid stability at the lowest cost is the name of the game.
Yeah I don't really have a problem with this. They are just going to a large electricity user and saying "we'll give you some free electricity later if you stop using it now." When the alternatives are brownouts or giving the grid operator the power to force businesses to suspend their operations, this is better.
As usual lots of hot takes from people who just read the headline
As usual lots of hot takes from people who just read the headline
Why doesn't the market decide, if the bitcoin miners got 20,000% rate increases they would stop on their own for free. If the rate goes to zero, they will surely jump in an use as much as their hardware can burn. Why are payments outside of rate changes necessary at all?
Because the grid needs a very large consumer to stop RIGHT NOW and be under contract to do so when told. To get this contract, the grid promises money. There is no guarantee that adjusting the power cost for a customer will make them stop consuming power when it is needed. The contracts are in place to ensure the grid operator KNOWS that they can achieve stability when they need it.
Why can't the grid operator just put "we reserve the right to shut off your power if demand exceeds our capacity" in their TOS?
I think aluminum smelters have big costs to restart after shutting down, so it makes sense to pay them. Bitcoin would take less than a minute to start mining again.
The entire reason aluminum smelters are happy to play along is because their restart costs are relatively cheap.
An aluminum smelter uses a humongous amount of power to, well, smelt aluminum. Due to the square-cube law the aluminum takes a really long time to cool back down. When you are trying to heat something from 20C to 700C it doesn't really matter if you take a break in the middle of the heating cycle. Sure, it might cool down from 500C back to 450C, but who cares?
As long as they get paid enough to make up for their workers getting an unscheduled paid day off, the smelter really isn't going to complain too much.
An aluminum smelter uses a humongous amount of power to, well, smelt aluminum. Due to the square-cube law the aluminum takes a really long time to cool back down. When you are trying to heat something from 20C to 700C it doesn't really matter if you take a break in the middle of the heating cycle. Sure, it might cool down from 500C back to 450C, but who cares?
As long as they get paid enough to make up for their workers getting an unscheduled paid day off, the smelter really isn't going to complain too much.
Not only that... they were "paid" in credits. Not USD as many suggest. Though more likely they don't understand what that means.
For a crypto miner, electric credits are as good as cash.
No, they aren't. Credits aren't as liquid or fungible. You can't pay for equipment, salaries, vendors, etc. in energy credits.
I know it's common to have contracts that are cheaper per Kwh, but have requirements to not draw if/when high load occurs. Is it really common to pay to not draw power?
For grid stability yes. The ideal grid stability solution is a customer who can sink AND source large amounts of power for a reasonable amount of time and very quickly. Like a giant battery farm can, although these are quite expensive.
Yeah but aluminum is useful.
So is a distrubuted, trustless system allowing transferring information securely. It shouldn't take much to see how this is solution is valuable to certain problems.
> So is a distrubuted, trustless system allowing transferring information securely.
Not really. NIST did a really good introductory blockchain technology whitepaper in 2018:
* https://csrc.nist.gov/pubs/ir/8202/final
Figure 6 is a decision flowchart laying out whether blockchain or some other technology would be best for various needs / use cases, and blockchain is very niche:
* https://twitter.com/justincormack/status/1048875246437195778
Needing both distributed and trustless simultaneous isn't common, especially in the modern, connected world:
> Reputation is a technology. We solve a lot of problems in the economy by turning one-shot games, in which people literally take the money and run, and turn them into repeated games, in which people have an incentive to behave well in this period so that you will trust them next period, and so on, into the future.
* Paul Krugman, https://www.youtube.com/watch?v=Y_IYGeZLLhI&t=10m30s
Not really. NIST did a really good introductory blockchain technology whitepaper in 2018:
* https://csrc.nist.gov/pubs/ir/8202/final
Figure 6 is a decision flowchart laying out whether blockchain or some other technology would be best for various needs / use cases, and blockchain is very niche:
* https://twitter.com/justincormack/status/1048875246437195778
Needing both distributed and trustless simultaneous isn't common, especially in the modern, connected world:
> Reputation is a technology. We solve a lot of problems in the economy by turning one-shot games, in which people literally take the money and run, and turn them into repeated games, in which people have an incentive to behave well in this period so that you will trust them next period, and so on, into the future.
* Paul Krugman, https://www.youtube.com/watch?v=Y_IYGeZLLhI&t=10m30s
Such a system would be extremely useful but Bitcoin is terrible at all those things.
It’s so slow and cumbersome that the vast majority of use is via banks, making it effectively quite centralized. It’s so volatile it’s a horrible currency so its main role is gambling and speculation. It’s pretty useless as the basis for decentralized communication or storage systems for a variety of reasons.
The only decentralize app even tangentially related to Bitcoin I’ve ever seen be marginally useful is Nostr, and Bitcoin has nothing to do with how that works. It’s tacked on as an afterthought. Nostr is just a straight up relay-facilitated peer to peer network with cryptographic addressing.
If anything the presence of Bitcoin makes Nostr worse by flooding it with spam, endless chatter about sending Satoshis, and stupid laser eye memes.
It’s so slow and cumbersome that the vast majority of use is via banks, making it effectively quite centralized. It’s so volatile it’s a horrible currency so its main role is gambling and speculation. It’s pretty useless as the basis for decentralized communication or storage systems for a variety of reasons.
The only decentralize app even tangentially related to Bitcoin I’ve ever seen be marginally useful is Nostr, and Bitcoin has nothing to do with how that works. It’s tacked on as an afterthought. Nostr is just a straight up relay-facilitated peer to peer network with cryptographic addressing.
If anything the presence of Bitcoin makes Nostr worse by flooding it with spam, endless chatter about sending Satoshis, and stupid laser eye memes.
I don’t think a qualitative measure of “valuable” is too helpful when it’s consuming a very quantitative 120+ terawatts annually
Bitcoin uses less electricity than christmas lights. Are they more "valuable"? How do you decide?
How do you figure? Bitcoin uses 116 TWh of energy each year, and that number is going up. The US, probably the largest user of Christmas lights in aggregate (probably more than Europe and South America put together, with Asia and Africa having negligible Christmas light usage), used 6.63 TWh of energy on Christmas lights in 2007 when only 5% of them were LEDs, and that number is dropping as people switch to energy efficient LED lights, which use 10% as much energy as incandescent Christmas lights.
That number is way low. I'm guessing it's only looking at indoor lights and not outdoor displays. Holiday lights in the US used 130 TWh in 2016.
Where does that number come from? Mine comes from the DOE. https://www.energy.gov/sites/default/files/maprod/documents/...
I’ve never done anything useful with Christmas lights or bitcoin. But Christmas lights are much prettier to look at than bitcoins.
Is that true though, globally, over the course of a year, mining + transactions?
I'd like to see the data sources and working on that.
As for decisions, a WWII European economy would have ditched both as excessive and not contributing to the current situation.
I'd like to see the data sources and working on that.
As for decisions, a WWII European economy would have ditched both as excessive and not contributing to the current situation.
Does it? that seems at odds with million dollar server farms of ASICs just mining Bitcoin all day, but I haven't logged into Xmas lighting power draw
That claim seems to have originated in this article from Mawson Infrastructure Group [1]:
> According to a recent post by the Center for Global Development, Christmas lights consume a whopping 6.63 terawatt-hours in the US alone, with the global figure being substantially higher across the rest of the world. The tiny state of El Salvador, for instance, consumes 5.35 terawatts for Christmas lights, and the African nation of Tanzania stands at 4.81 terawatts. Add these figures up all over the globe and you soon discover that bitcoin isn’t the energy consumer that many environmentalists lead people to believe – at least not compared to Christmas lights.
This is the post from the Center for Global Development that Mawson is using [2]. The relevant paragraph says:
> A 2008 study from the US Energy Department’s Energy Information Administration (EIA) found that decorative seasonal lights accounted for 6.6 billion kilowatt hours of electricity consumption every year in the United States. That’s just 0.2% of the country’s total electricity usage, but it could run 14 million refrigerators. It’s also more than the national electricity consumption of many developing countries, such as El Salvador, Ethiopia, Tanzania, Nepal, or Cambodia.
That is followed by a chart that shows the national consumption of El Salvador, Ethiopia, Tanzania, Nepal, and Cambodia, and the decorative holiday light consumption of the US.
Mawson misunderstood that chart. The 5.35 tWh for El Salvador and 4.81 tWh for Tanzania that Mawson cites as for Christmas lights are those country's total consumption.
In other words, what Mawson has shown is that the electricity used for Christmas lights in the US plus the electricity used for everything in the rest of the world (including Bitcoin!) is more than the electricity used for Bitcoin.
Let's see if we can figure out the right numbers. First, just for the US. Most figures I have found are for cryptocurrency in general not just Bitcoin, but I believe that Bitcoin is by far the biggest proof of work cryptocurrency so we probably won't be far off to attribute almost all of it to Bitcoin. Adjust all the following by whatever you think the ratio of Bitcoin to total cryptocurrencies is.
Annual US electricity use is about 4 x 10^12 kWh per year [3].
US Bitcoin electricity use is 0.9 - 1.7% of US electricity use [4]. Going with the bottom of the range that is 3.6 x 10^10 kWh per year.
The CGD number of holiday lighting, 6.63 x 10^9 kWh per year, were from 2008. Assuming that the proportion of electricity that goes to holiday lighting is about the same, that would be about 5% more now.
That puts US Bitcoin electricity use at least 5x higher than US Christmas light electricity use, and possibly up to 10x.
For the whole world, total electricity consumption is around 2.3 x 10^13 kWh per year [5].
The US is estimated to host about 1/3 of the crypto operations [4], so the global Bitcoin energy use should be about 3x that of the US or about 1.1 x 10^11 kWh per year, which is about 0.5% of world electricity use.
I haven't found anything on Christmas lights around the world. In the US, Christmas lights are around 0.2% of the total. If we assume that is the case for the rest of world that would give crypto uses 2.5 - 5x the Christmas light usage.
Also note that the Christmas light data for the US was from 2008. I'd expect Christmas lights nowadays to be much more likely to be LEDs, which would use less power.
[1] https://mawsoninc.com/bitcoin-mining-uses-less-energy-than-c...
[2] https://www.cgdev.org/blog/us-holiday-lights-use-more-electr...
[3] https://www.statista.com/statistics/201794/us-electricity-co...
[4] https://www.whitehouse.gov/ostp/news-updates/2022/09/08/fact...
[5] https://www.iea.org/reports/electricity-information-overview...
> According to a recent post by the Center for Global Development, Christmas lights consume a whopping 6.63 terawatt-hours in the US alone, with the global figure being substantially higher across the rest of the world. The tiny state of El Salvador, for instance, consumes 5.35 terawatts for Christmas lights, and the African nation of Tanzania stands at 4.81 terawatts. Add these figures up all over the globe and you soon discover that bitcoin isn’t the energy consumer that many environmentalists lead people to believe – at least not compared to Christmas lights.
This is the post from the Center for Global Development that Mawson is using [2]. The relevant paragraph says:
> A 2008 study from the US Energy Department’s Energy Information Administration (EIA) found that decorative seasonal lights accounted for 6.6 billion kilowatt hours of electricity consumption every year in the United States. That’s just 0.2% of the country’s total electricity usage, but it could run 14 million refrigerators. It’s also more than the national electricity consumption of many developing countries, such as El Salvador, Ethiopia, Tanzania, Nepal, or Cambodia.
That is followed by a chart that shows the national consumption of El Salvador, Ethiopia, Tanzania, Nepal, and Cambodia, and the decorative holiday light consumption of the US.
Mawson misunderstood that chart. The 5.35 tWh for El Salvador and 4.81 tWh for Tanzania that Mawson cites as for Christmas lights are those country's total consumption.
In other words, what Mawson has shown is that the electricity used for Christmas lights in the US plus the electricity used for everything in the rest of the world (including Bitcoin!) is more than the electricity used for Bitcoin.
Let's see if we can figure out the right numbers. First, just for the US. Most figures I have found are for cryptocurrency in general not just Bitcoin, but I believe that Bitcoin is by far the biggest proof of work cryptocurrency so we probably won't be far off to attribute almost all of it to Bitcoin. Adjust all the following by whatever you think the ratio of Bitcoin to total cryptocurrencies is.
Annual US electricity use is about 4 x 10^12 kWh per year [3].
US Bitcoin electricity use is 0.9 - 1.7% of US electricity use [4]. Going with the bottom of the range that is 3.6 x 10^10 kWh per year.
The CGD number of holiday lighting, 6.63 x 10^9 kWh per year, were from 2008. Assuming that the proportion of electricity that goes to holiday lighting is about the same, that would be about 5% more now.
That puts US Bitcoin electricity use at least 5x higher than US Christmas light electricity use, and possibly up to 10x.
For the whole world, total electricity consumption is around 2.3 x 10^13 kWh per year [5].
The US is estimated to host about 1/3 of the crypto operations [4], so the global Bitcoin energy use should be about 3x that of the US or about 1.1 x 10^11 kWh per year, which is about 0.5% of world electricity use.
I haven't found anything on Christmas lights around the world. In the US, Christmas lights are around 0.2% of the total. If we assume that is the case for the rest of world that would give crypto uses 2.5 - 5x the Christmas light usage.
Also note that the Christmas light data for the US was from 2008. I'd expect Christmas lights nowadays to be much more likely to be LEDs, which would use less power.
[1] https://mawsoninc.com/bitcoin-mining-uses-less-energy-than-c...
[2] https://www.cgdev.org/blog/us-holiday-lights-use-more-electr...
[3] https://www.statista.com/statistics/201794/us-electricity-co...
[4] https://www.whitehouse.gov/ostp/news-updates/2022/09/08/fact...
[5] https://www.iea.org/reports/electricity-information-overview...
The Watt is a unit of instantaneous power. Watt-hours are a unit of energy.
Ethereum is doing that without Proof of Waste.
Because there are no fundamental tradeoffs between the strategies. Because if it was so objectively more efficient and effective every project wouldn't move to that strategy.
it's waste if it's not used
Yes, it has been a massive boon to scammers, fraudsters and extortionists worldwide. It's paramount we keep it up!
Which problems today have been solved by blockchain? It's been around over a decade yet I still do not know of any noteworthy problem it has solved. What is a problem we had in 2009 that no longer exists due to blockchain?
Or, they could have implemented surge pricing and charged 10x, 100x for electricity during high demand for non-essential businesses and let market forces decide who wants to still use electricity.
Some lobbyists must have lined some pockets.
Some lobbyists must have lined some pockets.
That is basically exactly what is happening here, though. A bitcoin business considers itself non-essential and gives up power availability for a lower electricity price.
The issue with "pure" surge pricing is that it doesn't solve the actual problem. Most businesses - even "non-essential" ones will have fixed-price electricity contracts. They are not affected by surge pricing on the electricity market and have zero incentive to reduce their usage even when the market price spikes 100x.
In practice the "market forces" here are a bunch of Wall Street investors losing a shitton of money because their bets went sideways, but that is not going to fix the grid itself.
The issue with "pure" surge pricing is that it doesn't solve the actual problem. Most businesses - even "non-essential" ones will have fixed-price electricity contracts. They are not affected by surge pricing on the electricity market and have zero incentive to reduce their usage even when the market price spikes 100x.
In practice the "market forces" here are a bunch of Wall Street investors losing a shitton of money because their bets went sideways, but that is not going to fix the grid itself.
It's funny because it effectively reduces to Texas paying carbon offsets : )
Isn't this just getting compensated for being bumped off a plane?
I have read this before. Like a month ago.
Pretty sure it was in Texas. Hmm
https://news.ycombinator.com/item?id=37418866
https://news.ycombinator.com/item?id=37428605