Burying Chilled Water Piping

[Vicsidhu]

I am desigining a Computer Room with raised floor and suspended ceiling. I was thinking of running chilled water piping (supply and return from Chiller) underground (buried). The reason is I want to minimize the risk of flooding and avoid running it under raised floor.

This is in London, Ontario, Canada and frost line there is 4' below grade. Since this pipe will be under the computer room (which is in the lowest level), I can run it at 2' below.

My questions are
Is it ok to bury chilled water piping?
Is it normal to do this?
Do I need to insulate?
Am I into x-raying for leak testing or normal pressure testing is good enough?

Thanks.

 

[GMcD]

Fundamentally there is nothing wrong with burying the CHW pipes as long as the materials and detailing are done right. I would suggect that you look at the pre-insulated plastic PEX piping systems that are available from Wirsbo or Stadler Viega, or the more common pre-insulated steel pipe systems:

http://www.permapipe.com/product_catalog.asp?System=Low Temp Systems #sysapp

http://www.bekkonsult.lv/en/tehnologijas/izol_caurul.htm

http://www.heatinghelp.com/pdfs/71.pdf

 

[lilliput1]

Are you sure you have chilled water year round? If not consider glycol dry coolers & glycol economizer coils. It is common practice to run piping under the raised floor. Run a loop for reliability & provide taps for future. Provide leak detector & alarm. Provide drain under the raised floor with trap primer. Make sure access floor is deep enough to accomodate piping. Meet seismic code requirements.

 

[CountOlaf]

I agree with lilliput. It is fairly common practice to run the chilled water piping (as well as condensate drainage and city water for the humidifiers) under the raised floor in computer rooms. A low raised floor (say 12") does cause some design problems in terms of crossing pipes, etc. which are insulated thereby increasing their overall size.

Liebert (and I'm sure many others), who manufactures the air-conditioners as well as the UPS equip. for computer rooms also markets underfloor leak detection equipment (various lengths of cabling and the alarm/controllers). It is common to loop some cable around the perimeter of each AC unit under the floor and sometimes to add those "general area" detection cabling/systems. Which reminds me: even if you did run the chilled water piping underground, you would still have to come-up into the raised floor area at each AC unit.

The other benefit of having the water piping in the floor is for future connections or revised layouts. They always want more cooling in the future.

Now, regarding underground chilled water piping, it is fairly common (I imagine) for say central chiller plant type systems at a campus where mains are run outside underground to other buildings (although sometimes tunnel systems are built for easier access to the pipes). For larger size piping (3" plus), you could use ductile iron (similar to domestic water mains) and I've even seen black steel (sch 40) specified. For insulation with the steel pipe, I've seen Gilsulate 500 or Dri Therm specified which is not only insulative (is that a word?) but is a sort of backfill. There are also containment piping systems (pipe within a pipe) as manufactured by Rovanco. You can get these with insualtion, air spaces, etc. and various types of inner and outer piping materials.

 

[Vicsidhu]

Thanks GMcD, CountOlaf and Lilliput1 for the helpful information. Here is my feedback.

Why Burry?
We always design chilled water and all the other piping under the raised floor, with all the bells and whistles – leak detectors (Liebert Liquitechs) and lek-tec water detection loop around the room – so on and so forth. I will be doing this anyways as I will be coming into the AC units from under the raised floor.

In this case I am thinking of burring the chilled water pipes as I am already excavating for extension of sanitary line for floor drain under raised floor. We will only have to just make the trench a little bigger and lay in the pipes for good. It makes the overall design simple and foolproof, if my capacities are in line (which they are).

My current load is 15 tons. I am installing two 18-ton lieberts (redundancy) with future connections for 18-ton capacity heat exchanger for water based cooling of the high density racks. So capacitywise I am ok.

Why Chiller?
This is to provide redundancy in air-conditioning. Here my primary cooling will be chilled water – which will run chilled water in a coil in the ac unit. Secondary cooling is with DX system with condenser on the roof. Drycooler will just give me free cooling and no redundancy. Furthermore, my run is not big to go to drycooler.

I think I will be using Sch. 40 steel piping, but am debating on the insulation part. My question remain as: Why do I need to insulate the pipes? It will buried at a fairly constant low temperature (1-5°C).

Pressure testing of the pipes:
To ensure there are no leaks, I am calling for pressure test at 400 psi (design pressure is 250 psi) for 48 hours. Same pressure test is administered until after compaction and final pour of concrete. Any leak will be detected during the testing. Any ideas there?

Thanks for your input.

vicsidhu

 

[CountOlaf]

You'll have to convert to SI but here goes:

Normally (for me), I'd be running the chilled water at 45 deg F. I usually assume a 55 deg F ground temperature (ground source heat pump applications) or so, which means that the chilled water supply pipe will be picking up heat from the surrounding soil before it gets to the Liebert units, hence the insulation. You may of course have different conditions where you're at (but I'm in Detroit so we're not that different from a "climatic conditions" standpoint).

Additionally, if you run steel pipe, you have to consider corrosion, cathodic protection, etc. which is another way the insulation or containment pipe systems help out. If you really don't want to or are convinced it's unnecessary to insulate, perhaps you should consider ductile iron (I assume you'll need 3" mains for 36 tons) or even copper type K for the smaller branches (although I'm not sure about the pressure limitations) as these types of pipe materials (in addition to some of the plastic varieties) won't have the corrosion issues.

Now, your test pressure seams pretty high, as does your working pressure. I know the typical spec calls for 1.5 times working pressure, but I think some of the codes allow for a maximum if your components (valves, etc.) have a lower pressure rating. Are your pumps really generating that high of a pressure? Anyway, just make sure you isolate those components which could be damaged at 400 psig.