Advice for "aquarium" need (cooling water / technical setup / no fish)

[niels123]

I hope my question is allowed here. Please be kind.

I work as a chemist in the lab and want to use an aquarium cooler to cool things down. Mainly because I think they are good enough for my purpose and they are a fraction of the cost of a lab cooler (that goes to -80C / -112F). To get some "heat buffer" I want to use a 64 liter tank and fill it with water that I cool to 4C / 39F.

The equipment that has arrived: Teco 1000H cooler / Eheim CompactON 600 / Eheim CompactON 2100 / Eheim Universal 2400 L/h + hoses. Nothing has been installed, everything is still boxed.

Situation: the 64 liter tank is on a workbench. Next to it is something called a rotary evaporator ("rotavap"). It has a cooling spiral and I want to pump cold water through it. Then close to the workbench (with walking space in between) is a fume hood where I want to have a smaller tank (like 2-3 liters) that needs to be cooled as well. The tubing runs across the ceiling, hence the large 2400 L/h pump that can pump against some height.

The problem is that I'm a bit uncertain how to exactly set everything up and I'd like some advice here.

My first idea was to place the 2400 L/h pump in the large tank and have the hoses connected to this pump and running to the smaller tank. However, the problem is that I now think it looks like that placing the 2400 L/h pump in the big tank won't help much because it doesn't mix the water from the small tank with the cool water from the big tank.

So I'm trying to figure out what the best setup/solution is to mix the water between the two tanks. There are some solutions I am thinking of:
1) Make a T-joint close to the 2400 L/h pump with the T's in the big tank, so the water gets mixed. However, I am worried it can create a flow imbalance and can overflow the small tank.
2) Keep the 2400 L/h pump in a "closed loop" with this pump connected to a copper spiral (for effective heat transfer) placed in the big tank and the hoses ending in the small tank.
3) Use 2 T-joints on the cooler so you connect 2 pumps connected a single cooler. Each pump then circulates the water of its own tank then. But since the pumps are sort of connected together I am worried about a water imbalance and potential overflow of the small tank. Is this a possible risk? The setup is also running unattended and I dont want to come back the next morning to find all the water of my tank on the floor of the lab!

Suggestions / advice welcome!

 

[MnFish1]

I hope my question is allowed here. Please be kind.

I work as a chemist in the lab and want to use an aquarium cooler to cool things down. Mainly because I think they are good enough for my purpose and they are a fraction of the cost of a lab cooler (that goes to -80C / -112F). To get some "heat buffer" I want to use a 64 liter tank and fill it with water that I cool to 4C / 39F.

The equipment that has arrived: Teco 1000H cooler / Eheim CompactON 600 / Eheim CompactON 2100 / Eheim Universal 2400 L/h + hoses. Nothing has been installed, everything is still boxed.

Situation: the 64 liter tank is on a workbench. Next to it is something called a rotary evaporator ("rotavap"). It has a cooling spiral and I want to pump cold water through it. Then close to the workbench (with walking space in between) is a fume hood where I want to have a smaller tank (like 2-3 liters) that needs to be cooled as well. The tubing runs across the ceiling, hence the large 2400 L/h pump that can pump against some height.

The problem is that I'm a bit uncertain how to exactly set everything up and I'd like some advice here.

My first idea was to place the 2400 L/h pump in the large tank and have the hoses connected to this pump and running to the smaller tank. However, the problem is that I now think it looks like that placing the 2400 L/h pump in the big tank won't help much because it doesn't mix the water from the small tank with the cool water from the big tank.

So I'm trying to figure out what the best setup/solution is to mix the water between the two tanks. There are some solutions I am thinking of:
1) Make a T-joint close to the 2400 L/h pump with the T's in the big tank, so the water gets mixed. However, I am worried it can create a flow imbalance and can overflow the small tank.
2) Keep the 2400 L/h pump in a "closed loop" with this pump connected to a copper spiral (for effective heat transfer) placed in the big tank and the hoses ending in the small tank.
3) Use 2 T-joints on the cooler so you connect 2 pumps connected a single cooler. Each pump then circulates the water of its own tank then. But since the pumps are sort of connected together I am worried about a water imbalance and potential overflow of the small tank. Is this a possible risk? The setup is also running unattended and I dont want to come back the next morning to find all the water of my tank on the floor of the lab!

Suggestions / advice welcome!

In this case - a picture (i.e. diagram) is worth a thousand words. Additionally knowing what you're planning to use the system for is also helpful. I.e. it's unclear what the smaller tank is for - How are you planning to get the water from the larger tank to the smaller tank? How often are you needing to change the water etc etc?

 

[mdb_talon]

Option 2 is the only reasonable one out of what you suggest and what i would go with.

 

[niels123]

I made a sketch for 2 scenarios:

First:

Second:

In the first scenario, does the water in the 5L gets mixed in with the water in the 64L tank?

Would the second scenario work?
For simplicity I have omitted the rotary evaporator. It has its own CompactON600 that is in the 64L tank. The rotary evaporator has 2 hose connectors so the connection is completely analogous to how you would connect a cooler to a CompactON 600.

The problem is only how to cool the 5L tank. For simplicity I wrote 5L, but often it is even less (like 1-2L, sometimes bit more.)

 

[niels123]

In this case - a picture (i.e. diagram) is worth a thousand words. Additionally knowing what you're planning to use the system for is also helpful. I.e. it's unclear what the smaller tank is for - How are you planning to get the water from the larger tank to the smaller tank? How often are you needing to change the water etc etc?

Please see the sketches above.
The smaller tank I just fill once manually (and empty when done). The only thing I have is pure water, so no need to refresh it often. There is nothing in it. It's just for cooling suspending lab glass. (Yes boring I know).

 

[ZombieEngineer]

The way you are drawing this, there is no way to do either method without a flood. You don't have a reliable method for a gravity overflow for method 1 to work and method 2 will imbalance and end up flooding one of the two very quickly.

You either need to figure out a way to gravity overflow one tank into the other or you will need two chillers.

 

[niels123]

The way you are drawing this, there is no way to do either method without a flood. You don't have a reliable method for a gravity overflow for method 1 to work and method 2 will imbalance and end up flooding one of the two very quickly.

You either need to figure out a way to gravity overflow one tank into the other or you will need two chillers.

There is no way on earth that I am going to be able to gravity overflow tank 2 into tank 1. A last resort is to use the chiller for only one thing at a time, but it's far from ideal.

In the first scenario I don't see how tank 2 could possibly overflow. My assumption is that a pump will always take in exactly as much water as it ejects. This must be the case, or am I incorrect? Since the pump's in and out are directly into tank 2 with nothing else there can't be a change in water flow rate and overflowing risk?

In scenario 2 I can see that you can create an imbalance because there is some kind of connection between the 2 pumps. What I need for it are 2 T-joints and an adapter to a smaller diameter hose. I'd like to test it with 2 buckets of water in a bath tub. If it works, would that mean it's safe to use in the lab?

 

[MnFish1]

I made a sketch for 2 scenarios:

First:

Second:

In the first scenario, does the water in the 5L gets mixed in with the water in the 64L tank?

Would the second scenario work?
For simplicity I have omitted the rotary evaporator. It has its own CompactON600 that is in the 64L tank. The rotary evaporator has 2 hose connectors so the connection is completely analogous to how you would connect a cooler to a CompactON 600.

The problem is only how to cool the 5L tank. For simplicity I wrote 5L, but often it is even less (like 1-2L, sometimes bit more.)

OK - Great -- I'm confused. What is the purpose of the rotary evaporator - if you have the cooler?

2. Are you working mainly with the water in the 5 liter tank - and the water in the 64 liter tank is just so the you have a reliable source of cold water?

3. IMHO - option 1 would work - if you're just using the 64 liter tank to fill the 5 liter tank - BUT - both options seem overly complicated?

If you are only planning on cooling the 5 liter tank - why not buy a water bath - which you can set to 4 degrees - and put your tubes, etc in that? (That is the way we used to do it). Something like this: https://yamato-usa.com/product/low-constant-temperature-water-bath/

 

[MnFish1]

PS - thanks for answering the. questions - what exactly is the experiment you're trying to do?

 

[niels123]

OK - Great -- I'm confused. What is the purpose of the rotary evaporator - if you have the cooler?

2. Are you working mainly with the water in the 5 liter tank - and the water in the 64 liter tank is just so the you have a reliable source of cold water?

3. IMHO - option 1 would work - if you're just using the 64 liter tank to fill the 5 liter tank - BUT - both options seem overly complicated?

If you are only planning on cooling the 5 liter tank - why not buy a water bath - which you can set to 4 degrees - and put your tubes, etc in that? (That is the way we used to do it). Something like this: https://yamato-usa.com/product/low-constant-temperature-water-bath/

1) The rotary evaporator is a lab apparatus to remove solvents from reaction mixtures under vacuum. Your flask is heated to 40C in a water bath and constantly rotates. Some vacuum is applied and the solvent evaporates. A spiral is cooled with tap water (usually) or coolant and the solvent condenses again and flows into the receiving flask hanging on the left in the picture. When your coolant has a lower temperature, your distillation rate increases.

2) Yes, the 64 liter tank is a buffer. The amount of heat generated by the rotavap is quite big when condensing a lot of low-boiling solvents. Even my 1000H cooler is not able to keep up with it and some increase in temperature is ok. A large-volume water bath will likely work fine as a buffer and is much cheaper than a 1.5-2kW cooler.

3) I do reactions at lower temps, so I need to stir as well. As I wrote: 5L is just a number to represent a small volume. Sometimes I use a 50 mL reaction flask and a water bath of 500 mL that needs to be kept cool for 1-2 days.
Those water baths are impractical: they are big, you can't stir with a magnetic stirrer and stirring bar and they are expensive (several thousand dollars - I see the onces in your link are listed for around $5000).

 

[MnFish1]

1) The rotary evaporator is a lab apparatus to remove solvents from reaction mixtures under vacuum. Your flask is heated to 40C in a water bath and constantly rotates. Some vacuum is applied and the solvent evaporates. A spiral is cooled with tap water (usually) or coolant and the solvent condenses again and flows into the receiving flask hanging on the left in the picture. When your coolant has a lower temperature, your distillation rate increases.

2) Yes, the 64 liter tank is a buffer. The amount of heat generated by the rotavap is quite big when condensing a lot of low-boiling solvents. Even my 1000H cooler is not able to keep up with it and some increase in temperature is ok. A large-volume water bath will likely work fine as a buffer and is much cheaper than a 1.5-2kW cooler.

3) I do reactions at lower temps, so I need to stir as well. As I wrote: 5L is just a number to represent a small volume. Sometimes I use a 50 mL reaction flask and a water bath of 500 mL that needs to be kept cool for 1-2 days.
Those water baths are impractical: they are big, you can't stir and they are expensive (several thousand dollars - I see the onces in your link are listed for around $5000).

There are water baths that are .5-5 liters - and are priced between 500-800$ or so. Most water baths are 'circulating' so - they are being 'stirred'. You are right - there are lots of them that are extremely expensive as well.

 

[mdb_talon]

Back to your original options where 2 is a loop using copper pipe this will work(assuming you use enough pipe). You would want the pump in the small tank to looped copper pipe in the bigger already cooled tank. That will cool it and bring it back to your small tank.

 

[MnFish1]

1) The rotary evaporator is a lab apparatus to remove solvents from reaction mixtures under vacuum. Your flask is heated to 40C in a water bath and constantly rotates. Some vacuum is applied and the solvent evaporates. A spiral is cooled with tap water (usually) or coolant and the solvent condenses again and flows into the receiving flask hanging on the left in the picture. When your coolant has a lower temperature, your distillation rate increases.

2) Yes, the 64 liter tank is a buffer. The amount of heat generated by the rotavap is quite big when condensing a lot of low-boiling solvents. Even my 1000H cooler is not able to keep up with it and some increase in temperature is ok. A large-volume water bath will likely work fine as a buffer and is much cheaper than a 1.5-2kW cooler.

3) I do reactions at lower temps, so I need to stir as well. As I wrote: 5L is just a number to represent a small volume. Sometimes I use a 50 mL reaction flask and a water bath of 500 mL that needs to be kept cool for 1-2 days.
Those water baths are impractical: they are big, you can't stir with a magnetic stirrer and stirring bar and they are expensive (several thousand dollars - I see the onces in your link are listed for around $5000).

Again - focusing on the 5 liter tank. When you say you need to 'stir' - do you mean the flasks you're using - or the water in the bath needs to be stirred. I don't understand why the 2 tanks have to be related at all. But a small circulating water bath (lets say a 5 liter) it seems would solve that 5 liter tank issue - I guess I don't understand why the volume of water in the tank is important - as long as its kept at 4C all the time? Note not trying to debate - just trying to give you a potential option thats less complicated.

 

[niels123]

Again - focusing on the 5 liter tank. When you say you need to 'stir' - do you mean the flasks you're using - or the water in the bath needs to be stirred. I don't understand why the 2 tanks have to be related at all. But a small circulating water bath (lets say a 5 liter) it seems would solve that 5 liter tank issue - I guess I don't understand why the volume of water in the tank is important - as long as its kept at 4C all the time? Note not trying to debate - just trying to give you a potential option thats less complicated.

The reaction mixtures have to be stirred. I suspend roundbottom flasks in the water bath that sits on a magnetic stirrer. Water baths dont have a magnetic stirrer (as far as I know).

The volume of the big tank is a buffer: if you for a short period of time make more heat than the cooler can handle, the absolute temperature increase is inversely proportional to your water volume.

 

[niels123]

Back to your original options where 2 is a loop using copper pipe this will work(assuming you use enough pipe). You would want the pump in the small tank to looped copper pipe in the bigger already cooled tank. That will cool it and bring it back to your small tank.

That's what I think as well.
I also am super curious what happens if you connect 2 pumps to one chiller using T-joints and have one pump's hoses in one bath and the other one in another bath. Will you move water from the bath with the biggest pump to water with the smaller pump?

 

[MnFish1]

The reaction mixtures have to be stirred. I suspend roundbottom flasks in the water bath that sits on a magnetic stirrer. Water baths dont have a magnetic stirrer (as far as I know).

The volume of the big tank is a buffer: if you for a short period of time make more heat than the cooler can handle, the absolute temperature increase is inversely proportional to your water volume.

AHHHHHHH That makes sense There is a published article out there describing the copper tubing method recommended by @Mdtalon - that should work

 

[Beau_B]

Heat exchanger, plate or coil. Either direction from/to buffer depending on space restraints.

 

[niels123]

AHHHHHHH That makes sense There is a published article out there describing the copper tubing method recommended by @Mdtalon - that should work

Thanks! Do you maybe have a link to the article?

Do you have any idea how much water flow you will loose by adding 5 or 10 meters of copper tubing? Is it in the order of 20-50% or more like 80-90%?

 

[ZombieEngineer]

There is no way on earth that I am going to be able to gravity overflow tank 2 into tank 1. A last resort is to use the chiller for only one thing at a time, but it's far from ideal.

In the first scenario I don't see how tank 2 could possibly overflow. My assumption is that a pump will always take in exactly as much water as it ejects. This must be the case, or am I incorrect? Since the pump's in and out are directly into tank 2 with nothing else there can't be a change in water flow rate and overflowing risk?

In scenario 2 I can see that you can create an imbalance because there is some kind of connection between the 2 pumps. What I need for it are 2 T-joints and an adapter to a smaller diameter hose. I'd like to test it with 2 buckets of water in a bath tub. If it works, would that mean it's safe to use in the lab?

You are correct that the input into a pump and the output will be equal. The problem I see is that you are drafting the sketch in a way that would actually be like 2 inputs and 1 output.

If you are planning to use submersible pumps, the input would be the pump body and you have one output tube to work with. If you use a pump in an external setup the amount of cooling you would get from exposure to the 64L tank will be made null by the heat generated by the pump itself. You could increase the heat transfer from the 64L using hear exchanger coil, but this will be less efficient than a chiller and I doubt it will be any cheaper.

The only way to avoid this issue is if one of the tanks were pressurized. In that scenario you could pump water into the pressurized tank and the exact same amount would return.

My suggestion would be to just use two chillers. One for each tank. Makes things a lot simpler.

 

[niels123]

You are correct that the input into a pump and the output will be equal. The problem I see is that you are drafting the sketch in a way that would actually be like 2 inputs and 1 output.

If you are planning to use submersible pumps, the input would be the pump body and you have one output tube to work with. If you use a pump in an external setup the amount of cooling you would get from exposure to the 64L tank will be made null by the heat generated by the pump itself. You could increase the heat transfer from the 64L using hear exchanger coil, but this will be less efficient than a chiller and I doubt it will be any cheaper.

The only way to avoid this issue is if one of the tanks were pressurized. In that scenario you could pump water into the pressurized tank and the exact same amount would return.

My suggestion would be to just use two chillers. One for each tank. Makes things a lot simpler.

With 2 chillers there are some other problems:
1) The chiller requires 500 L/h. Imaging having that flow through a 500 mL water bath. It'll all splash over.
2) The chiller is not cheap and we are not yet certain our ideas (from the chemistry side, not the setup side) will work out to justify having 2 chillers.
3) The chiller is occupying valuable space

I'll just first try the heat exchanger trick and do a small test with 2 pumps on one chiller, just to see what happens (I'm curious). Even though I almost certainly won't use that setup in the lab, I will want to know what will happen.