Ok to Oversize a Drycooler?
I would like to know if anyone can speak to any disadvantages of using an oversized drycooler with our glycol based AC units. In setting up a secondary Data Center, we have a good amount of low cost/no cost equipment that is driving us in this direction if it will work.
Here is the basic scenario: Can you run a 150Ton drycooler with only Two (2) 20Ton AC units on it? Are there any drawbacks or concerns that would prevent us from doing this? (Down the road we would most likely scale the load to about 5 or 6 20Ton AC units
Thanks for the help! Everyone on this board is extremely helpful.
Tube velocities affect heat transfer coeffiecients. You should contact the drycooler manufactureer to find out how your specific application is affected.
Its done all the time!
Originally Posted by Razeman
Size the piping for the dry cooler and the pumps. This way when the additional equipment is added there is no adidtional changes required. It works the same if you have multiple units on one dry cooler and only one of the units are operating.
I'd use three way water regulating valves also with multiple units and the pump runs continuously.
Also a dual pump set up with auto change over is desired with critical applications.
You may want to search the white papers at Lieberts web site also.
No offense, but you're giving out bad, uneducated advice.
Originally Posted by cjcargo
Read this tech info on water velocities in heat transfer coils and maybe you'll re-think your post. http://www.usacoil.com/newsletters/aug.pdf
To the OP, the bottom line is that bigger is not always better in this instance. Your water velocities will be almost 75% lower than what the drycooler is rated atvery low, probably 0.5 FPS, which will not doubt affect heat transfer. If you want to find out the extent, the manufacturer can tell you that.
Last edited by flyrfan; 07-23-2008 at 03:55 PM.
I have to agree with cjcargo that this is reasonable if you have surplus stuff. You do want to do some planning to make your life easier when you add to it later, for instance I'd put in tees and valves as future tie-ins.
flyrfan, The way I see it is if you pump 150 tons worth of glycol in the mains and through the drycooler and just balance the flow through the units as you add them then the drycooler will see normal flow.
The downside is you waste some energy running an oversized pump. But if the drycooler is free, what the heck.
I have seen setups where the main is sized for say 150 tons with just a couple of 7.5 ton units and they have roughed in for both more indoor units and a matching drycoolers so with each expansion they plug stuff onto the loop on both sides.
Agreed. If you flow 150 TR of gylcol, that eliminates the water velocity issue.
Originally Posted by Pneuma
The main point of my post was that when you grossly oversize any component in an heat transfer system, you should consult the Mfr (especially in a data center)! To blindly say "it's done all the time" is not the right thing to do. That's my philosopy and it's worked for 20 yrs.
Agreed. But in that case you're matching the heat rejection system to the heat absorbing equipment. That's a totally different scenario than connecting a 150 Ton DC to a 40 Ton load. The piping can easily be oversized; same goes for the the pumps (too some extent).
Originally Posted by Pneuma
I see both your points, but think this way about the issue.
Originally Posted by Pneuma
What does it matter if you do not get full capacity due to low velocity? You only need partial capacity to satisfy the rejection requirement and as more load/flow is needed/added, the capacity will increase too. I would lean towards doing it from what little we know.
Here's my last word on this subject.
I have a client that has a 36,000 SF data center. They have N+50% redundancy on everything - CRAC units, drycoolers, pumps, glycol distribution loops...you get the picture.
Well I was hired to do an engineering evaluation on the system because they were having major problems keeping the GWS temperature at or lower than 115 deg. The DC's were designed for a glycol approach of 20 deg F. So on a 95 deg day, 115 deg GWS should be made (give or take).
As I start looking into the DC performance data & the system flows I discovered that the system flow for each DC (there were (18) 60 Ton DC's) is off the chart of the lowest flow listed in the catalog. So I suggested isolating (4) drycoolers that would get us the desired minimum flow through the remaining 14 DC's. When we did this, we were able to get the approach temps the DC's were designed for.
It's easy for one of us to tell a newbie poster one thingsince we won't have to live w/ the consequences of our answers. If Razeman was my client, we'd approach things logically, not by some "warm fuzzy feeling" approach. I sleep better at night & have less headaches that way.
I appreciate what you're trying to say about not just slapping this stuff in without having a good plan and getting some input from the manufacturer. I think that's good advice.
BTW not to one up you or anything, but since we on the subject, one of the most impressive places I've been in had 700 crac's, chilled water. Everything about the place was amazing, from the chiller plant to the back up generators which could probably power a decent size small town! They even had a well just to backup the city water supply to the towers. They kept two hours worth of chilled water in huge storage tanks...crazy stuff.
Funny you mention about the chilled water storage. At that same faciltiy I referenced, I looked at removing the DC's and installing closed loop towers. The owner wants 2 hours of storge for tower water make-up...thats like 10,000 gallons. We're looking at using one of thier retention basins!