Chiller idea... I might be insane

[brokentusk]

Quick and too the point.

Why can't I just make my sump out of a 40 quart (20 gallon) thermoelectric cooler?

I need to check ours to see if there are any metal parts coming in contact with the liner, but I can't think of why not. The liner is food grade I believe. I have to check. (easy enough for me, I work at Coleman)

But other than needing a temperature controller that won't freeze out my fish, I can't think of why not. I might even be able to mess with the controller that is already there.

[manderx]

do a search on iceprobe. it's a TEC designed for cooling liquids. i doubt the one built into a cooler has enough umph to do anything significant on anything but the smallest tank. remember that while you're trying to cool the tank it's still making more and more heat. refrigerators and the like are designed to remove a finite amount of heat from the food you put in them.

[brokentusk]

I've seen the slug size on a chiller at the LFS. Looks about 1/2 the size of the TE one.

This isn't a fridge. It's designed for someone to fill it with a warm object.

The amount of heat that needs to be moved by the chiller is small. On my 90 gallon it's like 4°.

I acutally think the 16 quart version would work better. The chiller is in the lid and constantly would blow cool air across the surface of the water. I think I could use it as it sits with a rio pumping back into the sump. Or I could plumb it (gee, wonder where I could get a hole drill in it) and leave the top sealed.

Not too wild on that idea. However you would be able to create an area of cool air. 50° off ambient is what they can do.

Low voltage too. Runs off 12 volt. There is an AC adapter for them.

http://www.coleman.com/coleman/colemancom/detail.asp?product_id=5615A734

Is the one I am thinking of now... Cooling unit is in the split lid. With a simple overflow added to the unit you wouldn't have to worry about the fan getting drowned. As for fan life, it is (was last I knew) a standard PC case fan. Actually, they were a thermo electric fan before they were a case fan.

Basically you aren't trying to freeze the water or keep it at some 40° level. You are just trying to remove slightly more heat than is being produced...

Then again when I get done with this weekends experiment, I may not need it. Everything that has an external pump or motor is getting heat tape and copper or aluminum heat sinks attached.

Don't worry, the copper is not going to be near the water. And they will be attached with thermal tap or thermal grease and wire ties.

[blgreef]

you may find that the cooler will not have enough capacity to cool the tank.

Thermoelectric cooler use peltiers and these are power hungry devices with more current required for greater cooling.


Since these unit would be designed to cool a specific capacity you may find that the peltier and the power it can consume would not be large enough to cool the tank.

but worth a try

[manderx]

a 4 degree change in 90 gallons is a lot of heat.

i know this isn't exactly thermodynamically precise, but it's close enough: imagine how much ice it would take to keep that cooler 50 degrees below ambiant for a day (as the specs say it can with it's built in TEC) . now put that same amount of ice in your tank and see exactly what it does to the temps. i'll bet it doesn't even register.

[brokentusk]

That's not quite the same thing. The ice will equalize to the water temp. It isn't regulated. It starts at 32° and climbs. This is constantly applied cold air.

If the ice stayed at it's constant temperature with the tank making it's constant heat, the water will change one way or the other.

The water will change temperature with this method. How much is the only question. A 10th of a degree or 10 degrees.

You could submerge the return pump in the cooler with freshwater in it. And have the saltwater passing through the pump only with no exchange with the was in the cooler. Basically a water cooled pump. That could make a big difference. If you had the pump at 100° and that was adding say 5° to the total water temp. I think you could knock it down to just one or two degrees added.

Total heat added is the tank water temp minus the average room temperature. Here that is about 5 degrees.

Worst thing that happens is I end up with an extra cooler.

[Vincerama2]

I thought about using one of those coolers too. I think the conclusion I came up with was that I'd just run a big fan across the top when it was warm.

My tank was at 86 degrees when I came home today (in San Francisco!! What's up with the heat?) But it was also hot yesterday, so I had frozen some salt water in my freezer overnight and when I got home, I put it in my HOB filter. It came back to 84 or so (ambient temp) where it is staying. My apartment has no AC, so it's brutal for the tank (and me)!

I am going to go to modify my tank hood (wooden DIY job) to take some fans, it's too traumatic watching your $100++ tank inhabitants (hey, even just the rock!) cook!

Vince

[Pyrojon]

Cut back on the next $100 tank inhabitant and get yourself an in window AC unit at WalMart. :)

[manderx]

ok, i'll fill in the blanks in my example since everyone hasn't had high school physics.

you have a cooler. it's 100 degrees outside in the shade. you want to keep the air inside of the cooler at 50 degrees. (to match the published and probably exaggerated specs of how much heat this TEC can move).

i would conservatively guess it would take a cup or two of ice every few hours to maintain this temperature gradient inside an insulated cooler, since the only heat coming back into the system has to come through the insulation. this means that the cooling power of the TEC is only equivalent to a few cups of ice, and there's no difference between cooling the tank with the TEC or dumping a tray of ice in the tank every few hours. go ahead and see how much ice you have to dump in to get a momentary .5 degree change, let alone maintain it.

even if my numbers are an order of magnitude off, there's no way it could move enough heat to do anything on a 90 gallon tank. maybe a 10 gallon, but not a 90. you would need at least a 500 watt TEC to do 1 degree to a 90 gallon tank. and by then you'd have a serious problem of getting rid of the heat on the hot side of the TEC. it just moves heat from one side to the other plus the heat generated by the energy wasted in the process.

[Vincerama2]

My 10 gallon tank was again at 86 degrees last night. In desperation, I found two 3 inch computer case fans and luckily the ac adapter for my crappy scanner happened to be 12 VDC output, so I rigged an electical connection (I taped the fan leads to the adapter plug) then I taped the two fans to a piece of wood. I took the hood off the tank and put the "fan board" on the corner of the tank, taped it to the outside so it wouldn't fall in, then ran them all night. This morning, the ambient room temp was about 79 and the tank was 74!! Oops! Too cold now!

Well, it seems to me that the power usage of a TEC is probably much more than a set of high performance (these Sunon case fans move ALOT of air, heck they are made for cooling computer cases!) fans so even for a 90, a row of these fans (they can be pricy from 11-20 dollars a piece!) OR a big desk fan, might be all you really need.

Evaporative cooling seems to work very well (ie; swamp cooler, mammal sweat glands, etc).

Anyway, it seems (for me and my little 10g) that the simpler solutions are the best!

Vince

[Entropy]

I think several people in this thread are going about this the wrong way. brokentusk works at coleman (as stated) which means he can more than likely get this stuff cheap.

There is one positive undisputable way to figure this issue out. Hook the sucker up and see what it does :D

[brokentusk]

First plan is to throw heat sinks on the return pump. And remove a couple other items that seem to be producing too much heat. IE the ballast is mounted under the tank. Moving it where the heat isn't rising right up onto the tank will help.

Then see where it is.

I actually have found some higher end TEC units for about the same price as the cooler. So I might just make a unit to blow cold air across my sump.

Oh, and yes I of course get a good deal. But we actually have it set up so we can share that with 'family members'. Just go to the www.coleman.com site and select the 'family store' option. The password is 'family'.

[Mutagen]

Brokentusk,

I think you will not be happy with the performance of the cooler you propose.

Thermoelectric chillers typically have a COP of around .4 to .8. A good refrigerator can easily have a COP of 4 or more. In other words, a good chiller might well be 5 to10 times more efficient that the peltier cooler. (COP is "coefficient of performance" and is the ratio of the energy removed divided by the energy consumed.)

If you wish to pursue the idea, I would think you could find out the power input to the cooler and its COP from Coleman and then you would know for sure.

Here is an example:

Say the Coleman unit has a max power draw of 150 watts and an estimated COP of .6 at the temperature differential required to chill your tank. Then the chiller will remove 90 watts of energy from the tank on a continuous basis, (while consuming 150 watts).

As a basis for comparison it takes about 300 Btu to melt a pound of ice and one pound of ice is just slightly more than a pint of water. Also, 300 Btu is about equivalent to 90 watt-hours.

The question then is how much heat is added to your tank and how much of that do you need to remove with the chiller. This question is much harder to assess since heat can leave your tank system via raditaion into the room (VERYminimal), conduction through the stand (VERY minimal), evaporation (which can be the majority of heat loss), or conduction through the glass into the room. But you can at least compare the hypothetical cooling rate to a worst case heat input. For example, you may have two 250 watt MH that produce at least 350 watts of heat. Since MH lights are high temperature sources they are very good at radiating heat into the water. In other words, the majority of 350 watts would be added to the water. Now add in the fluid energy from you pumps, plus the waste heat from motors if you have any submerged pumps and the 90 watt chiller could be easily overwhelmed. The chiller would be of use only as a supplement to the other natural means by which heat is removed.

The other thing that sucks with any chilling device inside the house is that the energy you removed from the tank must also be removed from the room. So inefficient chillers like Peltier devices can actually cost you quite a bit more than you might think.

Hope this helps with your decision, Mutagen