Demin Water System 1

[Icecoffee]

Hi all,

I am in process of designing a chilled demin water system and with some specs here:

Approx 1300 GPM, transient heat load: 10MJ, requirement is provide chilling capacity to dissipate 10MJ within 10 minutes. The design chilled demin water temperature is 50F.

I am thinking having a separate loop for cooling and another separate loop for resin bed (I plan to buy demin water to start and keep "re-cleaning" it along the way). I would like to ask for advise from all of you... about the feasibility of this setup and how to size the bed... and other pointers. THANKS!!!!!!

 

[kenvlach]

DI water will dissolve your system. Use softened water + a corrosion inhibitor.

 

[Icecoffee]

Kenvlach,

would you mind elaborate a bit more on: "dissolve the system"? the equipment and devices which cooling is required asks for DI water... Thanks!!

 

[kenvlach]

What are your materials of construction?

Is this a completely closed loop (recirculating) system?
If so, why need demin water? Also, please clarify what you mean by demin water (TDS, conductivity).

 

[Icecoffee]

Tank material is most likely fiberglass, or polyethylene; piping material are most likely 316 SS

In my design I am assuming a closed loop system.

The reason for using demin water (DI water) is the cooling water will be used for cooling some electronics and RF devices... was told that this is a requirement from the manufacturer.

Currently the requirement is conductivity max = 0.2 us/cm and as the cooling water is running past the equipment... it (I think and please correct me if i am wrong) will pick up ions etc and so conductivity will pick up... TDS I am hopeful on 0.2 ppm... please advise! Thanks!

 

[kenvlach]

The semiconductor industry uses 316L piping, electropolished on the inside, for ~high purity applications.

HDPE is perhaps OK for the tank, but it comes in several grades; an all virgin grade is necessary but don't recall the rest. But tanks are easy to find, and Kynar & Teflon fittings & 316L should be OK. What are the materials of the electronics being cooled, in contact with the water (hope not copper)?

 

[Icecoffee]

Oh ok I see...

Yea.. the problem is some ARE copper... since there are quite a bit of devices needed to be cooled so... therefore I figured a resin bed is need to keep the conductivity down... however I am struggling to size it and all the other assocessories that come with it (I am novice to DI system as you can prob. tell by now)... please advise...

Thanks a bunch!!

 

[kenvlach]

The solubility of copper in pH 6 water at the temperature of your electronics is probably about 20 ppm. To avoid deconstructing the electronics & then then removing the solubilized Cu via an IX bed, better to add some polyphosphate & raise the pH to 8.

 

[GarySCWSVI]

DO NOT ADD ANYTHING TO THE DI WATER. That high quality water is used all the time in semi-conductor and printed circuit board manufacturing applications. It is needed to prevent residue on the components. Much more detail is needed.

Gary Schreiber, CWS VI
The Purolite Co.

 

[bimr]

Icecoffee,

You need to go back to your equipment supplier and doublecheck the water cooling requirements.

Several items in your post do not make sense.

The flow capacity is unusually large for a cooling system.

Copper components are not recommended for demineralized water service.

Demineralized water is not generally used for cooling water.

 

[Icecoffee]

Hi bimr,

Well, the manufacturer specifically asked for DI water; and the capacity is quite large due to the large number of devices, high power RF equipment, that needs cooling. My current problem is .... more like they have given me a "black box" and I have to design around it, so please advise. THANKS!!!!

 

[bimr]

If you want to use demineralized water for cooling, then you must go back to all of the device manufacturers that require cooling water and confirm the materials of construction. Everything in the cooling system should be either non-metallic or made from something like 316SS.

Only a small demineralizer is necessary since the demineralizer will only be used to fill the cooing system. The blowdown from the cooling system should be minimal.

In the cooling loop, you should think about using a cartridge filter to control the crud. You will also need a UV system to control organics.

 

[Icecoffee]

Thank You bimr...

Alright, I have checked with the suppliers and manufacturers and they did say cooling channles are made from copper... and they said should be ok running DI water through...


I am having conflicting information, thought DI eats up copper, so should I size a much bigger resin bed to take care of that? Please advise.

Thanks to you all!!!

 

[bimr]

Demineralized water and copper will be a huge problem. Copper components will definitely not stand up to demineralized water. You would have leaks from holes in the copper within 6 months time from start-up.

Suggest that you get a complete water analysis.

As long as you have decent water quality, you should be able to operate without a demineralizer. You probably should consider a biocide as well as corrosion inhibitor as Kenvlach suggests.

 

[kenvlach]

For protecting copper from DI water (if certain that it's cooling water only, no processing water), use a triazole compound as sometimes supplied in a dilute potassium hydroxide [also gives a slight pH boost]. An initial dose of 5 ppm of the triazole is typically enough to coat the copper in a closed, chlorine-free system & maintain a residual 1-2 ppm.
Several suppliers:

http://www.metalprotection.com/alkaline_water_solutions.htm

http://www.rtvanderbilt.com/spec_hi_3.htm

http://www.alken-murray.com/826pib.htm

Google on triazole corrosion inhibitor for more.

Re biocide: I've used a ACTICIDE® product containing octylisothiazolinone from supplier below, but it was for an open system w/o triazole (an organometallic dye solution), so others may have better suggestions:

http://thor.adept.co.uk/usa/biocideproducts.asp?AppID=3

 

[arguellesmanuel]

hola icecofee, you system is very complex.

1.- DI water is not aproppiate for a chiller (use nitrite 500-700 ppm)
2.- If DI water touch a cooper material (is admiralty?), use azoles (2-3 ppm).
3.- considering a close loop system, < 130 GPM might you reposition water (less than 1%).
4.- TDS in recirculation water less than 30 ppm. (max conditions for welder system.

 

[Icecoffee]

I am running scenarios now... THANK YOU ALL!!!! please jump in if you feel there is more that I need to be aware!

 

[ScottyUK]

Looking at this from a power electronics background, the main reason I can imagine for using DI water is that the heatsinks of the components requiring coolant are likely to be electrically 'live' and the purity of the DI water is required to ensure that electrical isolation is maintained. Adding treatment chemicals to the cooling water will alter the conductivity of the water, and inappropriate choice of chemical could lead to equipment failure.

I have worked on a couple of fairly old and very high power electronic systems which used oil as a primary coolant running to a secondary heat exchanger with closed loop cooling water to an external fin-fan cooler. This eliminated the problem of maintaining isolation while ensuring there was no corrosion problem. Some processes would not tolerate the risk of contamination from a oil-cooled system. Is this a semiconductor application by any chance?


----------------------------------

Sometimes I only open my mouth to swap feet...