Bitcoin Forum

I had a random while reading about the development of ASICs and weighing the pros and cons of ASIC mining. It occurred to me that there are gigawatts of energy being consumed for heating hot tubs (of the outdoor style) and which, in most places, are heated year-round.

This last year I was mining.  Around the summer, I stopped mining because I believed it wasn't worth my time given the energy costs.  But about the same time, I decided to have my outdoor hot tub repaired (it came with the house when we moved in and I thought I'd finally give it a go).  I am guessing that the hot tub burns almost as much power keeping itself warm as my miners burned mining bitcoins because my power bill hasn't come down all that much.  (I had eight 5970's and four 5870's and calculated my usage to be 4kW).

I don't know how much money I spend on heating my hot tub, but I know it's a lot.

It occurred to me that if I could dump my mining heat into the hot tub, I would basically be getting free mining.

The only thing that stopped me was the hassle and opportunity cost of such a project.

That said, if there was such thing as a ready-made miner that I could strap on to my hot tub to dump the heat in there, it would be almost as good as free mining.

As I started this post, I mentioned ASIC mining.  ASIC mining is also relatively free.  But it is also, comparatively, "rocket science" compared to dumping heat into a hot tub.  It's expensive, and is vaporware at the moment.

If we are all clapping because "rocket scientists" are (hopefully) about to bring us free mining via ASICs, what about the thought of "joe the plumbers" bringing us free mining from Home Depot just by using some copper pipes to dump the mining heat somewhere where the heat is desired anyway?  If I felt that someone could deliver me and/or install a miner that heats my hot tub, and didn't look stupid-ass and wasn't loud, I would have no hesitation giving them a chunk of the proceeds of the power bill I was going to pay anyway.

And certainly I'm not the only person with a hot tub.

I'm not a miner but I've given this a lot of thought. (Even before bitcoin, actually. My refridgerator is right next to my stove, exhausting excess heat while I'm paying for more electricity to heat the elements just an arm's reach away.)

Sadly, I don't have a solution for you.

This could potentially work. I run a water loop for my 5970s and the water temperature is around 40-45 Celsius. A quick Google search puts hot tub temperatures in the 38-40 Celsius range. So, with a simple PID controller you could use GPUs for the bulk heating requirement for a hot tub and have a smaller heating element attached to the PID controller to maintain the water temp at the comfort range you want. The ideal solution would be a water/water heat exchanger so that your waterblocks aren't exposed to whatever chemicals/bacteria reside in your hot tub.

Watercooling is your only viable option.

You just need a water to water heat exchanger.  You will still need a water to air heat exchanger (radiator) "downstream" to remove the rest of the heat.  You need somewhere to dump "excess" BTUs.  Your hot tub can only take so many BTUs so the system can't rely on ONLY the hot tub as a heat sink.

One issue is that due to thermodynamics your hot tub will never be hotter than the coolant loop.  If you want your GPU running at 60C that means 130F to 140F hot tub.  If you are ok w/ 70C GPU you are looking at 140F-150F hot tub.

The issue is it isn't going to be (with current technology) cheap or maintenance free.  Water cooling will transfer essentially all the heat produced by the miners.  If you needed the heat anyways (to heat your hot tub, preheat your hot water heater, or warm your home) that energy cost is essentially free.  

That is the thought behind my watercooled server rack in the hardware section.  I don't have a hot tub but I intend to put a water-water heat exchanger on the cold water inlet of the hot water heater.  This should preheat the cold water entering the hot water heater from 60F to ~120F+ saving energy the hot water heater needs to expend.

I don't see why it wouldn't work, although it would almost certainly raise the temperature of your GPU water loop. Either that or you'd need a huge heat exchanger to move much heat with only a few degrees deltaT.

No you just need a water-air heat exchanger to provide supplemental cooling downstream of the hot tub heat exchanger.


GPUs ---> Pump ----> Hot tub heat exchanger ---> Water/Air heat exchanger ---> back to GPUs

It would be smarter and more practical to just plug your watercooling loop into the main hot water system of your house. Then you'd at least get use out of it at all times (rather than only when you use the hot tub) and your hot water is typically around 50C which is good for computer use. Hot tubs tend to be a bit hotter depending on personal preference.

I'm not sure it is worth the plumbing effort but it would be fun to play around with. I prefer to just keep it simple and mine for waste heat in the winter. In the summer, I will shut down all mining.

The typical household water pressure is 70-100psi, way to much for available gpu water coolers.
You may be able to come up with a custom solution, milling the water blocks yourself, but that's what I call risky.

One word.  HEAT EXCHANGER.  Ok I lied it is two words.

http://www.brazetek.com/files/imagecache/product/4x12_0.jpg

Exactly what I was thinking, though I'd have to think a while as to where exactly I would tie it in.

Would you though? If it takes x amount of watts for a hot tub to maintain temperature of y with it losing heat to evaporation/convection then so long as you input x or less watts then your water temperature won't exceed the original design. Issues would arise when you add more than x amount of watts to the water in the hot tub. Then the temperature would rise until the increased delta t between the water and ambient air caused an equilibrium to be reached through evaporation/convection. This isn't an issue for a heatsink but it is an issue for a heatsink that humans would like to enjoy sitting in. In that case then it would make sense to run a water/air HX before the hot tub so that you add less than x watts and then supplement to reach x watts.

You could get the GPU input water back down close to ambient using a water/air heat exchanger after the hot tub, but in order to dump heat into the hot tub you'd need the water out of the GPU to be hotter than the water in the hot tub loop. If you keep your hot tub at 40C and his existing water temperature is around 40-45C, he would need a pretty big heat exchanger to dump any meaningful amount of heat into the tub with only a couple degrees deltaT.

No that would be bad.  Heat exchangers work better with a higher Delta T.  Cooling the water before getting to the hot tub would make it useless for transferring heat into the hot tub.  If temp is less than hot tub heat would actually flow to the GPU loop.  If the temps are close you would get minimal heat transfer.  Also you have little control over how "cool" you get the water.  Lastly the temps leaving the heat exchanger will be close to the hot tub temp.  So your "cold" input temp for GPUs will be 60C-70C.  You are cooling the GPUs with water higher than the temps you want them operating at.

Pretty simple to solve the "humans don't like it that hot problem".  The first is that the hot tub will never get hotter than the GPUs.  If you use something like cgminer to throttle GPUs based on temps then water in loop will never be hotter than the temps humans like.  If you want max GPU temps then you need a bypass valve.  If hot tub is < desired temp then water flows through heat exchanger.  If it is >= desired temp valve closes and water flows around the heat exchanger.

GPU (60C) ----->  hot water heat exchanger (drops temps to ~ hot tub temp say 55C)  ---->  air heat exchanger (50C) ----> water returning to GPUs (50C)

GPUs then heat water back up to 60C and the cycle starts over.

Then run your GPUs a little hotter.  60C is nothing for a GPU.  60C vs 40C is more than enough Delta T.

Well yeah, but isn't that what I said initially? :)

Incidentally, I was looking at replacing my drain stack in my basement with a double output waste water heat exchanger to help recover some of the cost of heating water. I don't have a water setup, but it would be pretty trivial to implement a water loop into a system like that.

A heat exchanger indeed but not necessarily too big if you place the coil on the bottom of it link (http://www.youtube.com/watch?v=JRvxHA4AQbc), casascius is on to something here.

Maybe we could use existent solar heat exchangers (http://www.solartubs.com/solar-heat-exchanger.html) for hot tubs and and have it all installed in no time. The overall temperature would keep a constant level given the water surface with the ambient is far greater than a fan cooled radiator and you would have the water heated for free.

I suppose you already know that GPU water circuit can't be bothered with anything else than distilled water so a heat exchanger is the key part in doing all this.

Yeah maybe I misunderstood but you can't rely on JUST a hot tub was my point.  What happens when the hot tub is at 50C and the GPUs are at 50C?  With nowhere for the heat to go the temp in both loops will rise and rise faster in the GPU loop until you hit GPU killing temps.

Having a water-air radiator after the hot tub gives you control over the GPU temps outside of what the temp of the hot tub is.

This opens the door for other things heat related as well.  I use my miners for heaters in the house.  Have not turned the furnace on in months and struggle to keep the house under 72 lol

Yeah, I didn't mean that the hot tub would be only heat sink, just that the idea of heating the hot tub with excess GPU heat is sound. Even outside of other cooling for the GPUs, you'd probably need additional bypasses and controls to give you adjustment over the hot tub temperature.

Depending on where your GPUs are, you might even want to look into something like this.


GPUs --> Big Pump --> Hot tub heat exchanger --> Hot water tank cold input heat exchanger --> Cold water to shower heat exchanger --> Water/Air heat exchanger ---> back to GPUs

Run water/air exchanger fans on a temperature sensor so they turn off when not needed, and your GPU input temperature could very well be below ambient when you're taking a shower. ;)

The tub looses heat over it's surface and body if not insulated. There are formulas to calculate the temperatures the system will reach equilibrium with a given input energy (GPU temps) and natural dissipation surface of the tubs. Don't forget to take into consideration overall loses in the pipes and heat exchanger too.

The answer is NO, you won't be put in the situation of needing a water-aid radiator, imo

My thoughts exactly I think we were saying the same thing in different ways.

I may not go that elaborate for my setup but putting heat exchanger on cold input of hot water heater seems a no brainer.  That is the closest thing to "free energy" you can get.  

I suppose I should have used real temperature examples instead of abstract values. In my case I meant if the GPUs and hot tub started at 25C and you were adding more watts than the hot tub could dissipate that the water temperature would climb until an equillibrium was reached. So, if at some point your equilibrium is achieved at 87 Celsius, you could dump an unknown amount of watts before the water hits the hot tub so that your hot tub achieves a comfortable temperature. This way you don't throttle your farm(I mean we're here to make BTC, not heat a hot tub, right?) and you keep the humans happy. Hell, you could use a PID controller on the air/water HX to control the fans such that the outlet temperature of the water before it hits the hot tub is at the temperature you want the hot tub.

It would also appear that hot tubs run at around 60 Celsius. So, if your GPU inlet is 60 Celsius then you are looking at GPU temps around 75 Celsius which isn't half bad. You can't dramatically lower the inlet temperature unless you also dramatically lower the flow-rate as otherwise you won't have the time to heat up the water going through the blocks such that your outlet temperature is higher than your hot tub temperature.


Pretty interesting discussion. Would be fun to run a stirling engine from your waste heat to power the fans that provide cooling in a more traditional watercooled setup.

There's already someone here that heats his bathroom floors with his watercooled GPUs - sorry I've forgotten the nick. But there are pics on the forum somewhere.

EDIT: It's sturle, and here is the link: https://bitcointalk.org/index.php?topic=3707.msg53879#msg53879

a miner's wet dream... free energy or heating, whichever comes first

You could have a adjustable pump on the GPU side to increase or decrease the flow in case the cooling rate isn't sufficient, and make sure it's a big tub and have more than one heat exchanger, the rest is getting paid to enjoy such hot tub with family and friends :)


didn't saw that post rjk thanks, my fav line...


i should look into it more, yeah

Or radiant floor heat!

A 60C hot tub? Remind me never to visit your house. That's the temperature you use for Sous-vide. :)

Does anyone have any actual numbers on heat loss of such a tub? How much can I expect one to dissipate?

How about using a heat pump between the GPU cooler loop and the pool heating loop? It will add to the power draw, but the GPU:s will get cooler and the pool a lot warmer.

An outdoor hot tub can dissipate very large amounts of heat if it's not covered. I'm pretty sure it's more than 1kW. Most of the dissipation is done by the water when it vaporizes.

Really though, if your aim is to save money your first order of business should be to insulate your hot tub and buy a proper spa cover. I don't have exact figures, but a good covered spa doesn't take much power to keep warm as long as outdoor temperatures are reasonable. If you like to run your hot tub in the winter it's a different story of course, everything looses heat at -40.

You have too much unknown variables to be able successfully calculate all that, like primary and secondary circuit water volume and ambient exposure (insulation missing), inefficient heat exchangers and all that. Building it with a decent size tub, exchangers and GPU's, with the ability to add or remove some of them from the circuit easily, should give better results to reach some conclusions.

A rough estimate is that you would lose more than half of the energy you put in, so with 1kw rigs should give aprox 400-500w heat, and this value goes down with more energy we put in, better for the miner in this case.

hot tubs are around 40c (104 F). at least mine is.

as for heating your hot water with mining energy, a (very expensive, heh) roundabout way is have a geothermal HVAC system installed. in the summer when its in AC mode, excess indoor heat is used to heat your hot water tanks via its heat pump (so it can actually get it to 160 F if there enough warm air in the house). excess heat after the hot water tanks reach their temp is dumped into the ground. so, more or less free hot water in the summer :)

only practical for small scale mining setups (free gigs maybe?) but if youre gonna go geothermal (we are) its a nice bonus.

I know, not exactly what you guys are talking about... but..  figured Id toss it out. maybe someone here can devise a dedicated small scale home brew version. a heat pump seems to be one good way to regulating temps and moving the heat to where you want it. lotta sharp minds here, some one may think of something.

So I drew this up for fun. Uses a 4-way ball valve, which you might be able to get an electrically controlled version of and make the heater automated. This drawing shows the use of an external radiator in case the tub is too hot.

https://i.imgur.com/yoxCN.png

Maybe someone knows of some way to make this same idea, but with some kind of continuously variable valve. That would be the best thing to do.

couldnt you just run the water to air and water to spa loops in parallel (ie 2 loops). use stepper motors connected to valves to adjust flow to each, and some logic chips (with temp sensors in the spa) controlled via rs 232 port to control the steppers. computer program monitors the temps. water flow can be through one, the other or both loops, depending.

For me, the answer to deal with the problem of "what do I do when the hot tub is already hot enough?" needs no valves: it is a software solution to simply lower the hash rate when the extra heat is not needed.  Given that the original purpose is to mine with free power, there is no point in using more power than the hot tub needs for heat because that power is no longer free.

That makes it real simple then, just a couple pumps, a heat exchanger and a temp sensor with communications to your computer to control it.

what if you need to reduce the hash rate near zero? no fun it that.

outdoor temp in the 90's, with the sun beating down on it, spas can easily get near their 104 optimum temp with no extra heat input (dont forget using it increases its temp too as the spa's pump also heats it [friction] while its running, for scheduled filtration periods and in actual use). spas have an overheat cutout because of this.

If you want to be cheap no need for temp sensor.  Just use temp sensor on the GPU and cgminer.  With sufficient flow on both sides of the heat exchanger the two loops will reach equilibrium within a few degrees of each other.  So if GPU are 50C then the hot tub may not be exactly 50C it might be 45C or 42C but it isn't going to be 80C or 20C.

how long do you think you can dump a few kilowatts of heat into a 500 gallon insulated hottub (that already generates heat during its filter cycles even when its not in actual use with people)?

youre gonna throttle your farm down when the tubs at temp?

BTW hottubs are 104 F (40 C).

If I planned my mining to roughly consume the same amount of power as the hot tub uses now, I should be able to keep that hardware busy. Even if it was only two 5970s, it is FREE power.  Of course that presumes I'm interested - I recently sold my mining gear.

[lower hash rate to maintain temp]

yup, read that before. and I repeat:

youre gonna lower your hash rate to near zero?

OK, that will work, Ill admit.

BUT, arent we here to make some money too? not just save some by using waste heat.

Aren't these two equivalent?

If "you" provide a waterproof rig I can leave at the bottom of the pool and you "make" money and split it with me, that was worthwhile especially if it was just a couple gpu's. Do this with lots of people and you have money for little risk.

Sounds like a shockingly good time.

a couple cards is slightly more feasible. but even then 500 watts continuous is more than any decent hottub needs, unless you lave the cover off or something.

and I was under the impression some people in this thread had cooling kilowatt+ farms in mind.

this is a hottub we are talking about. we put people in them, not lobsters.

No.

Putting rigs underwater is unsafe but lets ignore that for a second.  Computing hardware is expensive.  A 1 KW pool or hot tub heater is pretty cheap.  Maybe $300.  A rig which uses 1 KW or power, has water blocks, circulation pumps,  in a nice sealed box with heat exchanger and connections to hot tub is more like $2000. 

So using it part time isn't really viable for anyone.  Rig should run 24/7/365 and provide unlimited free heat to the hot tub but also 24/7/365 hashing operation.  Otherwise given the limited power needs of a hot tub you are looking at a payback period in decades.

This isn't to say something couldn't be done but
a) underwater rigs high voltage rigs
b) "part time" rigs

simply don't make any economical sense.

exactly. to me, throttling to maintain the hottub at 40C doesnt make sense.

hence, a pump, two loops (one to hottub, other to waste heat radiator), steppers controlling valves to the loops. a few ICs and other stuff (just need a comparator circuit for the temps, really).

on demand heat. max cooling (and hash rate) at all times. NOW its making sense.

It doesn't need to be anything fancy with connections. Just a box you stick underwater when you aren't using the hot tub.

As for "shockingly good time" just unplug the box and take it out when using the tub silly. And use a gfi outlet so the power is cut if any leaves the box into the water as a secondary measure (gfi is already installed on most outdoor outlets anyway on modern construction).

I'm starting to see how you are thinking now. Most hot tub owners that I know of keep the thing on 24/7 so it is ready to go, but it sounds like you wish to only turn it on when you want to use it. In that case, you will need a lot of wattage, otherwise it will take forever to heat up.

Why not run a copper coil for the hot water instead of putting your electronics underwater?

I mean something like the opposite: I have to keep it warm 24/7 (or else it would freeze over), but only use it for 15 minutes here, an hour there.  The miner would keep it warm the rest of the time while it wasn't being used, and having it underwater would eliminate the noise problem.

The very fact that I HAVE to keep it warm in the winter (the only other choice is empty and unusable) means the power cost has already been justified the moment I decided to keep it full.

OK I was confused by your statement of unplugging the box and taking it out, sorry. Still the extra cost would probably not be worth it.

Let's start with this: how many watts is the current heater, and what is the approximate duty cycle that it is running at? I am assuming it is not running full on all the time, but is rather switching in and out to keep the temp steady.

Sure, the pool's heater will be switching on and off, but for all intents and purposes, let's say I can readily conclude that I can keep 500 watts of mining running full time, just by averaging its consumption for the month.

I don't know to be exact (I have never measured it) but let's assume that in the summer, it costs $50 a month in power, and in the winter, it costs $200.  That's 7 cents an hour in the summer, and 28 cents an hour in the winter.  7 cents an hour maybe buys me 500 watts.  In the winter, that buys me maybe 2000 watts (possibly more, because $/kWh is higher in the summer).

500 watts is what, a 5970+motherboard?  5970+5870+motherboard?  That's easily a gigahash.  Even with a 500-watt miner (relying on the hot tub's heater to provide the remainder of the heat for the winter), that equipment could be kept near 100% utilization on free power.  When power is free, equipment pays for itself much quicker.

Well unless you add a block to the CPU, don't count it in (same goes with PSU inefficiencies), since all the heat will only be coming from what you have blocks on. You could do 2 or 3 5970s, and even downclock them for a better mhash/watt while still getting enough to keep the pool warm.

crazy idea: Why not use a heat pump?

GPU-> Water to Water Heat exchanger -> pump -> GPUs

heat pump "cold side" would be warmed with the GPU temperature, giving you a better efficiency for heating the hot tub, while possibly making the GPUs inlet water temp below ambient, atleast at times.
Add water to air heat exchangers or other components as needed, just a basic crazy idea :)

Heat pumps can achieve close to even 1:6 ratios of efficiency, and i'm not kidding, there are a lot with those over 1:5 ratio for sale.

But yeah, your power consumption would still increase by at least 800W :( Plus the cost of making heat pump to heat the hot tub, no idea how much that will be.

If it's underwater, if the heat does not go into the water, where else will it go? =)

If heat is the desired product, there are zero inefficiencies.  The only thing there needs to be is something to make sure the CPU and PSU get their heat carried away so they don't burn up.  There will not be any net air flow - the air pocket will get hot in proportion to how well the heat can transfer out of the air and into the water.  A CPU block and a water-cooled power supply would probably be necessary.

So you you mean a completely sealed capsule/module that contains a miner that produces heat? And then you dunk it in?

Air is a very good insulator. 

Yes - drop it to the bottom, plug in power and ethernet, heat the hot tub and mine coins!

It would be nice if it could be WiFi too, but WiFi won't penetrate water worth a damn.  So, Ethernet is the only choice.

Get a big aluminum casing, IP68 and stick several FPGAs inside of it.

But you pretty much would have to develop your own board(s) (2 could fit in it on both sides)
And depending on the size of the hottub that's alot of fpgas, which would cost you.

The air in the case will not effectively conduct the heat from the heatsinks to the outside of the case, because as D&T noted it has properties that cause insulation and not conduction. It would simply overheat and shut down.

Hmm. A miner that can give me hot water (for the home as well as an outside pool) as well as Bitcoins (money)? That would be a very good solution. If only the instructions were like really easy for a guy like me who doesn't know much about computers ( I can setup GUIMiner but that's about it plus I still need to build an actual mining rig with those high performing cards HD7970s which would then be fitted with something that can pump water straight into the water heater which then goes to an outside pool.)