{"items": [{"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689818311452", "anchor": "fb-689818311452", "service": "fb", "text": "Is highly inefficient? It would not be sustainable as the value of bitcoins drop? It would be environmentally wasteful to run that many computers?
If, you, Jeff, want to open up a server farm large enough , say one that took at least 5 megawatts of power, and were able to shut down your machines through near instantaneous responses, you could hypothetically sign up to be a demand response resource for the New England power grid.", "timestamp": "1411995335"}, {"author": "Warren", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689818700672", "anchor": "fb-689818700672", "service": "fb", "text": "That's..... just crazy enough to work....", "timestamp": "1411995532"}, {"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689818840392", "anchor": "fb-689818840392", "service": "fb", "text": "I mean, we could probably do the math. How many watts per hour to run one server? How many bitcoins per hour per server? What is the average wholesale energy price in New England megawatt hour?", "timestamp": "1411995636"}, {"author": "Benjamin", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689819838392", "anchor": "fb-689819838392", "service": "fb", "text": "My first reaction is that that is an incredible waste of energy, using up valuable energy resources to produce something with an entirely artificial value. But if the flexible plants are really so inefficient, then something like this does seem like a good idea, so long as what you produce is something that is be needed and desirable on its own, stores well, and for which we are not likely to produce a surplus we don't know what to do with. Bitcoin would be my last choice, partly because I don't see any way in which the world is improved by mining a bitcoin, but also because of all this: http://www.antipope.org/.../why-i-want-bitcoin-to-die-in...", "timestamp": "1411996289"}, {"author": "Adrian", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689819928212", "anchor": "fb-689819928212", "service": "fb", "text": "I assume the present market equilibrium would make any incremental investment in miners yield zero return.", "timestamp": "1411996376"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689820746572", "anchor": "fb-689820746572", "service": "fb", "text": "@Adrian: The key is that by being willing to match your demand exactly to what is most efficient to supply at that minute you should be able to get cheaper power than the current market equilibrium is based on.", "timestamp": "1411996875"}, {"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689820811442", "anchor": "fb-689820811442", "service": "fb", "text": "Also, Jeff, can you please cite a source to back up your assertion that flexible plants are significantly more expensive than inflexible ones?", "timestamp": "1411996921"}, {"author": "Mark", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689820916232", "anchor": "fb-689820916232", "service": "fb", "text": "Let us assume that mining Bitcoin is profitable while paying retail prices for electricity. Factoring in transmission costs and the power company's profits on retail electricity, it follows that mining Bitcoin with the direct electrical output from a power plant is more profitable than selling that power to consumers. When our cities go dark and the power companies have all the Bitcoin, I'll know who to thank.", "timestamp": "1411996975"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689821335392", "anchor": "fb-689821335392", "service": "fb", "text": "@Mark: Perhaps the long term equilibrium for bitcoin is mines in the far north (cold) with their own nuclear power plants?", "timestamp": "1411997299"}, {"author": "David German", "source_link": "https://plus.google.com/111229345142780712481", "anchor": "gp-1411997619213", "service": "gp", "text": "What are the startup time requirements?", "timestamp": 1411997619}, {"author": "Chris", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689822188682", "anchor": "fb-689822188682", "service": "fb", "text": "Benjamin: My first reaction was also that it was a big waste of energy, but then I realized that it also generates carbon when we create paper money. I'm not sure which is worse.
I do feel like there needs to be some way that this excess electricity can make the world better. I suppose there's always earn to give.", "timestamp": "1411997904"}, {"author": "Ira", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689823161732", "anchor": "fb-689823161732", "service": "fb", "text": "Eric might know a bit about that!", "timestamp": "1411998585"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689823216622", "anchor": "fb-689823216622", "service": "fb", "text": "Note that generating coins with cheaper electricity should drive some people with higher electricity costs out of the generating market.", "timestamp": "1411998621"}, {"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689825896252", "anchor": "fb-689825896252", "service": "fb", "text": "Jeff, did you read Ben's link? It touches on the high environmental impact of mining coins and the unsustainable nature.", "timestamp": "1411999320"}, {"author": "Will", "source_link": "https://plus.google.com/115623320734956550423", "anchor": "gp-1412001366456", "service": "gp", "text": "What are the environmental costs of running our power plants at full capacity all the time? You mentioned that the \"base load\" plants tend to be more efficient than the variable load plants. Is the efficiency enough to offset pollution from additional production?", "timestamp": 1412001366}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689832642732", "anchor": "fb-689832642732", "service": "fb", "text": "@Ron: What I've written up here is a way to make money while decreasing bitcoin's carbon footprint.", "timestamp": "1412002705"}, {"author": "opted out", "source_link": "#", "anchor": "unknown", "service": "unknown", "text": "this user has requested that their comments not be shown here", "timestamp": "1412004125"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689836799402", "anchor": "fb-689836799402", "service": "fb", "text": "@Elliot: it's also useful for lower transaction fees and selling things in places where fraud is very common.", "timestamp": "1412005322"}, {"author": "Victor", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689837802392", "anchor": "fb-689837802392", "service": "fb", "text": "I can't tell if this is a serious discussion or a parody.", "timestamp": "1412006090"}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1412006197130", "service": "gp", "text": "Presumably power plants already sell discounted power at off-peak times, so the balance between flexible/always-on plants is already a result of how much use for off-peak power institutions have. This would just be a special case, right?", "timestamp": 1412006197}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1412006310997", "service": "gp", "text": "... which is similar to what you said, except instead of thinking of it as a \"startup that offers a service to power companies\" I'm thinking of it as a customer that buys cheap power at off-peak times [just like many other customers]\".", "timestamp": 1412006310}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1412006466234", "service": "gp", "text": "Note that the risk issues in investing in more inflexible power plants are the same whether the power company mines on its own or sells power to a miner. Confidence that mining will remain profitable (for you as the power plant) is the same as confidence that your mining customers will continue to have demand for cheaper off-peak power.", "timestamp": 1412006466}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1412006785196", "service": "gp", "text": "As somewhat of an aside, this is a nice piece on the economics of mining:\u00a0\nhttps://blog.ethereum.org/2014/06/19/mining/", "timestamp": 1412006785}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1412007842468", "service": "gp", "text": "@David Chudzicki\n\u00a0Just buying power off-peak is ok, but being willing to have the power company tell you exactly how much power to draw at a given time should be even better.", "timestamp": 1412007842}, {"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689840217552", "anchor": "fb-689840217552", "service": "fb", "text": "Jeff, what you propose essentially increases carbon footprint. Mining bitcoins is carbon dirty. Having excess capacity running during all non-peak times to mine bitcoins is, by definition, much more carbon dirty than having capacity on hand only utilized when needed at peak times.
Also... I haven't even gotten past that to other aspects, like how this would achieve the opposite market signals as desired in a deregulated power grid.
I'm also still looking for you to cite a source for your fundamental claim that on-demand, flexible resources are significantly more expensive than regular ones.
Also, I might add that your idea of putting nuclear power far away from energy consumers doesn't work. Your whole idea is to ramp up/down bitcoin production on demand. Nuclear power is the *slowest* energy fuel to ramp production up or down. If the plants are far from civilization, and you have a demand spike, so you shut your 650mw (1 nuke generator unit) of bitcoin all at once, suddenly you have 650mw of generation in the middle of nowhere, needed to rapidly make it to civilization where the power is needed. That is an expensive transmission line proposition.
In short: demand response needs to be located near where consumption is happening to be effective. Your bitcoin mines should be in or near cities.", "timestamp": "1412007892"}, {"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689840287412", "anchor": "fb-689840287412", "service": "fb", "text": "There is no clean way to mine bitcoins. Processing power per bitcoin doesn't change with a different fuel.", "timestamp": "1412007983"}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1412008074691", "service": "gp", "text": "I'm not sure why that would be better than the power company setting a spot-price for power, and the miner deciding how much to pay for?\n
\n
\nUnless you mean that they'd have a long-term agreement to sell to sell a certain amount of power at a certain price. That's valid, but forces discussion of each party's risk profile, credit worthiness, etc. (which is really the crux of the issue, I believe).", "timestamp": 1412008074}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1412008331922", "service": "gp", "text": "@David Chudzicki\n\u00a0Another way of putting it is I'm not really interested in the day/night difference or any other predictable differences. \u00a0That's all relatively easy to price. \u00a0I'm interested in the short timescale matching demand and supply.", "timestamp": 1412008331}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1412008555033", "service": "gp", "text": "(I wonder how things work at that timescale now?) I guess my argument was assuming a price that adjusts on such a short time scale, with buyers adjusting their usage on that same short time scale based on the price.\n
\n
\n...which is probably not how it actually works. But again--not sure why that would be worse than your proposal.", "timestamp": 1412008555}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689841210562", "anchor": "fb-689841210562", "service": "fb", "text": "@Ron: \"Having excess capacity running during all non-peak times to mine bitcoins is, by definition, much more carbon dirty than having capacity on hand only utilized when needed at peak times.\"
It depends how much less efficient your flexible plants are than your base plants, and what the source of the electricity was for the miners you displaced.
\"I'm also still looking for you to cite a source for your fundamental claim that on-demand, flexible resources are significantly more expensive than regular ones.\"
The \"peaking power plant\" wikipedia article is a good place to start [1]. Aside from hydropower, which is awesome but pretty much built out, you generally have a tradeoff between efficient plants (coal, nuclear) and flexible plants (oil, natural gas). Another way to look at it is that pretty much the only reason people are willing to put up with inflexible plants is that they're cheaper. Otherwise we'd just run flexible plants.
\"I might add that your idea of putting nuclear power far away from energy consumers doesn't work. Your whole idea is to ramp up/down bitcoin production on demand.\"
Sorry, that comment was considering an entirely different outcome, where bitcoin producers run their own power production and match their consumption exactly to the power produced. It's a far-future proposal; nukes take ages to build.
[1] http://en.wikipedia.org/wiki/Peaking_power_plant", "timestamp": "1412008722"}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1412009675684", "service": "gp", "text": "@David Chudzicki\n\u00a0\"But again--not sure why that would be worse than your proposal.\"\n
\n
\nWouldn't that allow arbitrage, where you have one circuit that has normal constant-price metering and another that has real-time prices? \u00a0The real-time one would usually be cheaper, but when it's not you just draw power from the other circuit instead?", "timestamp": 1412009675}, {"author": "David Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689842917142", "anchor": "fb-689842917142", "service": "fb", "text": "Jeff, if coal is \"efficient\" while oil & natural gas are \"flexible\", it sounds like that might be pointing to a huge difference between cost-efficiency (electricity per dollar) and greenhouse-gas efficiency (electricity per carbon), or other kinds of electricity-per-environmental-impact efficiency. That difference might be the cause of some confusion here.", "timestamp": "1412009707"}, {"author": "David Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689842972032", "anchor": "fb-689842972032", "service": "fb", "text": "If your proposal is actually a way to make coal (much less clean than natural gas) more cost effective, that puts things in a different light.", "timestamp": "1412009757"}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1412010269269", "service": "gp", "text": "It's not arbitrage unless you can sell the power back at the real-time price when constant is cheaper, right? (which you couldn't)\n
\n
\nAlso, I guess you'd only make the real-time one available to such customers?", "timestamp": 1412010269}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1412011476458", "service": "gp", "text": "@David Chudzicki\n\u00a0Why wouldn't you be able to sell it at the realtime price? \u00a0Isn't there a network of buyers and sellers?", "timestamp": 1412011476}, {"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689846689582", "anchor": "fb-689846689582", "service": "fb", "text": "Jeff, I work for the power company. I'll go through the Wikipedia article, but what you're saying is not really true. Gas has become very efficient and flexible.
And regardless, mining bitcoins is simply an exercise in trading fossil fuel for crypto into funny money.
People don't mine bitcoins to make money in an energy efficient or even financially efficient way ... they do it because they want untraceable money.", "timestamp": "1412012303"}, {"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689847822312", "anchor": "fb-689847822312", "service": "fb", "text": "Read the article. Some problems with your assessment:
1. Peaker plants are not the only flexible plants, so this article is not a good source to defend your claim that flexible plants are less efficient / etc. Many plants can ramp up or down, but are not peaker plants, who only run a handful of times per year.
2. The article specifically says, under \"Types\" ... \"The US peaker plants are generally gas turbines...\". This directly contradicts with you saying about a tradeoff.
If you look at what's being built in deregulated areas, it's mainly natural gas combined cycle plants. They're flexible, can use oil as a backup, and efficient. The old coal and oil plants are not financially viable to build in today's energy situation.
In short: there is no efficiency gap as you require to make bitcoin mining remotely economically feasible as the scheme you've described. And even if it was, bitcoin mining is dirty. And even if it wasn't, implementing on the scale you would need (10s of 1000s of mw between baseload and peak) would mean radical changes to both the transmission system and the deregulated market mechanisms.", "timestamp": "1412012917"}, {"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689849109732", "anchor": "fb-689849109732", "service": "fb", "text": "David, it doesn't make coal more efficient. It would mean coal plants, normally idle most of the year, would be running all year.", "timestamp": "1412013814"}, {"author": "David Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689850511922", "anchor": "fb-689850511922", "service": "fb", "text": "Ron-- how could that not make them more (economically) efficient? I'm aware that it wouldn't make them more efficient in other senses. That's my point above.", "timestamp": "1412014108"}, {"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689923730192", "anchor": "fb-689923730192", "service": "fb", "text": "David, because we don't live in a vacuum. Even if it made those plants more viable, it'd likely be a storm of environmentalists raising hell. And, again, we haven't even discussed market or regulatory implications.
When would planned work be done if we had no excess capacity year round?
How would government at all levels respond?
How would states and other fuels respond in demanding rules changes?
Ultimately, how would this not simply be carte blanche for any generator to crank up all units to max, burn whatever they can, since effectively, ratepayers like you and me would effectively subsidizing burning fossil fuels for a miniscule return in digital money that may not even exist in a few years?", "timestamp": "1412022128"}, {"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689924294062", "anchor": "fb-689924294062", "service": "fb", "text": "Jeff - don't take this personally. I like a lot of your ideas you post. The nature of FB means that you will only see a \"Like\" for your good ideas, and lots of discussion for more iffy ideas", "timestamp": "1412022373"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689925112422", "anchor": "fb-689925112422", "service": "fb", "text": "@Ron: One issue here is that my understanding of the economics sounds like it's out of date. Natural gas got cheap (fracking) which means now we're really only running coal plants because they're already built and cleared to run.", "timestamp": "1412022777"}, {"author": "David Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689925875892", "anchor": "fb-689925875892", "service": "fb", "text": "Ron, Jeff-- those economics are somewhat unique to the US. Gas isn't a global market (unlike oil).", "timestamp": "1412023229"}, {"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689927407822", "anchor": "fb-689927407822", "service": "fb", "text": "Jeff ... maybe, but what I've been trying to express is that there are far more factors at play than your original post acknowledged.
Another thought: Why bitcoin? Why not pick something else energy intensive that gives more bang for the buck, isn't horribly dirty, and can be turned off rapidly thus acting as demand response? Though, now that I'm challenging myself to think of one, I can't. Maybe this is just a difficult problem that can't be solved.", "timestamp": "1412024045"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689928350932", "anchor": "fb-689928350932", "service": "fb", "text": "@Ron: \"Why bitcoin?\"
Bitcoin is the only market for electricity-limited flexible computation I know of.", "timestamp": "1412024195"}, {"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689940870842", "anchor": "fb-689940870842", "service": "fb", "text": "There are supercomputing farms for hire. They could be dual purposed as demand response. But I think the sheer number of computers it would take to make even a single plant more efficient is enormous, let alone balance the grid on a broader scale.", "timestamp": "1412026717"}, {"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689941020542", "anchor": "fb-689941020542", "service": "fb", "text": "1 server at what? 1 kilowatt? So that's over 500,000 servers for one plant. Millions to make a significant dent in grid balancing, and it will drive electricity prices way up.
There's just no good way to do this.", "timestamp": "1412026864"}, {"author": "Neil", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689943685202", "anchor": "fb-689943685202", "service": "fb", "text": "You may as well use the extra electricity to create cool electric arc light shows, for all the value that you get from mining bitcoins.", "timestamp": "1412028492"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689943864842", "anchor": "fb-689943864842", "service": "fb", "text": "@Neil: You can sell the bitcoins you generate and make back a large fraction of the cost of the electricity you put into them.", "timestamp": "1412028649"}, {"author": "Neil", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689943929712", "anchor": "fb-689943929712", "service": "fb", "text": "You could sell tickets to your light shows too. But don't mind me, I have an axe to grind about the stupidity of bitcoin.", "timestamp": "1412028749"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689945137292", "anchor": "fb-689945137292", "service": "fb", "text": "@Ron: \"that's over 500,000 servers for one plant\"
These aren't really \"servers\". They're little special-purpose chips. And yes, you could potentially run a lot of them: http://www.thecoinsman.com/.../inside-one-worlds-largest.../
And here's where my understanding of the economics is shaky, but from talking to a mathematics PhD student who is working with an electric utility for his research I think this becomes useful at much smaller levels than \"a whole plant's worth\". The target situation is when there are plants that are capable of producing at multiple output levels and are usually turned up and down a lot over short timescales in ways that hurt their efficiency. This would let them be adjusted less often, keeping them a little on the high side and making up the slack with bitcoin generation.", "timestamp": "1412029432"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689945262042", "anchor": "fb-689945262042", "service": "fb", "text": "@Neil: Sounds like it would be hard to sell tickets to a light show that might or might not happen based on whether there was an awkward level of power demanded at the time.", "timestamp": "1412029506"}, {"author": "Andrew", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689945761042", "anchor": "fb-689945761042", "service": "fb", "text": "You could do this by operating a farm yourself in Iceland.", "timestamp": "1412029862"}, {"author": "Ron", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689948784982", "anchor": "fb-689948784982", "service": "fb", "text": "Jeff I still disagree with the concept and feasibility, but even more fundamentally... don't you have a problem suggesting we run *more* dirty fossil fuel plants just simply to trade capacity for demand response? Seriously ... where's the upside?", "timestamp": "1412032127"}, {"author": "Eric", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689949104342", "anchor": "fb-689949104342", "service": "fb", "text": "Thanks, James Ira. I feel obligated to clear up some misconceptions. TL;DR version: BTC is not as carbon-intensive as its detractors claim, but the economics of this business model do not work; ASICs are too expensive to hedge short-term shortfalls in electricity demand, and BTC is too volatile to hedge long-term shortfalls in demand.
My company, www.toom.im runs bitcoin-miner hosting in E. Washington State. Why? (1)EWA has ~ .02$/kWh, nearly the cheapest energy in the world, (2) it's all hydroelectrically sourced, e.g. close to operationally carbon-neutral. Iceland (geothermal) is close, at $.05/kWh. We could make bitcoin nearly carbon neutral within a year or two if we all wanted to, at negligible economic cost. So why the fuck do people still mine BTC using dirty expensive coal? Because they haven't seen our ads on Bitcointalk, yet, I guess. :)
But, Why is the power in EWA so cheap? The state and municipalities built dams for agriculture and power for locals, but never invested in high-voltage lines required to inexpensively sell power to grids in other parts of the country. So there's a huge mismatch between local demand and potential supply. YET, we are currently capped at 1 MW (with half that online) and expanding to 3 MW by the end of 2014. Why are we capped? Because the local substations (required to transform that power so we can use it, and are very expensive and take 1-2 years to build) are at their capacity, from the huge influx of businesses like ours. All miners are built to take wall power or less, so <= 260 V, about. The \"low voltage\" lines that transport power to substations? 115,000 V. (Long-distance \"high-voltage\" are 500 kV) Even post-substation distribution happens at 33 kV or so, hence the transformers in your neighborhood.
So: to use up \"spare power\" the Utility needs to put the BTC miners on the residential side of a substation. There is only a shortage of demand at the residential level if, e.g. Actual residential demand has fallen very short of projected residential demand. e.g. Detroit? To meet pre-substation supply with BTC-mining (e.g. by building a substation) is a huge gamble (Take a look at bitstamp.com). In the time it takes to build an extra substation you might not need, USD/BTC will either be at $1600+ or it will be at $.16, with about equal probability, depending on wether bitcoin becomes an adopted technology worldwide, or a bubble. Waaaay too risky a bet to justify building an extra substation (which is used for local distribution only); much better to just sell that un-met supply to other Utilities.
Okay so no good as a long-term hedge. What about soaking up supply at 3 AM in the summer? Short-term demand variability is the day to day stuff. Can BTC mining be profitably used as a hedge, here? I don't think so -- the economics don't work, because mining hardware is too expensive. The cheapest reliable miner in terms of upfront-cost per H/s is the AntMiner S3, which costs .58 BTC. In October I expect to make .16 BTC from my S3. I roughly expect to make 15% less in November, and so on every month after, due to increases in block difficulty, as more and more miners come online. So using geometric series with r=(1-.15), the lifetime ROI assuming free electricity is .16/.15, which is more than .58. Woohoo! I will almost certainly make a profit if electricity is free. What if I only run it half the time (e.g. Because I'm only running it on off-hours) -- you will note than .08/.15 < .58: I do not make back my initial investment, even with free electricity. Boohoo.
It's actually even worse than this, because even if you're only using off-peak power, the price isn't 0$, it's actually the price another Utility (e.g. In a big city or different timezone) is willing to pay for that power.
Sorry for the bad news, folks! But, seriously, if you're running a miner on electricity from coal or oil or LNG, you realize you're probably losing money? Go hydro-powered!", "timestamp": "1412032373"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689949872802", "anchor": "fb-689949872802", "service": "fb", "text": "@Eric: Thanks! Really useful to get feedback from someone who knows the numbers. So mining is much more hardware limited than electricity limited? Are hardware improvements still coming along quickly or do we expect this to slow down soon to where power matters more?", "timestamp": "1412032982"}, {"author": "Eric", "source_link": "https://www.facebook.com/jefftk/posts/689817977122?comment_id=689950651242", "anchor": "fb-689950651242", "service": "fb", "text": "No, we're getting close to diminishing returns on hardware deign improvements: there's only so much you can do to make SHA-256 happen faster. Also the hardware price isn't directly determined by how advanced the hardware is: Spondoolies probably has the mot advanced and efficient hardware (in term of H/ASIC and GH/J) -- their miners are more expensive than BITMAIN (who make the AntMiner): you can charge a premium for power density.", "timestamp": "1412033601"}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1412038131154", "service": "gp", "text": "I didn't think that just anyone could sell in that network. I guess also I was thinking that network \nonly\n has a variable price, and that variable price is always lower than the price consumers pay. Not sure if any of that if true though. ", "timestamp": 1412038131}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1412039537097", "service": "gp", "text": "@David Chudzicki\n\u00a0I'm pretty sure the true marginal cost spikes well above the consumer cost at times. \u00a0But I don't know if that's ever reflected in prices.", "timestamp": 1412039537}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1412046623881", "service": "gp", "text": "Yeah... with consumer prices set by regulators \u00a0(right?) I wouldn't be surprised. Makes it a hard market to reason about.", "timestamp": 1412046623}]}