Anyone know how thermoelectric coolers work?

[SaltyNuts]

Peltiers are inefficient in the sense that you will consume way more power per BTU than say, an air conditioner. That said, they do have their place, as they are solid state and probably very reliable compared to a standard refrigeration loop.

So depending how cold you need to go, it might make sense to get a small fridge or air conditioner, and circulate the water through it. You could make a heat exchanger. Depending how smart you want the system to be you could have various temperature sensors that control the pump and the chiller (whether it's a fridge, freezer or maybe even AC unit).

I'm guessing it does not need to be super cold, so I'm thinking a bar fridge would work. Setup a water tank inside the fridge, then coil up some copper tubing inside the tank to act as heat exchanger and have the fish tank water circulate through it. The idea is the tank inside the fridge acts as thermal storage. Then have a temp sensor in the fish tank that activates the pump when the water gets too warm. I imagine not shocking the fish is also important, so by using this indirect cooling method it should work decently while having good capacity. Suppose you could have the fish tank water go directly inside and out of the tank that is inside the fridge, but I feel it's probably best to keep it separate.

Disclaimer: I never did this before, so I can't guarantee it will work, but I'm thinking it probably would, worth a shot unless someone can confirm it won't work.

You certainly nailed one very valid idea Red Squirrel! Thanks!

 

[SaltyNuts]

On that seemingly awesomely efficient pool cooler ctbaars linked, could you not hook it up inline with several aquariums (or really just clearish plastic tubs to save money and not have to worry about drilling glass), all connected by tubing? Then put in a small pump (if the cooler does not have one built in) to circulate the water slowly. See below (hatched one is the cooling device).

 

[SaltyNuts]

So I'm reading more about the device ctbaars linked, and it keeps saying its "fan based". So is it literally just using fans to blow on water to cool it, like I mentioned the possiblity in my first post lol?!

 

[SaltyNuts]

In the Q&A on the device somebody said it costs maybe $10 a year to operate?!?! WOW. Plus it was developed in, and for weather in, Houston, where I am located lolol!!!

 

[SaltyNuts]

Bump!!!

 

[local]

On that seemingly awesomely efficient pool cooler ctbaars linked, could you not hook it up inline with several aquariums (or really just clearish plastic tubs to save money and not have to worry about drilling glass), all connected by tubing? Then put in a small pump (if the cooler does not have one built in) to circulate the water slowly. See below (hatched one is the cooling device).

No, that would lead to different temps in each tank. Unless that is what you are going for do it like this. Blue cold supply, red warm return. This is how chilled water A/C is piped for equal temps.



Put a valve before each tank to throttle the flow so you could balance the flow in each one.

 

[deadlyapp]

In the Q&A on the device somebody said it costs maybe $10 a year to operate?!?! WOW. Plus it was developed in, and for weather in, Houston, where I am located lolol!!!

As you mention, it is fan based, so basically it is just an air cooled heat exchanger. It operates on the fundamentals that the water must be warmer than the ambient air, although it does have some language about pumping over a medium within the glacier.

I did a quick search and found this : https://images.inyopools.com/cloud/documents/newer-glacier-pool-chiller-manual.pdf which has a bit more information and description of the cooling method, which appears to actually be more like a semi-closed loop chilling tower (like you'd see at a nuclear power plant). It appears that it sprays the liquid into the chamber while allowing air flow, so you get the benefit of evaporative cooling.

You could design basically the same thing by creating a small tower, with sprayers at the top, and fans at the bottom. As you see, they only claim a 10-15 degree drop in temp.

Basically, you can think like this:
- Basic air cooling : This would be running water through a heat exchanger cooled by ambient air. If your water temp was 10 degrees above ambient, and you had a super efficient cooler, then you could get the water temp within a few degrees of ambient
- Evaporative air cooling : This would be using the evaporative cooling effect to get the temperature much closer to ambient or slightly below ambient, however you lose some liquid in this method.
- Thermoelectric cooling : Similar to evaporative air cooling, you can get maybe 10-15 degrees below ambient, however it is inefficient and the electricity cost high depending on the capacity of the TEC.
- Refrigerative cooling : This would be using a secondary liquid (refrigerant) in a standard compression/expansion cycle to cool the liquid. You can get nearly whatever temperature delta you want depending on the refrigerant, the compressor, and the duty cycle (this would be like using a refrigerator to cool circulating water)


It all depends on how much cooling you need.

 

[SaltyNuts]

No, that would lead to different temps in each tank. Unless that is what you are going for do it like this. Blue cold supply, red warm return. This is how chilled water A/C is piped for equal temps.

View attachment 7041

Put a valve before each tank to throttle the flow so you could balance the flow in each one.

I am confused local. Does that device not have a single intake, where the hot water comes in, and a single exhaust, where the cooler water comes out? Why the two hoses? And why, if there is a single intake and exhaust, why would the different tanks have different temperatures (other than mildly different as the tanks further downstream would have slightly warmer water than the ones further upstream of course). Thanks!!!

 

[SaltyNuts]

As you mention, it is fan based, so basically it is just an air cooled heat exchanger. It operates on the fundamentals that the water must be warmer than the ambient air, although it does have some language about pumping over a medium within the glacier.

I did a quick search and found this : https://images.inyopools.com/cloud/documents/newer-glacier-pool-chiller-manual.pdf which has a bit more information and description of the cooling method, which appears to actually be more like a semi-closed loop chilling tower (like you'd see at a nuclear power plant). It appears that it sprays the liquid into the chamber while allowing air flow, so you get the benefit of evaporative cooling.

You could design basically the same thing by creating a small tower, with sprayers at the top, and fans at the bottom. As you see, they only claim a 10-15 degree drop in temp.

Basically, you can think like this:
- Basic air cooling : This would be running water through a heat exchanger cooled by ambient air. If your water temp was 10 degrees above ambient, and you had a super efficient cooler, then you could get the water temp within a few degrees of ambient
- Evaporative air cooling : This would be using the evaporative cooling effect to get the temperature much closer to ambient or slightly below ambient, however you lose some liquid in this method.
- Thermoelectric cooling : Similar to evaporative air cooling, you can get maybe 10-15 degrees below ambient, however it is inefficient and the electricity cost high depending on the capacity of the TEC.
- Refrigerative cooling : This would be using a secondary liquid (refrigerant) in a standard compression/expansion cycle to cool the liquid. You can get nearly whatever temperature delta you want depending on the refrigerant, the compressor, and the duty cycle (this would be like using a refrigerator to cool circulating water)


It all depends on how much cooling you need.

Thank you deadlyapp! I still need to study that link, but looking at your post, some initial questions:

1. I doubt the water temperature in my case (and in fact most cases) would be hotter than the applicable air temperatures - the device is designed for a pool, so of course pool water, which cools overnight and heats up slower, will likely be cooler than the outside air at least until the end of the day. Same with a large number of tanks piped inline I would think. Does the basic air cooling not work if the air is hotter?

2. Refrigerative cooling - is it pretty efficient energy wise, relatively speaking?

3. So, you mentioned building my own similar device. What might one look like? My initial thoughts: A pump pumps the water to the top of a trickle filter. The water trickles down through several layers of air/media (which cools it assuming the air is colder than the water, in addition to providing biological filtration for the water to reduce fish made ammonia, nitrites, etc.). While the water trickles down, several fans flow on it from several directions, causing further cooling through evaporation.

Thoughts?

Thanks!!!

 

[SaltyNuts]

Wholly shiat deadlyapp, I'm looking at that link now, and they have a diagram of the thing, and its almost like I said to a T - water trickles down in the device, and a fan blows on it from overhead. Good lord I am GOOD!!!

 

[deadlyapp]

Wholly shiat deadlyapp, I'm looking at that link now, and they have a diagram of the thing, and its almost like I said to a T - water trickles down in the device, and a fan blows on it from overhead. Good lord I am GOOD!!!

Yep, the overall method is very very simple.

generally, a cooler working on evaporative cooling can get you the 10-15 degrees they mention, however in very humid climates, it becomes more difficult (you need more cooling capacity since the amount of energy evaporation can "consume" is much less).

Refrigeration coolers are about as efficient as you can get, however will generally consume more energy than anything else.

 

[SaltyNuts]

Thank you deadlyapp!

Unfortunately, I live in Houston, and it is humid as heck here. So another question for you. If I build something like that pool device, even a very big unit, water pumped up to like as high as a single-story house, then it filters down through a trickle filter, with fans blowing on the water the whole way down (as energy efficient fans as I can find), how would you guess the overall effectiveness at cooling that water, both from a maxium temp drop and an energy efficiency perspective, would be compared to buy an energy efficient freezer, running the water through it with a copper (or something else very good at energy excahnge) through it would up over and over again, before it returns to the daisy chain of tanks?

Take a look at this link:

https://www.energystar.gov/productfinder/product/certified-residential-freezers/


Some of the least energy costly freezers use 172 K-Watts of electricity as year - compared to the refrigerator I linked earlier, which used 475 K-Watts per year, which according to EPA estimates equates to like $51 a year. So that would be only like $20 a year to operate the freezer! Now, since it will have hot water runnign through it, it will probably take a lot more electricity than that and cost a good bit more, but still, even if it costs several times that $20 number its not really a big deal.

Might the freezer method be the most efficient way to go given expecially that I live in Houston?

THanks!

 

[local]

I am confused local. Does that device not have a single intake, where the hot water comes in, and a single exhaust, where the cooler water comes out? Why the two hoses? And why, if there is a single intake and exhaust, why would the different tanks have different temperatures (other than mildly different as the tanks further downstream would have slightly warmer water than the ones further upstream of course). Thanks!!!

Two hoses to keep the warm water separate from the cool water and allow you to regulate the temperature to each tank. It's really no different from what you had except you eliminate the later tanks having just warm water. But that all depends on the flow. If you are flowing 1,000 gpm through five 55 gallon tanks it would be irrelevant but I am assuming a much lower flow rate. To move enough water to cool all five tanks in a series loop like you had would require quite a bit of flow and all of that flow would have to go through every tank. Running the lines like I did in parallel means you take your total flow rate and divide it by five to get the flow rate through each tank.

And for your previous question about the tower method keep this in mind, the big cooling towers you see on buildings only drop the water about temp 10 degrees. For the DFW area they are designed to have 95° incoming water and 85° leaving water, not sure about Houston. Crunching those kind of numbers requires some engineering that I have never cared to learn, I have engineers for that. The freezer method could very well produce much lower water temps but it would all depend on flow rate and what you are doing inside the freezer to transfer the heat.

Basically I am saying you either need some free engineering or you are going to have to do a lot of calculations yourself and probably a bit of trial and error.

 

[SaltyNuts]

Thank you local! On your revisions to my drawing, is the idea that the tank that gets the cold water last has its hotter water evacuated first, and the tank that gets the cold water first, has its hot water evacuated last, so that they stay closer to the same temperature?

I think everything else you said I follow completely. No doubt some trial and error will be in order!!!

Thanks!

 

[SaltyNuts]

Bump for local and deadlyapp!!!

 

[local]

Thank you local! On your revisions to my drawing, is the idea that the tank that gets the cold water last has its hotter water evacuated first, and the tank that gets the cold water first, has its hot water evacuated last, so that they stay closer to the same temperature?

I think everything else you said I follow completely. No doubt some trial and error will be in order!!!

Thanks!

Not really, all tanks getting the same temp water on the inlet side is the primary benefit. The fun part, since the tanks are not closed systems, is balancing the flow to be the exact same in and out if not the tanks will drain or overflow. Don't pay too much attention to the exact pathing in my sketch I was just copying your loop setup you could just as easily put them all in a straight line.

 

[JEDIYoda]

 

[Red Squirrel]

Not sure if math, or trying to summon something.

 

[deadlyapp]

Thank you deadlyapp!

Unfortunately, I live in Houston, and it is humid as heck here. So another question for you. If I build something like that pool device, even a very big unit, water pumped up to like as high as a single-story house, then it filters down through a trickle filter, with fans blowing on the water the whole way down (as energy efficient fans as I can find), how would you guess the overall effectiveness at cooling that water, both from a maxium temp drop and an energy efficiency perspective, would be compared to buy an energy efficient freezer, running the water through it with a copper (or something else very good at energy excahnge) through it would up over and over again, before it returns to the daisy chain of tanks?

Take a look at this link:

https://www.energystar.gov/productfinder/product/certified-residential-freezers/


Some of the least energy costly freezers use 172 K-Watts of electricity as year - compared to the refrigerator I linked earlier, which used 475 K-Watts per year, which according to EPA estimates equates to like $51 a year. So that would be only like $20 a year to operate the freezer! Now, since it will have hot water runnign through it, it will probably take a lot more electricity than that and cost a good bit more, but still, even if it costs several times that $20 number its not really a big deal.

Might the freezer method be the most efficient way to go given expecially that I live in Houston?

THanks!

As hinted at by an above formula, you'd need to do some calculation, and I'd have to charge you for that

Similarly, you need to look at many factors to determine the cooling effectiveness of your possible option. Tubing material, diameter, are there fins, what's the flowrate, is the air in the freezer stagnant, etc etc etc. I will tell you that you can simplify much of these equations for your purposes.

I would recommend doing some quick research around heat balance and heat transfer equations, heat exchanger equations, etc and you can get some general ideas of how to calculate. Some general points:
- The lower the flowrate, the more heat transfer will occur between the hot water and the surrounding air (Increasing the time of heat transfer)
- Increasing the surface area of any heat transfer surfaces will increase the heat transfer
- Increasing the difference in temperature between the two fluids will increase the heat transfer
- Improving the heat transfer coefficients (eg the materials, the convective transfer (forced air), etc) will increase the heat transfer

You can probably fairly rudimentary find out the freezer cooling capacity, assume perfect heat transfer, assume temperature in of your water, and set some general flowrates and surface areas, and come up with a temperature out of the water.

 

[SaltyNuts]

Thank you deadlyapp, much appreciated!!!

 

[SaltyNuts]

Not really, all tanks getting the same temp water on the inlet side is the primary benefit. The fun part, since the tanks are not closed systems, is balancing the flow to be the exact same in and out if not the tanks will drain or overflow. Don't pay too much attention to the exact pathing in my sketch I was just copying your loop setup you could just as easily put them all in a straight line.

I think I got it - so just to make sure, if they were inline, it would look like this correct? Thanks!

 

[local]

I think I got it - so just to make sure, if they were inline, it would look like this correct? Thanks!

Sure, could even run the lines right next to each other on the same side as long as the inlet and outlet in each tank were far enough apart.

 

[Red Squirrel]

Could even add individual valves for each outlet that are adjusted based on water temp. Just make sure they don't all get closed at once, if no cooling is needed then pump should just stop. Easily doable with Arduino.