Submersing a rig

[P4man]

2000w will more than likely impact 10,000 gallons, though it wholly depends a lot on surface area.

I dont believe so. the pond is not that deep. Its bowl shaped, no deeper than 1m60. Half the pond (by surface) is less than waist deep.   Surface area is 60m2. Then there is the fact I have a DIY "trickle tower"/ proteine skimmer thats designed to  expose the water to air and *lots* of aeration in general (heavy duty air pump in the filters and in the pond).  When I turn on all my pumps on a cold autumn day, I can easily decrease the water temperature by several degrees in a single day, 2000W isnt going to offset that.  Im not sure 20Kw would.

Anyway, heating the pond is not the objective. Heating the water in the filters would be quite a useful by-product, particularly in spring, although Im not expecting much difference even there. Too much flow. One can do the math, the filter barrels excluding filter material are probably around 200L and I believe are turned around every 4 minutes. At 2000W that is  ~ +0.5C. Better than nothing, but not impressive.

[DeathAndTaxes]

Insulator was the wrong word, but now I need you to explain something. Would the layer of water trapped by a wetsuit be considered an insulator, or is the insulating going on in the neoprene? Wetsuit without that trapped layer doesn't do anything.

The water in wetsuit isn't used for insulation.  A dry suit (not a dry wetsuit but an actual drysuit) will always be superior but dry suits are expensive, complicated (the air compresses at depth), and require a lot of maintenance.  It is the neoprene which provides insulating properties.  Technically it is the air bubbles in the neoprene which provide the insulation (air is on the best low cost insulators, and water is one of the best low cost conductors).  This is why as you dive deeper a wet suits becomes less effective.  The water pressure compresses the bubbles in the neoprene reducing the insulating value.  You can partially compensate with thicker neoprene or use a dry suit.

The "wet" in a wet suit keeps it simple.  The suit doesn't need to be "perfect fit" just snug enough to minimize water exchange.  The water layer aids in comfort and since water is a conductor is spreads around any heat loss.  For most diving it is simply "good enough", however if you have ever had a wet suit which is torn you will quickly find out how bad of an insulator water is.

My point was that dumping a constant 2000w into a body of water will heat it, and I am not sure ambient temperatures and airflow will be enough to 'exhaust' it, so to speak. Throw the sun into the mix and you are going to have some wild daily temperature swings to the detriment of the fishes health.

Yeah not sure how big the body of water would need to cushion the temp swings.   Since water is such a good conductor of heat what matters is how well the pond can "exhaust" the heat into the atmospshere.   I also have no idea how much of a temp swing the fish can tolerate.  

I guess one could get a ballpark figure by measuring the pond temp in summer at dawn and noon.  The rise in temp ~= thermal energy from sun.  Figure out solar insolation, and number of hours and you could turn that into wattage.  i.e. pond got x watts from the sun and the pond temp rose y.  That might give p4man at least the magnitude we are looking out.    With only a 5C rise in temp can the pond dissipate 100W, 1000W, 10,000W, etc?

[P4man]

I guess one could get a ballpark figure by measuring the pond temp in summer at dawn and noon.  The rise in temp ~= thermal energy from sun.  Figure out solar insolation, and number of hours and you could turn that into wattage.  i.e. pond got x watts from the sun and the pond temp rose y.  That might give p4man at least the magnitude we are looking out.    With only a 5C rise in temp can the pond dissipate 100W, 1000W, 10,000W, etc?

Im not worried about that in the least. Temperature difference between bottom (where I pump most of the water) and top of the  pond can be 5-10C difference just from the sun. Between morning and afternoon you can add another 5+C to that when my filters are running. Fish dont mind. They will just swim where they feel most confortable, and if anything, Koi like warmer water than what I can offer them. They thrive when the water is ~30-35C, Ill never get there, except for some weeks per year at the top of the pond...

Once best avoids sudden temperature shocks, like from a hot bassin to a cool pond without some acclimatisation time, but intra day variance is not an issue. They are river carp, its not like you will not find huge temperature differences in a fast flowing shallow river.

[DeathAndTaxes]

Im not worried about that in the least. Temperature difference between bottom (where I pump most of the water) and top of the  pond can be 5-10C difference just from the sun. Between morning and afternoon you can add another 5+C to that when my filters are running. Fish dont mind. They will just swim where they feel most conformable, and if anything, Koi like warmer water than what I can offer them. They thrive when the water is ~35C, Ill never get there.

Sounds good.  Also if you wanted an added layer of security you could rig a temperature switch to the power for the rigs.  If temp in pond got above say 40C it would kill power to rigs.

[PrintCoins]

Wouldn't an fish tank filter :



do just as well if not better than a radiator at cooling down the oil. You would of course take out the actual filters, and just have it as a way to pull oil out of the tank and pour it through the air back into the tank.

This puppy is on sale for $29 bucks and can turn over 350 GPH of water (probably less for oil)
http://www.amazon.com/Marineland-Penguin-Power-Filter-70-Gallon/dp/B0009IMDQM/ref=pd_bxgy_petsupplies_img_b

Don't forget to decorate your tank:

[RandyFolds]

Insulator was the wrong word, but now I need you to explain something. Would the layer of water trapped by a wetsuit be considered an insulator, or is the insulating going on in the neoprene? Wetsuit without that trapped layer doesn't do anything.

The water in wetsuit isn't used for insulation.  A dry suit (not a dry wetsuit but an actual drysuit) will always be superior but dry suits are expensive, complicated (the air compresses at depth), and require a lot of maintenance.  It is the neoprene which provides insulating properties.  Technically it is the air bubbles in the neoprene which provide the insulation (air is on the best low cost insulators, and water is one of the best low cost conductors).  This is why as you dive deeper a wet suits becomes less effective.  The water pressure compresses the bubbles in the neoprene reducing the insulating value.  You can partially compensate with thicker neoprene or use a dry suit.

The "wet" in a wet suit keeps it simple.  The suit doesn't need to be "perfect fit" just snug enough to minimize water exchange.  The water layer aids in comfort and since water is a conductor is spreads around any heat loss.  For most diving it is simply "good enough", however if you have ever had a wet suit which is torn you will quickly find out how bad of an insulator water is.

I've done a lot of diving with both wet and dry suits ('fuck Lake Superior' is all I can really say about that), I just wasn't hip to how the liquid fit into the insulating picture. I suppose I should have pieced it together, what with needing a new wetsuit every six months due to compression. Southern California's got some chilly water, but dry suits absolutely suck...turn your head too far and you're wearing wool soaked with icewater.

I was an AAUS diver for a bunch of years and people are always disgusted by the concept of pissing in your wetsuit...all I can say is when you have to dive at five in the morning in 10c water, it is the only relief you are going to get...even a 7mil with an insotherm and a hooded vest won't keep you warm in that shit.

[P4man]

Sounds good.  Also if you wanted an added layer of security you could rig a temperature switch to the power for the rigs.  If temp in pond got above say 40C it would kill power to rigs.

Above 40C? Heh. Maybe I need a picture here to help you visualize 50K liter. Here is half the pond half drained for maintenance this spring:






A few gpus arent going to cause that to overheat .

Im much more worried about leaking. Ill probably add a floater switch that shuts everything down if the oil level drops. And probably an electric valve as well.

[RandyFolds]

That is a baller pond. Start the aquaculture op!

[P4man]

Wouldn't an fish tank filter :
do just as well if not better than a radiator at cooling down the oil.

Definately not.  Its just a waterfall, all that matter is the surface area you produce, and that isnt much. Moreover, insummer inside the shed it already gets to 35+C without help from mining rigs, and I were to dump all the heat in there, it would get a lot hotter. And then you only have (hot) air to cool the oil, thats not nearly as efficient as cool water.

[P4man]

That is a baller pond. Start the aquaculture op!

Not sure what you mean. If you think there is not enough plants in the water, keep in mind thats early spring before most of the plants and lillies surface, and right after a thorough cleanup. in the summer you can almost walk across the pond without getting your feet wet .

[RandyFolds]

That is a baller pond. Start the aquaculture op!

Not sure what you mean. If you think there is not enough plants in the water, keep in mind thats early spring before most of the plants and lillies surface, and right after a thorough cleanup. in the summer you can almost walk across the pond without getting your feet wet .

No, I meant you should be farming food fish and vegetables in there. 10,000g with some upgrades to the biofilters should be able to handle a good five or six thousand pounds of tilapia, easy...just cut out some raceways and start growing greens and herbs for your local restaurants.

[P4man]

So you are suggesting I replace my precious Koi with these hellish creatures:



No thanks. Its so ugly I wouldnt even want to eat it .

If I were in it for the money, I even think growing Koi is the better investment. Though I would need warmer water... (or indoor winter tanks).

[RandyFolds]

So you are suggesting I replace my precious Koi with these hellish creatures:

No thanks. Its so ugly I wouldnt even want to eat it .

If I were in it for the money, I even think growing Koi is the better investment. Though I would need warmer water... (or indoor winter tanks).

Oh, there are some prettier species than that but they are all pretty fugly in the grand scheme of things.

I only suggested tilapia cause it can tolerate the highest density. You could grow out some walleye pike or something awesome like that, but like all the other delicious apex predator fish, they are super cannibalistic and have like, 90% mortality on an ok run.

I know a couple people who culture Koi in California and make a good amount of money on it, but for every nice koi, there are a thousand flushers.

[P4man]

Ive been reading and searching further on this subject to decide on the oil. I can not find a cheap local source of mineral oil, so Im gonna go for vegetable oil. Im aware this is less stable and can oxidize (go rancid). It seems the biggest factors contributing to oxidation are sunlight and humidity. The first is easy enough to tackle, I have no need for a transparent aquarium, by setup is going to be butt ugly anyway.

To counter humidity, Im going to try to add a layer of mineral oil on top of the vegetable oil as sealant. Mineral oil is lighter and should therefore float on top, but Ill have to see if it mixes or not.

If anyone has other suggestions, Im all ears. Is there something lighter than mineral oil that is stable and not conductive to use as sealant?

[SgtSpike]

Ive been reading and searching further on this subject to decide on the oil. I can not find a cheap local source of mineral oil, so Im gonna go for vegetable oil. Im aware this is less stable and can oxidize (go rancid). It seems the biggest factors contributing to oxidation are sunlight and humidity. The first is easy enough to tackle, I have no need for a transparent aquarium, by setup is going to be butt ugly anyway.

To counter humidity, Im going to try to add a layer of mineral oil on top of the vegetable oil as sealant. Mineral oil is lighter and should therefore float on top, but Ill have to see if it mixes or not.

If anyone has other suggestions, Im all ears. Is there something lighter than mineral oil that is stable and not conductive to use as sealant?

I like this idea.  Way to think outside the box!

[P4man]

Its not so far out of the box, since I got the idea while reading about mineral oil often being used to seal other liquids. So why not vegetable oil
Anyway, Ill go shopping this weekend and Im buying some extra parts to build a rig around an old 8800GT to experiment with.

Some other points I came across:
- thermal grease will dissolve in oil. Before plunging in the cards, youlll have to remove all thermal grease. I hope the vrm pads will manage.
- I came across one report of overheating with a S775 CPUs, despite the oil being cool. apparently caused by air trapped in the socket; though Im not quite sure I understand how that would work, Im assuming this is not a problem with GPUs as there are no sockets.

[Defkin]

What about if you used a metal box like an ammo box that you can just fit the cards into. Leave the fans on the cards to circulate the oil. Because the cards fit snugly into the box not as much oil is needed to submerge them, this way you can use the mineral oil rather than run the risk of vegetable oil.

Place the metal box into a larger plastic container of water. Place it the top of the metal box is sitting above the top of the plastic container so any accidental water overflow goes onto the floor not into the box. Pump water into the plastic container and use gravity to take it back to the pond via an overflow. The oil/metal/water interface should be a good heat conductor.

Make sure you put a zinc in the water to stop galvanic series eating the box away.

I cannot see the issue about thermal grease being eaten away, if it was wouldn't just be replaced by the oil that ate it??

[P4man]

What about if you used a metal box like an ammo box that you can just fit the cards into. Leave the fans on the cards to circulate the oil. Because the cards fit snugly into the box not as much oil is needed to submerge them, this way you can use the mineral oil rather than run the risk of vegetable oil.

Place the metal box into a larger plastic container of water. Place it the top of the metal box is sitting above the top of the plastic container so any accidental water overflow goes onto the floor not into the box. Pump water into the plastic container and use gravity to take it back to the pond via an overflow. The oil/metal/water interface should be a good heat conductor.

Make sure you put a zinc in the water to stop galvanic series eating the box away.

While this may work, you effectively reduce the surface area for the thermal transfer to the size of the box. I suspect an oil cooler would have a larger surface area, and for my setup at least, its much easier to drop a cooler in my existent filters where I already have waterflow, than create a box and somehow circulate large amounts of water through there without overflowing it due to pump head. With gravity return that means HUGE pipes or low flow.  You would also risk a fair amount of water evaporating in a place where I dont want humidity, ie, in the shed where my rigs are.

The idea of an ammo box is excellent though.


Oh, and no zinc in the water. My Koi dont like it

I cannot see the issue about thermal grease being eaten away, if it was wouldn't just be replaced by the oil that ate it??

Yes, the problem is thermal grease can be conductive. You dont want it dissolved in the oil. But indeed, for temperatures it should not be a big concern as it would be replaced by oil.  To make sure no air gets trapped, you will want to smear some oil on the die before re-mounting the heatsink.

[Defkin]

I think you might be surprised how much heat a metal surface area can transfer with liquids on both sides. It would be easy enough to test if you have a submersible heater or jug element.

I am trying to picture what you are going to do. Are you only putting the cards into the pond if so how will you protect the MB and PSU etc?

[P4man]

I think you might be surprised how much heat a metal surface area can transfer with liquids on both sides.

Its no different than with a submerged radiator. Assuming your ammo box has equal thermal transfer characteristics as an oil cooler, which I doubt considering the thickness, coating etc, all that matter is surface area and temperature difference between the liquids.  A radiator has a huge surface area for its volume, while with your approach, surface area is directly linked to the amount of oil and scales rather poorly with it (cubic root) .

Mind you, Im not saying it wouldnt be enough but I fail to see the advantage, particularly since I wouldnt be able to flow nearly as much water through the "containing container", so Id have higher water temps (and evaporation).  The rigs will be above pond water level, so I cant use gravity feed, I need pump fed. Since you cant pressurize the container, you need gravity return. Unless I plumb at the very least 50mm water return pipes to the shed, thats going to be a serious limitation on flow rate or cause headaches with the water level. To give you an idea, I use 130mm return pipes in my filter, and with my pump capacity even that creates a 10+cm increase in water levels in the filters. Now I dont think youd need anything like the kind of waterflow I have through the filters, but it sure doesnt hurt, and I already have it, so why not use it? The higher the flow, the lower the water temps.

I am trying to picture what you are going to do. Are you only putting the cards into the pond if so how will you protect the MB and PSU etc?

Not sure what you mean by protect. Protect from what? Take a motherboard with a few gpu's (probably using extender cables), turn it upside down. Drop the cards in an oil bath. Even if oil would splash on the MB or PSUs or, or the oil creeps up, it doesnt matter. Ill probably prevent this by creating a lid over the oil box though, mostly to prevent dirt from getting in.  I will wire the cables through the lid and seal it (and may use PCIe power extender cables that are fixed in the lid, so the psu cables plug in to the lid), but thats not even necessary.


Here is sketch made by the 5 year old in me to explain :