[labnetwork] chiller water

[Keith Bradshaw]

Thanks, John....this information is the best thing the labnetwork does.

cordially,

Keith

On Sat, Oct 15, 2011 at 10:57 AM, John Shott <shott at stanford.edu> wrote:

> **
> Keith:
>
> While I'm confident that others will contribute some important insights and
> recommendations to this discussion, let me take an initial stab at
> discussing some of the issues and approaches that we have tried.  Note: I
> don't consider myself an expert in this area, but have certainly had to deal
> with a lot of problems in these areas over the years .... both in tools with
> local chillers and heat exchangers and in tools connected to the
> building-wide process cooling water system.
>
> We've likely had more problems over the years with the building-wide system
> and have tended to go to local chillers and/or heat exchangers when we
> either encounter a corrosion problem or have a resistivity requirement that
> cannot be met by the building-wide system.  That said, however, we also have
> periodically seen problems on our tools with local chillers and/or heat
> exchangers.
>
> Note:  Just for reference, our building wide system is treated with Nalco
> TRAC107 at a 2000 ppm concentration, a pH of 7.5 to 8.0 but it ends up with
> a pretty low resistivity of about only 2000 ohm-cm (500 microSiemens/cm) by
> a closed-loop control system.  We recently had a problem due to a failed pH
> probe that resulted in water with a pH of 6-6.5 .... which begins to be bad
> news for aluminum and mild steel.  So, even a "professionally managed",
> closed-loop control system can have problems and is an ongoing struggle.
>
> Back to your local chiller ....
>
> 1.  I think that it's always good to understand the metals in the system
> .... including parts of the chiller as well as the materials in your machine
> through which the water flows .... to understand how much of a galvanic
> reaction problem you are likely to have.  While we can't always control
> these things, systems that have mixtures of aluminum, yellow metals, and
> mild steel tend to be more problematic.  A lot of tool manufacturers seem to
> include quite a range of metals in contact with the water.
>
> 2. Can you measure the pH of the water in your recirculating tank?  If it
> has gradually drifted away from 7-ish .... particularly if it gets below 6.5
> or so ... that may help to explain more recent problems.  If pH is either or
> low or high I'd try to fully exchange the distilled water.
>
> 3. Do you have a resistivity requirement .... for example, if this chiller
> is cooling RF supplies, then you need pretty high resistivity water.
>
> What are some other options?
>
> 1. Analyzing the chemical content of the stuff that you are capturing in
> your filters is often useful to try to determine what is the primary
> corrosion material .... or whether you may be seeing organic material due to
> algae growth.  I believe that the little systems with open tanks are more
> prone to algae problems than truly closed loop systems because of ongoing
> oxygen entrapment in the water.
>
> 2. Adding a little deionizing cartrigde, resistivity monitor, and solenoid
> so that water can periodically be passed through the deionizing cartridge.
> I believe that a lot of people will tell you that something in the range of
> 0.5 - 1.0 Mohm-cm is about optimum.  High enough resistivity to reduce
> galvanic corrosion, but not so high that the DI water corrodes everything in
> sight.  Note: if you go this route, you generally have a second solenoid
> controlled loop with the deionizing cartridge in it in parallel with the
> main flow loop.  That is because you can't typically get 2-3 GPM through a
> deionizing cartridge so you have a bypass flow channel that routes a little
> bit of the flow through the deionizer when you need to increase the
> resistivity.  Plus, you are only shooting for 0.5 - 1.0 MOhm-cm and don't
> want to be routing too much of the flow through a deionizing cartridge. I
> haven't priced this in a while, but I think that you can get a cartridge
> housing, a resistivity controller, and a resistivity probe for on the order
> of $1k.
>
> 3. There are additives that claim to be able to be added to these
> recirculators to inhibit corrosion with minimum monitoring or precise
> control.  For example, there is some stuff called Optishield (and Optishield
> II and Optishield Plus ....) that you mix in at about a 10% concentration
> with your distilled water.  That is supposed to be good for a year and then
> you empty the lines and refill with a fresh mix.  I have no personal
> experience with this stuff .... but maybe someone else has and cares to
> comment.
>
> 4. We have used 50/50 mixes of DI water an ethylene glycol in a number of
> tools .... particularly if they are going to be cooling to temperatures near
> freezing.  This mix, by itself, doesn't really include much in the way of
> corrosion protection .... and, as I understand it, if it is run to too long,
> can break down into a variety of acids that will once again create problems.
>
> Note: when we've had systems that have gotten significant build up, we have
> used a variety of radiator flush type products ... with a fair amount of
> trepidation .... to try to clear some of the "stuff" out of a system.  I'm
> not particularly proud to admit this ... but desperate times call for
> desperate measures and a machine with a plugged (or nearly so) cooling
> channel is a desperate situation if you can't easily disassemble and/or
> replace cooling channels.
>
> I fear that my conclusion is that we've tried a lot of things over the
> years and still don't have a particularly good approach that has resolved
> these issues on a permanent basis.  Maybe some or our more enlightened
> colleagues will provide a more comprehensive approach.
>
> Good luck,
>
> John
>
>
>
> On 10/14/2011 6:31 PM, Keith Bradshaw wrote:
>
> We have used distilled water in our chillers recirculating tanks.
> Recently we have suddenly seen the inline filters clogged with corrosion.
> We have seen more problems in the last 5 months than we had in the previous
> 6 years.
>
> Any advice on this?
> Should we be using an additive?
>
>
> cordially,
>
> Keith Bradshaw
> Garland, Texas
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20111015/2e29b6a0/attachment.html>