PDA

View Full Version : freecooling



briand
26-02-2009, 01:17 AM
hi
can anyone help with how to calculate the airflow requirements and motor sizing of a system to cool a data centre using ambient air only.plan is to cool data ctr when ambient temp low enough to permit this. ac units still used when neccessary.
size of room 8 mtrs by 8 mtrs by 3 mtrs. heat load 130kw. had several quotes but can tell these would never meet the demand, any help on how best to calculate this very much appreciated.

multisync
26-02-2009, 08:00 AM
Intel have done a lot of research in the desert on just using ambient only.

http://www.itbusinessedge.com/cm/community/news/inf/blog/intel-tests-free-cooling-in-the-data-center/?cs=20341

We are much cooler and yet we have not seen one new one using this method. Most farms are now using integral cabinets from Rittal or similar HP, Airedale Knarp (sp).

http://www.42u.com/rittal-lcp.htm

Thermatech
26-02-2009, 11:31 AM
For sensible heat load
KW = 1.21 × M ³/s × ∆t


So for example your outside ambient air is introduced into the room at 10°C & you want the room temp not higher than 25°C.
then you may be looking at about 7M ³/s supply air volume.


But if the outside air is 20°C & you still require room temp at 25°C then you may be looking at about 21M ³/s supply air volume.

For very low outside ambient temperature the AHU will need some RA mix with SA as you would not want for example -5°C supply air into the room & then have condensation issues.


The implications for air velocity in the room & the size / control of the AHU plant are clear.
There will be some practical limitations for what is realistically achievable.

Introducing some cool outside air for some free cooling will make usefull reduction in power consumption but trying to provide 100% free cooling for most of the year I would suggest be very difficult.

I have myself never yet seen such a system installed in the UK but if you are able to post more details that would be interesting as the resident design experts could take this further.

Ke-Fibertec UK
26-02-2009, 12:35 PM
The draugt problem could be overcome by using a low impulse textile duct, for 7 m/s it would need to be about 1100mm diameter.

Ke-Fibertec UK
26-02-2009, 12:36 PM
The textile duct itself would remain condensation free without himidity control.

Ke-Fibertec UK
26-02-2009, 12:41 PM
http://i17.photobucket.com/albums/b98/adamtaylor69/Animation2.gif
15 posts, and now i can put a link in :-)

Brian_UK
26-02-2009, 11:47 PM
OK, so you now have 15 posts but please do not keep expecting free advertising for your products :(