Minimum air off-coil temperature

[engrrk]

Dear engineers,

For one of my projects, the vendor has selected a computer room air-conditioning unit (CRAC). The selection parameters are as below:
Supply air temperature from the unit = 19 deg C
Return air temperature to the unit = 30 deg C
Net sensible cooling capacity = 110 kW

But, as per the revised design requirements, the return air temperature shall be 23 deg C. With the same CRAC unit selected above (110 kW) and with the revised return air temperature of 23 deg C, the vendor mentioned that the unit can supply the air only to as low as 17.5 deg C but not below it (and hence it can deliver only 56 kW cooling capacity).

Can anyone clarify, which parameter of the coil limits the lowest supply air temperature from the unit? Why can't the same unit supply the air at 11 deg C, so that the sensible capacity is 108 kW or supply at 9 deg C, such that the sensible capacity is 126 kW? (The chilled water supply and return temperatures are 10 and 17 deg C).

This question may seem basic/preliminary, but hope to get a clarification from any of the experts.

Thanks.

 

[tys90]

There are a number of reasons to supply higher temp air in data centers. It's usually for efficiency and so you don't condense on the coil. Why don't you ask the vendor?

 

[cdxx139]

What medium are you using to do the cooling and what are its conditions? CHW, refrigerant etc? This will be the primary function of what your air conditons can be.

knowledge is power

 

[SAK9]

You have stated chilled water supply temperature is 10[sup]0[/sup]C.You would require a coil of infinite size to get to 11C and forget 9C as you it would be a violation of basic science.

 

[ChasBean1]

Assuming your data is with everything at steady state and make-up air has very little influence (typical for data centers):

Your CHW supply is 10°C, your supply air is 19°C and you have an air supply-return air differential of 11°C.

With the same air and CHW flow rates, it would stand to reason that if you provide CHW at 3°C (consider a 10-20% glycol mix) you could achieve a supply air temperature of 12°C and therefore a room (return air) temperature of 23°C.

Any CHW coil should be equipped to drain so condensation should not be a concern assuming the make-up air to the space is very low, as is typical. After time, the amount of condensation would be very low as the space comes to equilibrium at its new condition.

 

[engrrk]

Dear SAK9,

I missed the point that we cannot cool the air to 9C when the water is supplied at 10C. Thanks for bringing up this basic principle.

You have mentioned that to cool the air to 11C, we will need coil of infinite size. Is that right? After going through ChasBean1's response, I realize that when water is supplied at 10C and the entering air is 23C, then the air can be cooled only upto 17.5C, which is the average of 10 and 23. This means, the chilled water heats up and the air cools down until the equilibrium of 17 or 17.5C is reached on both the medium. In that case, I don't think even a coil of infinite size will cool the air down to 11C, unless the chilled water is supplied at a lower temperature. Any comments?

 

[engrrk]

Dear ChasBean1,

Thanks for your reply. I understood your point. In order to cool the air at 19C to 11C, I need to supply CHW at 3C. The equilibrium in air and water side will be reached at 11C. Right?

But how does this apply when the CHWS/R is at 10C/17C and return air/supply air are at 30C/19C? How does chilled water gains 7C, when air loses 11C? What is the relation between these 2 heat transfers?

(If it is a one degree-to-one degree heat transfer between air and water, the equilibrium should be reached at 20C, when both air and water are at that temperature. If the air is cooled down 1C further, then the design CHW return of 17C will become 21C?

If it is NOT one degree-to-one degree heat transfer, then it is not mandatory to supply CHW at 3C to cool the air from 19C to 11C. I can supply the CHW at a slightly higher temperature, say 5 or 5.5C?)

Sorry if my queries are stupid. Just want to get clarified. Thanks.

 

[tys90]

That's not correct. An infinitely large coil is capable of supplying air infinitely closer to the chilled water supply temperature.

 

[engrrk]

Dear tys90,

Thanks. From my reply above, I understand that the temperature at which CHW and air becomes same is 17.5C and hence there cannot be any heat transfer beyond this. This could be the determining factor of how low a CRAC/AHU can supply. Agree?

Regarding your point on efficiency at high supply air temperature, in one of my projects, the CHW is supplied at 7.2C and the supply air is delivered at 14C/15C in order to maintain the cold aisle temperature of 22C. As I am new to data center design, I do not know what is the best practice employed in choosing the supply air temperature.

Dear cdss139, Thanks. The coolign is by CHW.

 

[SAK9]

What drives the heat transfer is the temperature difference (TD)between the two media.As air gets cooled,TD gets reduced across the coil.If you plot the counterflow temperature curves with all temperatures and work out the LMTD it will be clearer to you.By varying the physical properties of the coil(such as number of rows,fins)it is possible to cool the air closer to the chilled water temperature.In air handles that supply 100% outside air,air is cooled from 35C to 13C using chilled water at 6.7C.This is usually done with two sets of coil in series.

In your case if you want to achieve supply air temperature of 11C,you need to drive down chilled water temperature to around 5C.If this does not cut it,try increasing the coil rows to 5 or 6 with closer fin spacing.