DIY Chiller

[FC]

KL,
The fans will not help to bring down the temperature at this state. Remember, their performance is tied strictly to the ambient and tank water temperature differential, which is now (great enough and) greater than what the fans can do.

It is absolutely right that the chiller tank (the chill source and power) is colder than the tank (the target where the environment is heating at). Let me put it this way. If one measure the temperature of the air-con close to the louver at slow fan speed, it will read easily below 0 degree celcius so long as the compressor is kick-in. However, the room will never reach anything below 20* degree celius. Now that's 20 degree celcius (or more) different!
* - given ambient temperature above 26 degrees with typical sized air-con.

Now how to bring the tank temperature below 24 or the room below 20 degree celius (I will not cover the room issue here)?
One got to study 2 areas:
1) the cooling target (the tank) and its' environment
2) the cooling source and its' effectiveness in heat exchange

For point 1), there is nothing much one can do to influence (nobody gonna wrap the tank with insulators). The volume of the tank is the only factor useful, for choosing the cooling equipment power.

For point 2), here is where one can do alot and you have done alot. However, the sequence should be:
1) the first one is obviously power, I mentioned:

KL,
How much power is this fridge rated? I suspect your tank's size would required at least 100 watts model.

From observation, yours seems to be slightly below idea but serviceable.
2) locate a cool place (all day) for planting the fridge.
3) As mentioned:

Ronnie's idea of using the ice bags were good to kick start the fellow

4) study the method to achieve the best possible heat exchange efficiency at a given budget.
You should target nothing but all that is in the fridge.
Anything out of the fridge would not yield much return simply because the mother of all heat (from ambient) receiver is the tank and there is nothing you can do about it. The surface of hose and the filter exposed to the elements is neglible and therefore deem unworthy of any attentions.

For your situation, you are left with point 3) to do. And I beleive the possibility of bringing the tank down a degree below 24 is high. Remember the ice forming at the fridge wall is pulling down the heat exhange significantly.

[whuntley]

Let's go back to basics, here.

If the temperature difference between the fish tank and the chiller tank settles down to about 2 degrees, that says the flow is slow enough and the thermal conductivity of the vinyl hose is good enough to sustain that low difference. You don't need metal tubing unless you want to increase the flow rate substantially.

What about the 20 degree plus difference between the freezer walls and the chiller tank? The temperature difference, under stable equilibrium conditions, is a measure of the thermal resistance of the air gap, just as the 2 degrees measures the losses across the tube boundary (and in the external plumbing).

Tape a probe with a foam cover to the coldest wall of the freezer. If it is nearly the same as the chiller tank temp, then it isn't worth trying to fix the conductivity, there, and the freezer is working at full capacity.

Basically, you need to measure all temperature drops and get them each as low as possible. When the freezer's walls and the fish tank stabilize at nearly at the same temperature (+ a few degrees), you have reached good efficiency and only a larger freezer (or external heat-sink fins) can do any more. Any big temperature differences anywhere between freezer inner wall and fish tank represent a loss that is reducing the efficiency of the system.

Use a bit of alcohol (isopropyl rubbing alcohol is very cheap, ethanol has less aroma) to stop freezing in the freezer compartment. That was a great suggestion, for the ice is a good thermal insulator and you want none on the freezer walls to insulate the freezer's coils from the chiller tank water and certainly none in the chiller tank. Thermal convection should be all the pumping you need if solid ice cannot form. [Slush is probably OK, but IDK.]

Wright

[TyroneGenade]

Tyrone, I don't think it would make a difference. As I've said, I can get the chiller to become very cold in a short time by switching off the filter but it won't help because the fish tank will warm it up as soon as the filter starts running. The point I've been stressing is that the chiller tank and the fish tank are not 2 entities, as I've initially thought. They are one.

KL, your freezer is the heat sink. For it to work there needs to be an efficient exchange of heat between it (the freezer) and your tank. By having the extension of your tank (the hose coil) isolated in a glass tank (the chiller tank) not only does a temperature gradient form between the hose and the chiller tank but also between the chiller tank and freezer. Both gradients are filled with air which is a damn good insulator when not in circulation. By taking the hose out of the chiller tank and putting it in direct contact with the freezer you are removing that one gradient which will almost certainly buy you a degree or two and this is why:

The rate of temperature exchange is proportional to the difference in temperature multiplied by itself four times as well as the heat capacity of the substance. A large temperature difference exists between the hose (your tank) and the chiller tank so heat exchange is relatively fast; but only a small difference exisits between the chiller tank and the frezer and to make matters worse there is all they still air which isn't conducting heat so the heat exchange between the chiller tank and freezer is slow. This means that the chiller tank cannot keep pace with your tank as a consequence the thermodynamic equilibrium reached between the hose and the chiller tank is not the equilibrium that would be reached if the chiller tank were disposed of.

Slush is probably OK, but IDK.

My dear Wright, slush is much better.:-) The temperature of an ice slurry is lower than ice alone. I think several degrees lower. Mix in some alcohol and potassium salts and you can make that ice slurry very very cold. Just ask your good friend Mr. Farenheit.

Ciao

[whuntley]

[quote ]Slush is probably OK, but IDK.

My dear Wright, slush is much better.:-) The temperature of an ice slurry is lower than ice alone. I think several degrees lower. Mix in some alcohol and potassium salts and you can make that ice slurry very very cold. Just ask your good friend Mr. Farenheit. [/quote]

I suspect you meant Fahrenheit?

We are not talking temperature, here, but how to move heat from fish tank to exterior of the freezer (the room). The drop in temperature is a direct measure of how much thermal resistance is present at each stage. If a large temperature difference exists between the wall of the freezer and the chiller tank, then slush may be much worse than liquid at reducing that temperature drop, because convection currents can't aid the transfer of BTUs. That's what I did not know.

Frost, ice and (maybe) slush are your enemies, here. The function of the antifreeze is to keep thermal transfer high by keeping the fluid liquid. The absolute temperature is not lowered by making slush (except in water where ice has an anomalous coefficient of expansion). Once you add much antifreeze, the actual temperature of the slush can be any old thing, but you want to avoid any condition that allows a bigger temperature drop in the heat path. In this case, convection currents move warmer fluid to freezer walls to reduce the temp. difference. Slush may slow that process and allow a higher temperature drop, hence reduced efficiency.

In a somewhat ideal case, tyou are trying to raise the inner wall temperature of the freezer to something like 16C. If you have a 2C drop to the chiller tank, and another 2C drop to the fish tank, you will have the desired 20C in the fish tank. [Note that this would not allow any ice to form if dynamically stable.]

As I understand it, there is frost on the freezer wall so it is at about 0C and the bulk of the heat resistance is between the wall and the chiller tank. Reducing that loss (temperature difference) deserves a lot of thought.

BTW, there is no practical way to get the coils in direct contact with the freezer walls as shape and volume dictate that 99% of the coils would be in air, and the same heat block would be present and maybe worse.

Therefore:

a) You need a heat-conductive fluid coupling the freezer walls to the tubing walls. The present drop of essentially 20C is unacceptable and very inefficient. It makes the system more a room heater, and not a tank chiller.

b) You need to avoid formation of ice, frost, or slush by adding some antifreeze if they ever tend to form. This will tend to be true when you get the system so efficient that the freezer control cycles it on and off. You need to allow for the deeper drop (maybe below 0 in spots) during the on cycle to keep the 16C average.

Wright

[TyroneGenade]

I worry that with the coolant fluid the tank's temp may drop much more than 1 or 2°...

The ice on the sides of the freezer suggest to me that the chiller tank is the problem. If there isn't ice on the side of the chiller tank then the chiller tank is not cooling enough to be effective and this has nothing to do with the tank's temp. Did ice form on the chiller tank walls when the water flow was off?

tt4n

[whuntley]

I worry that with the coolant fluid the tank's temp may drop much more than 1 or 2°...

That's what the temperature-control thermostat on the freezer is designed to handle.

The ice on the sides of the freezer suggest to me that the chiller tank is the problem. If there isn't ice on the side of the chiller tank then the chiller tank is not cooling enough to be effective and this has nothing to do with the tank's temp. Did ice form on the chiller tank walls when the water flow was off?

Ice on the freezer wall (0C or below) and none on the chiller tank (at 22C) is just proof that the air is a good insulator (assuming glass temp. drop is minimal) and that one needs to get it out of the system. The chiller tank is doing an excellent job of keeping the fish tank at nearly the same temperature (2C difference). Removing the air insulation is needed to move both lower. One hopes that does too much, so the freezer control can be set so it doesn't run all the time and burn out early.

Wright

[timebomb]

KL,
The fans will not help to bring down the temperature at this state.

You know what, Freddy? You're right on the dot with that one. The fans not only did not bring the temperature down but for some strange reason, the temp went up by 1 degree the next morning. So I switched the fans off and turned the settings of the freezer to midway. Not much point setting it to maximum when that doesn't help much.

This morning however, the thermometer went below 24 C for the first time. It's probably due to the fact it rained last night but hey, who cares? I got 23.2 C. Hurray!!!!!!!



There was even some condensation on the sides of my fish tank. I checked the chiller tank and the temp there was 21.5 C.

You should target nothing but all that is in the fridge.

Okay, friend, I'm taking your advice. But don't swear at me when you find out what I just did. I decided to do something about the water in the chiller tank. I added a powerhead. Okay, I know, all of you have been saying this won't help to cool the tank but warm it up instead. But hey, I'm a skeptic, remember? I don't believe anything until I see real evidence . Here's a picture of my chiller tank with the powerhead installed.



Wright and Tyrone,

I appreciate your suggestions very much but I have to be honest and say I don't really understand what you wrote. I flunked Physics when I was in school, remember? Sorry about that, old chaps but you have to forgive me as I'm pretty dumb with many things. Anyway, the picture above shows some ice on the sides of the freezer walls and they stick to the glass. There's a big patch of it right behind the power head. The corners of the chiller tank are free from ice though and I wonder why this is so.

Loh K L

[PohSan]

The corners of the chiller tank are free from ice though and I wonder why this is so.

Maybe the cooling coil is inside the center of the four freezer walls. Since the ice is touching the tank, there is no more insulation between the freezer and the tank. I believe that the DIY chiller will be now working more efficiently.

Regards,
Ong Poh San

[TyroneGenade]

Ever the prophet of doom that I am, but... with all that ice forming you should perhaps be worried about that chiller tank cracking in two or three or four... I still say get rid of the chiller tank.

[timebomb]

I still say get rid of the chiller tank.

I can't, Tyrone. The chiller's compartment is not meant for holding water. The guys selling them all say it will leak. As for the glass cracking, I was kind of worried about that but someone said it won't happen because it's equivalent to leaving a mug of beer in the freezer, he said.

Loh K L

[TyroneGenade]

No the tank cracking from ice inside but cracking from ice pressure forming from the outside. Ice forming around the hose is not problem because the hose can deform but the ice build up against the glass can perhaps cause problems. I am not a structural engineer though so my opinion isn't well weighted (to borrow a SG term).

[whuntley]

I agree with Tyrone that ice against the outside of the glass chiller tank could eventually crack it. It is a potential problem, but not at this stage where it is just thick frost.

Looking at the picture with frost touching the glass, I'd be inclined to guess that the internal pump will be a big help, depending on its power and efficiency. Even a small airlift system, like for UGFs would be enough to stir the water with minimal heating. The frost touching the glass isn't much better than air for effective heat transfer, BTW.

Let me try to clarify what Tyrone and I have been discussing. [I do not agree with some of Tyrone's points, BTW.]

Think of heat flowing as like an electrical current flowing through a conductor. A resistance will drop some Voltage, just as a thermal resistance will cause a temperature drop.

When heat flows out of your aquarium and is sent to the outside world by the freezer's output (hot) coils, the energy must flow through different parts of the system. When heat energy flows from one place to another (without any outside input of energy) then the flow will be from the warm tank to a place with lower temperature. The temperature difference is a direct measure of the thermal resistance between the warm and cool place for that amount of heat transfer.

Right now, it is only 2C or less from fish tank to chiller water, so the heat transfer efficiency of the coils is quite good at the present low heat flow rate. Since the freezer walls are making ice from condensed atmospheric moisture, they have to be at 0C or colder. All of the heat that flowed from fish tank to chiller water is being removed by the freezer. The 22C drop from chiller to walls is a good clue that the air is strongly insulating the chiller water from the freezer's coils.

It may actually be rocket science, but to any electroniker/laser guy, rocket science is pretty trivial.

Freezers should not be designed to leak. A freezer where a power failure allowed the ice to melt or the ice cream to run into the mechanism wouldn't last long in the market. If the inner case is strong enough to hold ice cream, it should be able to hold a considerable weight of water. Of course they are not designed to hold water, but they should, anyway. [A really good skeptic would question that store advice. :wink:]

Even if there are slow leaks from seams, a bit of good old aquarium cement (silicon seal) will easily fix that.

Can you add a few inches of water to the freezer compartment without flooding the carpet? If you can, try it.

I would expect the fish tank temperature to drop, the chiller tank to drop considerably more (i.e., more than the current 2C difference) and the freezer control would have to be set up to kick it off to keep your fish tank above 20C.

The leak risk is limited by the chiller tank, so only the small amount of water surrounding its base would be at risk. A few towels could absorb that much, even if it failed catastrophically. No?

Look at where the frost is forming, and assume that is close to the freezer cold coils. The water would probably have to be about that high for my suggestion to work, so the risk could well be more than you want to take.

What are balconies for?

Try it, You'll like it!

Wright

Who is safe in CA if your downstairs neighbor comes after you with a cricket bat.

[timebomb]

[A really good skeptic would question that store advice. :wink:]

Ouch you got me there, Wright.

Okay, I'm going to do as you suggested. But I will not fill up the freezer compartment with water as there are gaps in the seams. It will surely leak.

I checked this morning and found that the ice has retreated from the chiller tank. Where the ice was touching the glass before, there's now a small gap. The powerhead is doing a good job preventing the chiller tank from cracking under pressure and good thing about it all, there's hardly any difference in the temperature. Here's a pic:



I went out just now to look for giant plastic bags. They are not easy to find but I managed to buy some from a shop along Haig Road. Here's a pic:



I will be replacing the chiller tank with this plastic bag. I'll double-bag of course as there's a chance the plastic will be punctured. I don't know if there's enough time to work on this today. I promised to bring the family to a movie. We're going to watch "Predator vs Aliens". I love the Alien series of movies; my daughters and I are big fans.

Maybe tomorrow. I'll let you all know the results. If this works, I'm going to take a cricket bat to the guy who suggested I use a glass tank. I'm going to claim the $40 I spent on the tank from him .

Loh K L

[FC]

KL,

The fans not only did not bring the temperature down but for some strange reason, the temp went up by 1 degree the next morning.

Yes, that’s expected.

So I switched the fans off and turned the settings of the freezer to midway.

Yes, so long there is ice at the Wall and nothing is done to twart it, the fridge can practically be switched off and still getting the 24 degree celius at the aquarium.
Let me put it another way:
The cooling coil (behind the Wall) is running at easily below -10 degree celcius. By allowing ice to form at the Wall will, effectively, build an insulation slap that limit the temperature at a useable 0 degree celcius. It is worse than the insulation of the air gap (between the Wall and the Glass) because air can go far below -10 degree celcius.
- That’s why there is a heater (not many people are aware) in our no-froze fridge used (at a set program) to melt any ice formed on the coil.

I decided to do something about the water in the chiller tank. I added a powerhead.

The 2 degree celcius difference between your Chill Tank and the Aquarium indicated the 50 m hose is doing a fantastic job in the heat exchanging. Trying to bridge the gap further is meaningless and ineffective.
You see, the difference between the cooling coil (assume -10) and the Chill Tank (22) is 32 degree celius! Now, it is obvious this is the place where any slight effort would yield great benefit.
- Imagine that the difference were to be improved by just 10 degree celius, that’s the Chill Tank will be 12 degree celcius (all the time), the Aquarium will easily be lowered by at least 4 degree celcius to 20.

There's a big patch of it right behind the power head. The corners of the chiller tank are free from ice though and I wonder why this is so.

I concur with Poh San’s explaination.

[timebomb]

You see, the difference between the cooling coil (assume -10) and the Chill Tank (22) is 32 degree celius! Now, it is obvious this is the place where any slight effort would yield great benefit.

Okay, fellas,

I did as you all suggested.

Yesterday, before leaving for the cinema, my wife and I siphoned out the water and took out the chiller tank from the freezer. We then replaced it with the big plastic bag I bought. I decided to use only one as double-bagging may create air pockets which you all said are no good for chilling the water.
Here's a pic:



It was a bit difficult to tie up the bag nicely so I cable-tied both ends of the opening, rolled up what was in between and use clothes pegs to hold it in position. Before that, I put in the powerhead. Here's a pic:



After everything was rigged up nicely, I switched off the filter and started the freezer running. I then went for the movie with my family. Great movie but it can't compare to the Alien series.

Okay, back to topic. When I came home from the cinema, the temperature in the chiller bag was about 13 C. I started the filter running. This morning, I woke up hoping to see something really cold. Shucks. The temperature of my fish tank was 24.5 and that of the chiller bag was 22.7 C.

Well, folks, so it looks like I was right all along. We're back to square one.
I should have been more of a skeptic

Loh K L

[timebomb]

Hi, folks,

Using a plastic bag instead of a glass tank to hold the water in the freezer didn't make any difference but I noticed something this afternoon. It could be I just wasn't paying attention before but today, I saw condensation on the filter hose, the one that goes from the freezer into the fish tank. Here's a pic"



Does condensation indicate anything, other than the water coming out from the freezer is cold?

Loh K L

[TyroneGenade]

Condensation suggests you have high humidity and a cold hose.

I would be patient with the chiller bag... You have really upset the apple cart and it will take a while to stablize. It took 2 days wasn't it for the chiller tank to take the tank down from 28 to 24°C so it may take that long to get it down further. Maybe you should get rid of the powerhead and replace it with an airstone as someone suggested. You might win back 0.5°C.

By all rights you should have much better heat exchange between the tank water and freezer. It should be cooler. Much cooler. Perhaps one must just be patient.

Let us know how things progress.

On the bright side you now have another fishtank.

[timebomb]

I would be patient with the chiller bag... You have really upset the apple cart and it will take a while to stablize.

That could be a factor, Tyrone. I'll wait another few days and if the situation does not improve, I will put back the chiller tank. I feel it's safer with the tank as chances of it leaking are lower.

On the bright side you now have another fishtank.

That's a good one I had a good chuckle over it.

Loh K L

[whuntley]

KL,

Before this last change, you had set the thermostat higher on the freezer. Did you remember to turn it back down to max cold for this experiment?

If you did, then I think you just moved the temperature drop out to the walls, but have proven that your freezer only has the capacity to cool your tank a few degrees. Insulating tubing and cannister will get another few 10ths but possibly isn't worth the trouble.

The only thing that can do more is externally cooling the freezer with finned aluminum heat sinks or a fan or two.

Wright

[FC]

I strongly think that the present setup is gonna bring the aquarium temperature to at least 23 degree celcius, 22 is very very likely. Give it 1~2 days more when the bag's water goes below 20 degree celcius.

I guess it would reach an equilibrium of 18 degree in bag and 22 degree in aquarium with full thermostat setting. If it does not, it goes back to my first skeptic:
fc wrote:

KL,
How much power is this fridge rated? I suspect your tank's size would required at least 100 watts model.

The pump you place in there will only be needed when the bag's water achieve below 18 degree celcius.

Remarks:
Sealing the bag is a good thing to do as it prevent moisture escape from it that was the water source for the ice formation at the Wall.