Data Center, Colocation, Cloud Computing, Storage, Dedicated Servers Forums

Data Center, Colocation, Cloud Computing, Storage, Dedicated Servers Forums (http://www.datacentertalk.com/forum/index.php)
-   Data Center Design, Development, Building Systems and Operations (http://www.datacentertalk.com/forum/forumdisplay.php?f=37)
-   -   Data Center Cooling – PCU Temperature Discrepancy (http://www.datacentertalk.com/forum/showthread.php?t=76171)

decarn 09-14-2017 08:44 AM

Data Center Cooling – PCU Temperature Discrepancy
 
Hi Guys,

I have two precession cooling units (PCU1 and PCU2) in the Data Center. At any one time, only one PCU will be running. My issue is that when PCU1 runs, the overall data center temperature is higher at 25°C as compared to 21°C when PCU2 runs.

The PCUs are located side by side about 4 feet apart. Both set point were set at 18°C and are running at 100% capacity.
The vendor used a data logger to measure the average temperature return and output of the PCU.

PCU1
Hot Air Return: 22°C | Cold Air Out: 13.7°C | Difference: 8.3°C

PCU2
Hot Air Return: 19.8°C | Cold Air Out: 10.8°C | Difference: 9°C

The vendor explain that since the Hot Air return into PCU1 is higher than PCU2, it is normal that the temperature of the Cold Air out will also be higher than PCU2. Since the difference of the Hot and Cold Air for PCU1 and PCU2 (8.3°C vs 9°C) is about the same, both PCU is running normally. The vendor also attribute the higher temperature around PCU1 to the higher density racks.

We are skeptical on the vendor's assessment due to the following reasons:

1) Why is PCU1 unable to bring down the temperature from 22°C to 10.8°C?
2) When PCU1, the room temperature rises from 21°C to 25°C. However, when PCU2 takes over, it is able to bring the room temperature back from 25°C to 21°C.
3) It is true that the higher density racks are near PCU1 but if a hot spot really exist, why doesn't it impact PCU2?

I perform a few other measurements on my own:

1) Using a basic anemometer, the air flow from the perforated tiles for PCU1 and PCU2 is able the same.
2) I have increased the temperature set point for PCU2 from 18°C to 21°C while PCU1 remain at 18°C. The average temperature from the perforated tiles when PCU1 is running is about 16°C to 19°C. For PCU2, it is 12°C to 15°C.
Surprisingly, PCU2 is still able to supply colder air after I have increased its temperature set point.

Question
1) I will like to seek the experts here how can I bring the temperature of PCU1 to be in line with PCU2?
2) Is the vendor assessment correct?
3) How can to address my skepticism?

Any advice or comments is appreciated.

Thanks.


All times are GMT. The time now is 02:45 PM.

Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2024, vBulletin Solutions, Inc.