Heat Transfer with Recirculating Chiller Problem

[kishanb_01]

TL;DR

Using a recirculating chiller to cool an insulated container via copper tubing clamped to an aluminum plate. Chiller cools water 30°C to 10°C in 5 mins, but plate takes 40 mins to reach 12°C. Seeking insight on why actual cooling is slower than calculated estimates.

Hi All,
So I am trying to use a recirculating chiller to cool/heat an insulated container. This chiller sends deionized water through silicone tubing that enters the insulated container it then is passed through copper tubing in the container before it goes back through silicone tubing to go back to the chiller. The copper tubing fits into a slot on an aluminum plate that I am effectively trying to cool. The copper tubing is also clamped to this plate. The chiller has a temperature display on it that tells me the apparent temperature of the water. It tells me that effectively the temperature can go from 30 °C to 10 °C in 5 minutes. I have attached a temperature sensor to the aluminum plate to find that it takes around 40 minutes to reach 12 °C. I am trying to calculate if this temperature change should be this slow but have found from my calculations that the temperature change should be much faster. Im hoping someone can offer me an explanation as to why my math doesn't match my actual results.

 

[russ_watters]

Are you trending the return water temperature or just the supply? If you have the return water temperature also(wait - do you also not have flow rate?), you can calculate the exact heat removal. Having only the supply temperature and volume tells you very little.

My guess is you're not on the chiller performance curve because of a mismatch in assumed conditions -- likely flow rate.

 

[kishanb_01]

Are you trending the return water temperature or just the supply? If you have the return water temperature also(wait - do you also not have flow rate?), you can calculate the exact heat removal. Having only the supply temperature and volume tells you very little.

My guess is you're not on the chiller performance curve because of a mismatch in assumed conditions -- likely flow rate.

The chiller is operating at a flow rate of 12 LPM. If I were to figure out the return waters temperature with respect to time how could I then find the calculated time to cool?

 

[russ_watters]

The chiller is operating at a flow rate of 12 LPM.

Measured or assumed?

If I were to figure out the return waters temperature with respect to time how could I then find the calculated time to cool?

If you measure it you have the real cooling rate over time:
Q=M* C_p (T_return - T_supply)

Q=heat flow rate
M=mass flow rate.

Trend it and then numerically integrate it and you'll have total heat removed.

If I had to guess, I'd guess that your flow rate is less than you think, so you are removing less heat than you think.

 

[kishanb_01]

Measured or assumed?


If you measure it you have the real cooling rate over time:
Q=M* C_p (T_return - T_supply)

Q=heat flow rate
M=mass flow rate.

Trend it and then numerically integrate it and you'll have total heat removed.

If I had to guess, I'd guess that your flow rate is less than you think, so you are removing less heat than you think.

I am assuming the flow rate given this chart given by the chiller manufacturer and the pressure gage reading on the device

 

[kishanb_01]

Measured or assumed?


If you measure it you have the real cooling rate over time:
Q=M* C_p (T_return - T_supply)

Q=heat flow rate
M=mass flow rate.

Trend it and then numerically integrate it and you'll have total heat removed.

If I had to guess, I'd guess that your flow rate is less than you think, so you are removing less heat than you think.

This makes sense my follow up question is if I am removing less heat than I think what is a possible way I could increase this to get to the cooling capacity specification

 

[erobz]

I am assuming the flow rate given this chart given by the chiller manufacturer and the pressure gage reading on the device View attachment 354695

What is your pressure differential across the chiller or the pump. Inlet pressure - outlet pressure? I ask because you say "pressure gauge reading" - singular.

 

[kishanb_01]

What is your pressure differential across the chiller or the pump. Inlet pressure - outlet pressure? I ask because you say "pressure gauge reading" - singular.

That I'm not too sure The gauge on the device is measuring the outlet pressure of the fluid but I have no way of measuring the return pressure in my current set up.

 

[russ_watters]

This makes sense my follow up question is if I am removing less heat than I think what is a possible way I could increase this to get to the cooling capacity specification

Bigger pump or bigger pipes/less restriction. Though -- positive displacement pump? That's surprising for a chiller. It came with the chiller?

 

[kishanb_01]

Bigger pump or bigger pipes/less restriction. Though -- positive displacement pump? That's surprising for a chiller. It came with the chiller?

The pump did come with the chiller. I may look into getting a new one in that case should I look for one with a higher cooling capacity or better pump?

 

[erobz]

That I'm not too sure The gauge on the device is measuring the outlet pressure of the fluid but I have no way of measuring the return pressure in my current set up.

To me it seems like they are giving you optional pump curves. Anyhow, unless it is a differential pressure meter already plumbed across the pump you need a two measurements to determine the flow. The pressure on both sides of the pump. A one sided pressure in a closed loop is not helpful in regards to flow specification.

 

[Lnewqban]

TL;DR Summary: Using a recirculating chiller to cool an insulated container via copper tubing clamped to an aluminum plate. Chiller cools water 30°C to 10°C in 5 mins, but plate takes 40 mins to reach 12°C. Seeking insight on why actual cooling is slower than calculated estimates.

It seems to me that the heat transfer at the insulated container-copper tubing-aluminum plate side of the system (warming up the water) is much weaker than the heat transfer at the chiller (cooling down the water).

 

[Tom.G]

I see several possibilities:
1) Coolant entering the box is not as cold as you think
2) Coolant flow is slower than assumed
3) Poor thermal coupling between coolant line and Aluminium plate
4) Poor thermal coupling between Aluminium plate and box contents
5) Thermal leak into box from ambient

Check that there is good thermal contact between the Copper tubing and the Aluminium plate. You may have to change to a Copper plate and solder or braze the tubing to the plate.

For instance what is the temperature of the tubing just before it enters the cold-box, and what is the temperature of the Aluminium plate (or the tubing at the exit from the box)?

You may get further information by checking the temperature at various places on the Aluminium plate.

If the Aluminium plate stays close to the coolant temperature, try adding a fan in the coldbox.

For what it is worth, your present, experimental, time constant is a bit over 13 minutes. Which is a close match to the 40 min. to reach ~ 90% of the target; it takes 3 time-constants to get to 93%.

Cheers,
Tom

 

[Chestermiller]

Let's see your analysis/calculation for the plate temperature as a function of time. What surrounds the aluminum plate inside the insulated container? Is the plate sitting on the bottom of the container, or is it somehow supported within the container?