Wikipedia:Reference desk/Archives/Science/2017 December 7

Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

December 7

Effect of CO2 emissions on the metabolic energy obtained from glucose

One of the basic biology factoids is that glycolysis + Kreb's cycle + oxidative phosphorylation gets "up to" 38 ATPs out of a glucose molecule.

It occurs to me that, just as the temperature of cooling water affects the efficiency of a power plant, the concentration of CO2 should affect the efficiency of respiration. Searching I found many articles on plants, but I wanted to look at the same from an animal point of view.

Back of the envelope calculation: carbon dioxide in Earth's atmosphere has gone from 280 to 407 ppm. Gibbs free energy includes a term for RT ln (407/280) = 8.134 J/K * 300 K * 0.374, which gets me 0.93 kJ per, I think, 1 mol of CO2 that crosses a membrane between these two concentrations. Since there are 6 CO2s per glucose, that gets 6.00 kJ of energy difference between the two conditions per starting glucose! If so, well, adenosine triphosphate cites a change in free energy of 3.4 kJ/mol, so that would be 1.64 fewer ATPs per starting glucose than in pre-industrial times. (Note this goes by ln, so it takes larger and larger increases in CO2 to reduce the ATP count further; a 1000000 ppm concentration would, by this calculation, reduce it by about ... 38 ATPs).

First, did I do something stupid in the calculation? I haven't used these chemical concepts this way before.

Next, is there a way to call shenanigans on it for external reasons? For example, the CO2 concentration in the lungs will be much higher on exhalation. However, I'm thinking that this might go up somehow in proportion to the air concentration ... actually, I have no idea one way or the other.

Could a decrease in ATP production tend to turn everyone into couch potatoes, make us feel like aerobic exercise is too hard, cause overeating and so on? Just woolgathering here. Wnt (talk) 02:34, 7 December 2017 (UTC)

This appears to say the first noticeable without equipment effects don't start till 1000 ppm. Sagittarian Milky Way (talk) 03:11, 7 December 2017 (UTC)

I think that type of value is determined by acute exposure rather than lifetime exposure. Also, note that it is possible to waste significant amounts of cellular energy (e.g. uncoupling agent) with relatively little perceived effect. Something like 2,4-Dinitrophenol could be marketed as a "dieting aid" (indicating substantial loss of energy), though it causes symptoms up to and including death at sufficient levels. If the math above is correct it describes less than a 5% reduction in energy from carbohydrates (less from fats because the H2O is not affected), which is scarcely noticeable for dieting purposes. Wnt (talk) 14:06, 7 December 2017 (UTC)

It'd be interesting to do an analysis of the rate of world record breaking in as many probably steroid-free less-team/people vs people (too many variables) outdoor aerobic sports as possible. I'm not sure if you could see a signal in the noise of numerous other factors i.e. once-a-generation+ x players causing more kids to want to be pro x players, and x players with sudden big improvements on the previous best are still appearing in major sports (like a man can be many inches too tall (Usain) and still pulverize 100m records, a man can shoot well while being bumped when no one else could (LeBron) and that huge Australian high school football player could possibly not be too slow to play NFL at that size. Are the outdoor more individual aerobic sports (i.e. marathoning) also still capable of having something like that happening in the near future screwing up trying to use their world record progression to see if something more than the usual slowdown of record-breaking is happening? If your math is correct it's probably been thought of by now and you'd probably hear about it at least as much as ocean acidification or runaway clathrates warming but it'd be nice to see research on someone kept at CO2s between 280 and 1000 for a long time. I now wonder if any important sports events had enough building-related CO2 or oxygen abnormality to affect the performances lol. Or if it'd be allowed to mess with the air of sports venues to increase your team's or event's performance or even favor the home team. Do the Denver Nuggets want to try installing airlocks and O2 injection during games? Does Tokyo want more world records in their games? Make them play in 40% O2 at 60% pressure! Sagittarian Milky Way (talk) 18:40, 7 December 2017 (UTC)

You problems are: (i) you do not take into account the limited efficiency of the Kreb's cycle. 38 ATPs contain much less energy than that is released during oxidation of 1 molecule of glucose. The remaining energy turns into heat. So, when you calculate balance of Gibbs (or free) energy you need to take this into account. (ii) In addition you seems to confuse Gibbs (or free) energy and energy itself. The free energy is only useful when you want to determine the direction of a reaction: will glucose oxidase or will it form instead!? On the other hand the true energy balance does not depend much on the external CO2 concentration. So, the Kreb's cycle consists of a fixed number of steps. Each step gives the fixed number of ATPs. This will not change as the CO2 concentration increases. The changes of the Gibbs energy that you refer to are essentially changes in entropy multiplied by temperature. Ruslik_Zero 20:47, 7 December 2017 (UTC)

@Ruslik0: The Krebs cycle involves a fixed amount of reduction and phosphorylation producing 10 NADH, 2 FADH2, and 4 ATP, it is true. But oxidative phosphorylation is less certain and is said to produce "up to" 34 ATP (3 per NADH and 2 per FADH2). There are costs in moving precursors and products (e.g. NADH, ATP) to keep them on hand to make the reaction go forward. Additionally, alternate mechanisms exist that generally have a lower yield. The result is that estimates tend to say things like "30 to 38 ATP", though it looks like eukaryotes can't hit the top figure.

Whenever there is a choice of regulated mechanisms, this means that the final energy figure has a chance to talk back to the reaction. For example, if ATP is produced abundantly in the mitochondrion I think it might be valuable to let it diffuse out through a channel, though I don't know that. But if anemic amounts are produced, a proton from the gradient might be spent wringing out ATPs from a less rich storehouse inside the organelle. Similarly, one of the alternate biochemical mechanisms might serve to add some extra free energy to the reaction to keep the process going. My memory of metabolism is limited, but it is surely not a rigid reaction that can be allowed to sputter to a halt based on a small change in free energy. Note that although a typical figure lists a metabolic efficiency under 40%, it is more like 50% in cells due to free energy considerations. [1] In other words, if ATP and ADP had exactly the same amount of energy in their chemical bonds, the cell has 10 times more ATP, so any reaction ATP + X -> ADP + Y would still be pushed forward, which means that that 10% of the energy was never actually lost to begin with. Now to be sure, yes, even if it is 50% I can say a 5% difference in ATP production is a 2.5% difference in efficiency, but it seems just as relevant no matter what I compare it to. Wnt (talk) 22:36, 7 December 2017 (UTC)

The Kerb's cycle efficiency can vary, of course, but it does not depend on the ambient CO2 partial pressure unless this pressure is extremely high. This is my point. Ruslik_Zero 20:26, 8 December 2017 (UTC)

If the free energy change was sufficient to extract nearly two extra ATP from the cycle when CO2 concentrations are low ... why didn't it evolve that way? The difference in free energy has to be reflected somehow, whether it is in the number of ATP produced or the maintenance of a high ATP/ADP ratio (which stores a large fraction of the output free energy as explained above) or some other form of chemical energy. Wnt (talk) 03:43, 11 December 2017 (UTC)

Natural satellites of planets

I glanced through the Wikipedia material on the satellites (or moons) of planets. It gives very specific numbers, all of them in the order of one to a few dozens, for the known satellites of Jupiter, Saturn, Uranus and Neptune. But it also says that there's no agreed-upon threshold for size that objects have to pass so they can be considered satellites. And there doesn't seem to be any specification that a satellite mustn't be located in the planet's ring system. So why don't we say instead that these planets have billions and billions of satellites, all but a handful of which are tiny ones found in their ring systems? --Qnowledge (talk) 07:33, 7 December 2017 (UTC)

You are right to include the qualifier "natural". If you look at space debris you will see the term "satellite" is usually reserved for larger objects. 92.27.49.50 (talk) 11:06, 7 December 2017 (UTC)

Also, see Natural satellite#Definition of a moon. Dolphin (t) 12:54, 7 December 2017 (UTC)

I'm not sure, but I think some of that has to do with the Roche limit. A ring system like Saturn's contains particles that AFAIK would continue to break up under tidal forces were they not held together by chemical forces, i.e. electromagnetism rather than gravity. I think such assemblages are reasonably disqualified from being moons. (I'm not sure at the moment what distinguishes ring particles from shepherd moons, should look up later if someone doesn't explain) Wnt (talk) 14:11, 7 December 2017 (UTC) (withdrawn - see below)

Probably being big enough to be seen and have a significant shepherding effect. Ring particles in this solar system are only the size of a house or smaller. Since shepherd moons are big enough to be seen by spaceprobes the Roche limit isn't absolute and natural satellites big enough to be seen by them and thus named and numbered can have enough structural integrity to avoid being broken up by tidal forces. Sagittarian Milky Way (talk) 17:44, 7 December 2017 (UTC)

A Ring system can be regarded as one or sometimes even multiple satellite entity for originating from a satellite that broke up but also in perspective to eventually bake together a new satellite, just like all other planets and satellites evolved. Its a placeholder for a potential satellite if you like and it also has a mass and motion fitting Kepler's laws of planetary motion. Its an ongoing exploration and discussion. You can read about it in Kuiper belt and Oort cloud which can be regarded satellites of our sun or orbits with multiple satelites, maybe even planets, without one body dominating the part. In that you can also find the distinction, what is generally counted as planet, satellite, meteorite of a star- or planetary system and what is not: The origin of its matter and its orbit. --Kharon (talk) 21:30, 7 December 2017 (UTC)

It appears my confidence in the Roche limit was misplaced. According to [2], "If you broke up all the satellites within the Roche limit of Neptune, you'd get a ring system that would not look too terribly different from Saturn's." So this clearly is not a criterion to say they aren't moons. It is a good reason to watch your step on Larissa (moon)! (N.B. [3] says Larissa is inside the Roche limit while our article says it will break up someday when it passes it. Since Roche limit depends on density I'm not sure good sources don't disagree, but can someone confirm an error?) Wnt (talk) 22:00, 7 December 2017 (UTC)

Astronomy is a very bad science that is completely unable or unwilling to keep theory and facts separately. Just some weeks ago they found a super massive black hole so old that its assumed formation process can not be added up with the age of our universe according to the big bang theory. So maybe the big bang theory is bullocks but no one dares to say it out loud!! Its a very mainstream centered science like economics where you are either neoliberal or outcast (to put it slightly exaggerated). So don't expect to much! --Kharon (talk) 22:22, 7 December 2017 (UTC)

Well, if it's a choice between bad science in astronomy and bad science in our current Wikipedia draft I know which I find a priori most likely. ;) Wnt (talk) 22:37, 7 December 2017 (UTC)

The Big Bang is not bollocks, it's how'd it get from ~380,000 years after the Big Bang to galaxies that's not well understood. Sagittarian Milky Way (talk) 22:49, 7 December 2017 (UTC)

That's a new record in Dogmatism. Now even whole (known) galaxies must be wrong if they don't fit the mainstream. Btw. the mainstream core argument is not the background radiation but the Redshift. --Kharon (talk) 23:16, 7 December 2017 (UTC)

I suggest you work on your reading comprehension and check again what Sag has written. He comments on our incomplete state of knowledge, not any galaxies that "must be wrong". And while the redshift was one piece of evidence for an expanding universe, there were competing models - see Steady State theory. The CMB, on the other hand, is very well explained by the big bang theory (it's the red-shifted image of the surface of last scattering), but does not fit into e.g. Steady State. --Stephan Schulz (talk) 00:52, 8 December 2017 (UTC)

This is way off topic, but I should note [4] describes the black hole and links the original paper (I didn't check Sci-Hub though). It's 10% younger than the previous record holder. There's something there about "episodic hyper-Eddington accretion". I have no idea, but my gut feeling is it seems odd to say that a region of twisted space 800 millions times the mass of the sun forming in 800 million years would be perfectly logical, but 690 million years buggers all belief. Wnt (talk) 00:11, 8 December 2017 (UTC)

Modified Atmosphere Article -Clarification Questions & Suggestions

Hi,I am no expert in this subject, but thought I would give some feedback in the hope of making the article easier to read & understand. I was unable by clicking the Talk button to access anything other than my account page User talk:Mpe123

severity of preparation (give example to make the meaning clearer) does it mean something like processing such as washing salad or grating carrots?

A paragraph/comparison chart outlining the difference between EMAP and MAP could be helpful,to me they sounded pretty similar.(respiring product,permeability,"an equilibrium modified atmosphere will be established in the package and the shelf-life of the product will increase.")

When gas flushing & compensated vacuum are 1st mentioned (paragraph above scientific terms) it would be good to have a note mentioning that more details are given further in the article.

Isn't a potato a vegetable? I checked 2dictionary definitions & it says that they are vegetables "An example of a gas mixture used for non-vegetable packaged food (such as crisps)"

"breathable" films called EMAP are mentioned at the beginning of the article, but later(packaging films section) are referred to as "MA/MH films" are they the same?

Thanks in advance! — Preceding unsigned comment added by Mpe123 (talk • contribs) 17:58, 7 December 2017 (UTC)

Article is Modified_atmosphere and as someone has already explained Mpe123 has hit the wrong Talk-Button. Just one thing: potatoes are vegetables (a living thing which emits and produces gases), crisps are definitely no potatoes and even less vegetables. 194.174.76.21 (talk) 18:24, 11 December 2017 (UTC) Marco Pagliero Berlin

Why isn't bitcoin mining causing the price of bitcoin to level off?

Bitcoin mining has now become hugely profitable at a bitcoin price of nearly $20,000. You could already make a modest profit at a bitcoin price of around $3000. So, why is the price going up and why isn't the bitcoin boom being accompanied by a boom in the sales of fast computer processors? Count Iblis (talk) 21:08, 7 December 2017 (UTC)

• Maybe it's a function of Bitcoin being a scam. ←Baseball Bugs ^{What's up, Doc?} carrots→ 21:25, 7 December 2017 (UTC)
• Bitcoin "mining" really is a misnomer. What is really going on is the creation of new blocks in the blockchain (each of which gives the successful creator a pre-defined reward). The bitcoin network automatically adjusts the difficulty of the block creation so that, on average, one new block is created every 10 minutes, independent of the total amount of compute power in the net. More miners spreads the same reward among more parties, but does not increase the supply. --Stephan Schulz (talk) 21:41, 7 December 2017 (UTC)

Your central assumption is wrong because the gratification for the mining is adaptive. There is no fixed rate like you would always get say 1 Bitcoin for solving/verifying a block. With the current high prize the gratification is most likely very very low now because there is only a limited number of blocks to be verified and many computers and computer pools try to solve one of these. Also if you add more computers and computer pools that is only more competition, more supply but not more demand, which is at the end essentially biting its own ass according to the rules of supply and demand. --Kharon (talk) 21:58, 7 December 2017 (UTC)

Are you replying to me or to the Count? The reward for mining is changing, but with a long-term predefined schedule - the reward is cut in half every 210000 "mined" blocks (which is once every couple of years). What is "adaptive" is the difficulty of creating a correct block (you need to find a nonce that will produce a hash with a given number of leading zeros, and that required number is adjusted to keep the rate of block creation roughly constant). --Stephan Schulz (talk) 22:11, 7 December 2017 (UTC)

No, i just saw an edit conflict and i did'nt want to rework it, in parts because of fearing to run into the next edit conflict and thus getting trapped into an adapting loop. --Kharon (talk) 22:31, 7 December 2017 (UTC)

• One must be aware that the mining of Bitcoin gets you two kinds of money: the "Bitcoin mining" part where you create new Bitcoins that you get to keep, and the "transaction fee" part where people pay miners to validate their transactions in priority. I think I had read somewhere that the latter is what really gets you the money these days, but the only semi-serious source I could find is [5], whose numbers do not allow to compare the recent mining/commission parts in a meaningful way. Tigraan^{Click here to contact me} 10:19, 8 December 2017 (UTC)

  While true, transaction fees don't increase the Bitcoin supply - they only move Bitcoin from one market participant to another. It may cause people to keep mining, even if the built-in rewards are no longer cost effective, but it has not direct effect on the supply/demand situation. --Stephan Schulz (talk) 11:27, 8 December 2017 (UTC)

  Well, it does impact the supply and the demand of Bitcoin mining even if the impact on the Bitcoin possession market is limited, and the OP is explicitly about mining. (Maybe my indentation choice was questionable, since I was not really replying to your post after all.) Tigraan^{Click here to contact me} 12:22, 8 December 2017 (UTC)

Alternatively, many economists make the case that the price is overwhelmingly speculative, and is therefore detached from supply-and-demand economics. The actual availability of the "resource" - constrained by mathematical details, or otherwise - is no longer a contributing factor the market-price at which people are buying and selling.

Furthermore, fully 100% of the "bitcoin-to-dollar" conversion price is a snapshot of a secondary market - not "some" or "most," but fully all of that price is sustained on such a market. And it is an entirely unregulated secondary market! So this means that price arbitrage can occur with catastrophically enormous price-spreads - ratios that would be orders of magnitude larger than any other conventional marketplace.

In my opinion, I think I have composed my explanation using certain technical terms that are more ... shall we say, precise than the word "scam," but to the informed investor, these descriptions ought to carry equal weight.

For even more verbosity on the topic, here's Susan Athey, an economist specializing in internet commerce: Bitcoin Pricing, Adoption, and Usage: Theory and Evidence (2016). She's written several well-researched commentaries on bitcoin over the year. She couches her statements in even more jargon: given "the presence of frictions arising from exchange rate uncertainty," ... "the idea of bubbles seems salient for Bitcoin..."

Again, the language is florid but, in my reading, the implications are equally lurid. ...Scam.

To put it more bluntly: if you want to invest in bitcoin, just try to put a non-trivial amount of money (let's say, U.S. Dollars) into an exchange on some proverbial Monday; wait for the price to vary by some non-trivial amount; and try to get your money back out on the proverbial Friday.

See, in a regulated market, they have to give you your money. In fact, as of right now, in the United States, starting in 2017, they have to give it to you within two business days: this is called T+2 and it dramatically changed the financial marketplace this year - even though it got almost no media coverage outside of specialist investment and economics publications! But Bitcoin exchanges follow no such regulatory oversight. They can arbitrate your withdrawal, at any price, on any schedule. You won't be able to withdraw your proverbial investment return of 5%, or 50%, or 50000%, because the exchange maker sets the schedule for paying you.

The exchanges that convert bitcoin to hard-currency are ponzi schemes. What you will find is that you might be able to pull a few hundred dollars of "earnings" out of them, at a massively inflated price (so that they can sucker in the next guy with unrealistic inflated growth statistics). But macro-economics does not work via "a few hundred dollars." Even a 50x growth in an investment of a few hundred dollars still won't buy you a private jet! As soon as you attempt to invest any nontrivial amount of money, and try to reap your well-invested earnings, you will find your arbiter mysteriously goes bust in a bank run. This has already happened multiple times, but new suckers keep buying!

Nimur (talk) 22:36, 7 December 2017 (UTC)

• I believe that @David Gerard: is something of an expert on cryptocurrencies. He may be able to weigh in here. --Jayron32 01:17, 8 December 2017 (UTC)

The "boom"-part is nothing specific to bitcoin or other cryptocurrencies. There was a similar development in the Shadow banking system and we all read about how some companies that serve the tax havens lost their pants lately and what became visible. Behind all of this is a world financial system which contains more wealth than the whole world industrial economy can craft in 100 years.[6] One obvious side effect of this is a flood of "investors" desperately trying to put their wealth somewhere "save". Even a 10-year German government bond with negative interest rates was sold out in hours. There is going to be a huge financial "bloodbath" somewhere again soon and bitcoin looks like build to survive it almost as save as German government bonds. --Kharon (talk) 03:15, 8 December 2017 (UTC)

A few exchanges, including Coinbase, one of the largest, are regulated. Of course, I have no idea what kinds of standards are enforced by the regulators or how strictly they are enforced. Also, although it's true the vast majority of Bitcoin holders hold Bitcoins through a broker, Bitcoin intentionally doesn't require this. You can run Bitcoin wallet software on your computer and transact directly with others. Of course, then you are taking on the settlement risk yourself. --47.157.122.192 (talk) 03:19, 8 December 2017 (UTC)

Purely addressing the computing aspects of this: it's impossible to make a profit anymore mining Bitcoin on general-purpose CPUs, assuming those are what you mean by "fast computer processors". All "serious" Bitcoin mining today is done with custom hardware based on ASICs designed for the Bitcoin algorithm. Have you checked the price of those? --47.157.122.192 (talk) 03:19, 8 December 2017 (UTC)

A sharp rise in the price of tin has resulted in the re-opening of a mine in Cornwall. However, one mine which isn't going to be opened is this one: [7]. 82.13.208.70 (talk) 11:38, 8 December 2017 (UTC)

Some of our regular readers know that I've spent a portion of my infamous career indulging in the art, science, and business of mining and prospecting, and oh boy, do miners have a reputation for selling lodes! If you've never been involved in a dig, it's time to familiarize yourself with the long con, also known as the "solid business plan," in all its glory and wonder. There is no better resource to refer you to than the nonfictional account of Roughing It, in which comic masterpiece we acquaint ourselves with the original goldminers: metal prospectors in the Sierra Nevada. "What could a man say who had an opportunity to simply stretch forth his hand and take possession of a fortune without risk of any kind and without wronging any one or attaching the least taint of dishonor to his name?" Or, translated into 21st-Century-ese: "I've got a guaranteed stash of bitcoin buried in a landfill, and I just need some investors to help me with the up-front capital costs to start digging. If possible, I'd like to expense my jet travel, and a mule, too."

The deliberate choice of "mining" terminology to describe the software processing of these digital currencies greatly contributes to the amusement-factor - at least, for anyone who understands the in-joke!

Nimur (talk) 05:36, 11 December 2017 (UTC)

I'm calling a major on your claim that the surge in bitcoin pricing is primarily because people are looking at a safe haven. Sure it may have originally been a factor, but while I guess there may be a few who agree with you, most experts on economics question whether it's a significant factor with the current insane rises, generally suggesting as Nimur said it shows all signs of a bubble. Let's not forget other safe havens have not experience anything even close.

At a basic level, it's not that hard to understand either. If something has risen 10x in a few months, it's easy to see the attraction. If you'd only put $10k in it not long ago, you'd now have $100k. You've already missed the part of the boat, you don't want to miss the whole thing. And even those predicting doom are generally reluctant to say the value isn't going to go 10x more before there's a correction.

While Nimur had some points about the state of the market, I haven't seen strong evidence that you really can't get in or out as long as things stay as they are and you're not talking about too much money. (E.g. I'm not saying the Winklevoss twins could really easily get their $1 billion or whatever it is now.) Sure you may lose a silly amount compared to what you feel you should get but if you spent $10000 and are now getting $80k that's still a great deal. You're fine as long as there's enough suckers, er other people, who want to get in. When the shit hits the fan is if the value does collapse. You could easily find your $10k (or whatever) nearly gone.

But even then, this doesn't even mean that it's always a bad idea. The evidence suggests plenty of people who do think it's a bubble and fairly experienced investors are getting in, treating it similar to other high risk high reward things like VC. If you have enough money you can afford to put some in with the hope you'll get lucky and get out before it collapses. You have several different investments of the sort and you just need one to pay off for it to have been worthwhile. The issue is those who don't really understand the risks and are putting money in they can't actually afford to lose (or would otherwise be better places to be choosing less risky investments).

Nil Einne (talk) 12:06, 9 December 2017 (UTC)

The answer to the original question: bitcoin "mining" is actually on an approximately predetermined schedule of one block every 10 minutes (average), and whoever figures the block first gets the mining reward (12.5 BTC). Throwing more computing at the problem mines faster very temporarily - because every 2016 blocks, the mining difficulty is adjusted to be approximately every 10 minutes again. So, mining is an evolutionary arms race.

But this won't supply more bitcoin to drive the price down, because it's a somewhat-regular release of bitcoins - David Gerard (talk) 20:44, 8 December 2017 (UTC)

So is it more likely to mine a bitcoin at certain times like northern hemisphere summer (since the cost of computer cooling's higher there) and the 4th of July (since that's when the most populous rich country has its only major summer holiday)? (are there any miners that *don't* run 24/7/365? since they're pro now and have big hardware investments to recoup) Sagittarian Milky Way (talk) 21:12, 8 December 2017 (UTC)

Cooling does not add much cost difference. For example common mining/3D graphics cards need 250 watts and only 5-10 watts are needed for the cooling fans. The main factor is the price of electrical power which is why the biggest farms work in cooperation with a local power plant and why they are only located in countries like china and island, where electrical power is very cheap. That is also why most amateurs dont stand a chance in the long run unless they produce their own power very cheap. --Kharon (talk) 05:47, 9 December 2017 (UTC)

While you're right that power usage of the components tends to be significantly higher than the cooling costs, it definitely can be a factor. That's why you get things like [8] [9] [10] [11] [12] (the last one deals in particular with bitcoin). Just because you can get away with just leaving your desktop computer (which is probably idle most the time anyway) in your house with just the HSF doesn't mean it works at a large scale. Of course if your house uses AC you probably are paying for it anyway albeit it's likely almost lost in the noise if you just have a one or two computers which are idle most of the time. Of course as others have noted no one uses GPUs for serious bitcoin mining now anyway. Nil Einne (talk) 11:47, 9 December 2017 (UTC)

The price of electrical power for heavy industries is below 20 USD/MWh in Iceland.[13] [14]. 75% renewable energy there - mostly geothermic. --Kharon (talk) 15:24, 9 December 2017 (UTC)

What does that have to do with anything? The only source I linked to which talks about Iceland mentions that the price of power, the use of renewable energy and the cool climate are all factors. All the rest do not deal specifically with Iceland, including the one which talks specifically about Bitcoin. It's true that a number of them are dual use cases i.e. by reusing the heat you could reduce the cost by getting someone to pay for the heating, although it's clearly not happening in some cases e.g. the Norwegian one it's mentioned there's no charge to the people with the devices other than an initial setup cost. Maybe a per use cost is/was the long term plan, but as some of the other sources demonstrate, the cost of cooling is often a factor in large data centres so simply getting rid of the heat without having to pay can be useful. You haven't shown any sources that the cost of cooling isn't a factor or that it only adds 5-10W for cooling fans in large scale use, as you implied. As I said, you seemed to make the mistaken assumption that what works with a computer at your home (which is probably idle most of the time anyway) works when you have a lot of devices in a small area with very high constant use but it doesn't. And as I also said even with the home case it's not necessarily true that the only cost comes from running the fans, it probably doesn't if you have an AC. Nil Einne (talk) 07:11, 10 December 2017 (UTC)