Hi all,
Thanks so much to the sixteen of you who responded to my June 28
question. It's great to see that there's a consensus solution to the
problem, and that we're not the only ones having difficulties with our
chilled water infrastructure.
I'd like to point out that this post dovetails with the one Carlos
Pacheco recently posted, entitled " Chillers for water-cooled
compressors for cryoprobes ". His nice June 14 summary included a lot
of specific information on particular kinds of closed-loop
recirculating water chillers. My question pertains to different ways
of thermally interfacing the cryoprobe with air or water in the
building, and including such a chiller is one of the more popular
solutions.
There seems to be a consensus that chilled water from the building is
rarely used to cool cryoprobes directly because of the instabilities
in pressure, temperature, and flow associated with sharing the water
with other labs. Most people handle this by using a "water-to-water"
heat exchanger that uses a closed loop of water to remove heat from
the cryoprobe and transfer it to the building's chilled water. This
ensures that the NMR facility has complete control over the material
circulating through the cryoprobe, thus avoiding particulate problems
common with shared water lines, and reduces the impact of fluctuations
of pressure and temperature in the building system.
Disturbingly, some people reported that their chilled water systems
were so problematic that they resort (part- or full-time) to using tap
water for their supply and letting their warmed water go down the
drain. At about 2-5 gal/min, this is a lot of tap water! You'd think
that host sites would recognize this waste and rig up a better
solution, but too often institutional barriers make it impossible.
I've left allusions to this solution out of the responses below at the
request of some respondents.
Here are the edited replies I received. Thanks again for all your help.
- Josh
****
I wish I had solutions but instead I just have more anecdotal tales.
Our facilities has spent more than five years and millions of dollars
installing a very complicated chilled water loop with redundant
backups (for our 24/7 resources), heat exchanger to isolate our
department from the main loop (to reduce / eliminate contamination),
on-demand pressure pumps, blah blah blah.
It doesn't work. Too hot, not enough pressure differential, not
enough flow, junk in the lines, pumps break down, filters clog, fail-
safes fail, programming doesn't . . .
Way way way over engineered, maybe way way way under engineered. I
don't know. From what I gather, your experiences are not uncommon.
****
if there are issues with the house cooling water it is advisable to
install a secondary closed loop. You then have a closed loop
connecting a pump, heat exchanger with the cryoplatform and the heat
exchanger is connected to the house loop.
This way you have control over the quality of the cooling water for
the platform.
****
At XXU we also use a Haskris closed circulating water bath for our
Varian cold probe. When our 700 was installed space was at a premium
and I was concerned about the amount of additional heat I would be
pumping into the room so I opted to install the Haskris in the boiler
room across and down the hall from the NMR lab. We plumbed the Haskris
in using copper pipes which are insulated. The pipes terminate with a
ball valve and red rubber hose is used to connect the ccc to the valves
on the pipe. We have only used water in the Haskris and have been
running
in this configuration for a bit over 5 years with no sign of degradation
of the hoses. Putting the Haskris in a different lab, away from the
spectrometer,
I think, was a wise decision. However, had other space been available I
would have chosen not to put the Haskris in a boiler room. It is always
hot there and the heat takes a toll on the Haskris. I just replaced the
refrigeration compressor on the unit.
***
We have a Haskris closed loop system that is cooled by the building
cold water which is on a branch off the main loop.
That avoids the hot air problem you have. We just let the building
water flow through the Haskris, with a coarse particle filter, but we
have not had much in the way of junk in it. The Haskris then regulates
the cooling of the closed loop system by regulating flow of house
water through heat exchanger.
We have been that way for about 3 years and it seems to work well.
.
<<JK: To be sure, the Haskris you're referring to is a simple W-W heat
exchanger, not one of their units with an additional refrigerator? >>
Yes its just a W-W exchanger and has no additional refrigeration.
We run this setup on both a 600 and a 900 cryoprobe. Although the 900
has only been in for a few months.
***
< one person didn't want their whole story posted, but made reference
to the vulnerability of the "capsule" inside a Bruker cryoprobe.
Apparently, this component can be damaged by heat, so if the cooling
water fails and is later restored, the cryoprobe may start back up
again, but the damaged capsule may fail much sooner than expected. >
***
we primarily use "water to water" systems.
Haskris makes a ww1 ww2 & ww3 model.
We filter the building circulating water with 100 micron filters, split
into two separate paths so the filters can be changed periodically
without disrupting flow. The closed cycle bath is filtered through 5
micron filters and remains clean for the compressor heat exchanger.
The best advantage the WW systems offer is they don't put any
significant heat into the room. The house chill loop must be between 55
& 80 degrees F to operate properly.
<< JK: I noticed the WW units come in refrigerated and non-
refrigerated types. Which do you use? >>
Non because our house H2o is _at_ 50 degrees
***
We have two cold probes in our facility and a building chilled water
loop. We decided to use the Haskris heat exchanger system to cool the
He compressors and it has worked very well. The Haskris has it's own
water reservoir and pump and circulates clean water to the He
compressor. We also have a 5 micron filter in this line to ensure the
water going to the compressor is clean. The water in the Haskris
itself is then circulated through a heat exchanged which is cooled by
the building chilled water. This indirect system seems to be able to
absorb fluctuations in the building water quite readily and is not
quite so dependent on clean water (we have a sieve to filter out big
particles, but that is it).
***
We have a bank of filters protecting our Cryoprobe compressor. There
are 2
stages, 10 micron then 1 micron. The 24" long filters have plenty of
flow
alone, but we put 4 in parallel to get more time between filter changes.
I have not seen problems with rubber hoses, but your building
maintenance
people should be able to tell you what they're putting in there and
what is
safe to use with it. Silicone tubing seems to be the most
chemical-resistant. The compressor heat exchanger will grow crud
regardless
of what they put in the water, so you might need a regular schedule of
back-flushing it anyway. We backflush ours every 6 months with a
small lab
pump designed for acidic commercial heat exchanger cleaner.
How bad are the flow fluctuations? I would see if the flow can be
sorted
out and stay on the cooling loop if possible. Because the building is
new,
sorting out the cooling loop problems now could mean a lot to not only
your
operation, but also others. Adding any more pumps is just more
maintenance
headache, IMO. If you do decide to add a pump, why not just put it in
series with your cooling loop and boost their water through your
compressor
instead of relying on their system pressure? You could use a flow
switch to
turn on the pump to boost the loop flow only when it is low. As long
as you
are returning their water to the loop, there should be no problem with
you
running your own pump.
We use tap water, and that might be a good back-up source for you. You
could put a flow sensor in the cooling line, and automate when it
senses low
flow to open a valve to augment cooling with tap water. Always put
check
valves in any design that combines 2 water systems!
***
We have been using building chilled water, but due to high pressure
and the possibility of floods from hose leaks, have put a heat
exchanger in the building line. We have only a small reservoir of
water flowing in a close loop controlled by a small electric pump to
cool the compressor. As the warm water exits the compressor, it flows
through the heat exchanger and is cooled by the chilled building water
flowing in the other side of the exchanger. This setup allows some
flexibility in choosing hoses, connectors, filters, etc. that can
match the lower pressure compressor loop. Importantly, if one of the
lines leaks, only a few gallons of water can escape, rather than a
continuous flow from the building (those hoses and connections are
heavy duty).
The heat exchanger is "Flat Plate" brand, model "FP5X12-10 (1-inch
MPT) MMC" and this cools a Leybold coolpak 6200 compressor. I did not
install it, so am unaware of the reasons for the chosen model, however
I could try to find answers to questions you may have.
***
We have cryoprobe on our 700. The configuration is a PolyScience 4200
liquid to liquid heat exchanger running a closed loop to the cryoprobe
compressor. The building cooling water for the heat exchanger is
another loop to a condenser located on the roof. I was told by Bruker
installer do not run building water through the compressor directly
because of water conditioning problems. The building water has glycol
and other stuff in solution and I would only assume that it will kill
a cryoprobe compressor over time. I also can set the heat exchanger
temperature to keep the compressor happy and not depend on the
building cooling line temperature. The building cooling temperature
varies from very cold in winter to very warm in the summer. The
compressor to heat exchanger water lines are braided tygon hose.
****
From my time at XYU I have experience with chiller. Here at ZZC we
use outdoor air-cooled He pumps.
They use a Haskris chillers on a building chilled water closed loop.
In that system, there are two closed water loops. The Haskris loop,
and the building loop. It is crucial to follow the Haskris
recommendation on antifreeze in that loop to prevent corrosion of the
system. I don't remember any rubber hoses. As I recall, it was all
hard piped. Could be wrong there. They do use water filters (two
inline) with gauges measuring the pressure drop across the filters.
That set-up worked fairly well. They did have some problems with the
building water becoming too warm and shutting the system down. Though,
I've had about the same rate of failure here using the outdoor air-
cooled He compressors. We've had a couple bad compressors that
resulted in shutdowns about two or three times a year. However, one
system has been up and running here for two years with no unscheduled
down time.
***
I use campus chilled water via a water-water heat exchanger to ensure
the water through the helium compressor is very clean.
I currently have a Thermo Neslab System III heat exchanger which works
really well. I have a filter on the campus water to remove the worst
of the muck which I change about every two months. I have a bypass
loop for when I'm changing the filter.
Originally I had a Thermo Neslab System I, which didn't really have
enough capacity. It needed maximum flow on the campus water, which
meant I was changing the filter weekly. And at top performance the
loop temperature was about 20C. With the system III, I run the loop
temperature at 15C, but it will chill to about 10C.
The most important part is the water in the closed loop. I use
purified drinking water from the supermarket.
<<JK: Is your W-W heat exchanger a simple thermal contact device, or
does it have a refrigerator component like some Haskris units do?:>>>
It is a simple thermal contact unit. It is the system III described
here: http://www.thermo.com/eThermo/CMA/PDFs/Various/File_26289.pdf
***
Our N00 ColdProbe uses a chiller with a closed-loop of Distilled/DI
water to
go through the helium compressor, and it uses building chilled-water
(through a heat exchanger in the Haskris) to cool the internal closed-
loop.
Our other Cold Probes (on the X00 & Y00) each have a self-contained
Haskris,
that has its own refrigeration unit to chill the DD/DI water in the
cooling
loop (using a standard refrigeration system, with a compressor and
condensing unit outside).
We are building a new Biotech research building, which will have
adequate
house-chilled-water, but we will be purchasing another Haskris that
will use
the house-chilled-water to cool the DD/DI water for the CryoBay. We
do add
~5% Propylene Glycol to our DD/DI water as an anti-freeze and anti-
microbial
agent.
I would never put house chilled water through the Cryo-Platform
because I
don't trust the Facilities people to have it be clean (no matter what
they
tell me).
***
We have had three cryoprobes since about 2003, and we've had many
water issues. We also use building chilled water to carry away the
heat, but we don't pipe it directly into the compressors. We use
Neslab water-to-water heat exchangers, which run a closed loop of
water through the compressors, and exchange heat with the building
chilled water. We aren't allowed to directly use the building chilled
water in any equipment, and also some parameter of our building water
(pressure or flow rate) is outside the compressor specs.
We've had terrible problems with our cryocooling units. The coldheads
would start ratcheting after a couple of years of service, and no one
could figure out why. The Bruker people would come and decontaminate
the whole system, and then it would start ratcheting again three weeks
later. They replaced many coldheads and made countless service visits
for about three years. Finally they figured out that the helium inside
the system was getting contaminated with hydrogen, which they said
could only have come from the capsule oil breaking down (that is - the
capsule in the compressor).
The connection with the cooling water is that it was too hot, making
the capsule oil break down quickly. When we set the system up, the
water temp was within spec, but not by much (probably ~26 deg C). But
our building chilled water got warmer, and intermittently got very
warm. The water temp into the compressor would sometimes spike to 35C
or higher (the upper limit of the compressor spec is 28C). We didn't
pay close enough attention to it, and it seems to have been the source
of all of our problems. Since then we have induced the department to
upgrade the building heat exchanger and regulate the building chilled
water temp. It's now rock solid at 10 deg C.
The Neslab System I heat exchangers that I have are not really good
enough for the purpose of cooling the helium compressors, and if I
could buy new ones, I would get larger ones - probably Haskris. But my
bosses don't want to spend the money. But I have done something that
has helped a lot: I bought external heat exchangers and installed them
in the lines between the Neslabs and the compressors. They are just
metal boxes with connectors that exchange heat between two separated
streams of water. By putting an external heat exchanger in line, I was
able to lower the water temp going into the compressor by four or five
degrees C. I also had the department put in larger diameter water
lines into the room which doubled the flow of the building chilled
water. So now my water temp into the compressors is around 20C, and
since I've done all this I haven't had any problems (it's been 15
months now).
So my advice is to get something that can keep the water temp well
within the specs. I think Bruker now recommends that the water into
the compressor should be 15C. I wouldn't get the Neslab System I. Its
water pump isn't strong enough, and its heat exchanger isn't big
enough. I don't have any experience with actual chillers, so I can't
give any advice about them. But I have seen some models that are water-
to-water heat exchanger and also chillers.
....
The external heat exchangers are like these:
http://www.houseneeds.com/shop/HeatingProducts/HydronicHeating/heatexchangers/mischeatexchangers/heatexchangerbuyage.asp
They look like little metal boxes with four house connectors on them.
I have one of these plumbed between each Neslab and its compressor. It
brings that water stream in thermal contact with a water stream from
the building chilled water. So for each Neslab, I need two connections
from the building chilled water: one that goes into the Neslab itself,
and one that goes into the external heat exchanger. It doesn't
increase the flow or the pressure through the compressor, but it's
like having a Neslab with an extra large heat exchanger. It really
helps - I'd say that it lowers the temp of the water loop from the
Neslab to the compressor by about 4 deg C (versus the same setup
without the external heat exchanger).
***
We have water-cooled Haskris units cooling the helium compressors of
both of our Varian cold probes; these Haskris units are in turn cooled
by our (large) building's secondary chilled warm loop (this loop does
heat exchange will a primary cooling loop that flows from the
University's Central Plant). I think they (our facilities people)
take pretty good care of the building's secondary chilled water loop;
we've never had a problem w/ "crud" clogging the Haskris (we're
responsible for keeping the Haskris loop clean; 15% ethylene glycol
seems to be doing the trick). The only drawback is that, even though
our Haskris (and every other component of the consoles and cold
probes) is on UPS and/or emergency power, the pump for the building's
2nd-ary chilled water loop is too big to do this, so when it loses
power, chilled water stops flowing, Haskris heats up, He compressor
heats up, cold probe warms up. If your local chilled water loop is
small enough to put on emergency power, maybe you could find someone
to trade your air-cooled Haskris for a water-cooled unit, then use the
water-cooled Haskris as a buffer against your variable pressure,
variable solids-content local chilled water loop? (The Haskris might
be more forgiving of you chilled water loop's pressure fluctuations
than the helium compressor of your cold probe; it certainly wouldn't
throw chunks into your helium compressor.)
<< JK: To be sure, which kind of Haskris W-W units do you have? They
have both "simple" types and ones with refrigerators associated with
them. I'd guess you have the simple one, but I'd like to be more
certain.>>
Ours have refrigeration units; they cost about $15k (guess I should have
mentioned that, this could make it prohibitively expensive for you). On
our 600 we have an R300 model (serial HB15729) w/ options C and R4; I
think it's a bit of a special design since our building chilled water
loop
isn't very cold (in a moist climate like Houston, you have to keep
condensation to a minimum or you have major mold problems). Doug Wagner
(doug_at_haskris.com, 847-956-6420, x243) at Haskris may have the details.
We also have a dual pump system on our 800 (in case we ever stumbled
across the money to put another cold probe on our 500). I can't seem to
find the manual for it right now; maybe you could ask Doug (we got the
R300 for our 600 in late 2004, and the 2nd Haskris for our 800 in late
2005, if that helps Doug find it). Hope this helps.
***
Our facilities guys spent a whole bunch of time working on the system
in December- and it worked beautifully for about 6 months. Then it
became summer, the water started being super cold, so I decided to
live dangerously and asked them to set it back to 10C like it was all
winter. Which of course set off a whole new round of adjustments and
failures.
There is a fancy system of valves that mix a chilled stream and a
stream that bypasses the chiller to keep the temp constant- I believe
managed by some kind of Siemans online program, which is my opinion is
the source of all my troubles over the past few years. What the
program says apparently often does not correlate with what is really
happening- whether it is because the manufacturer is the problem, or
the installation I don't know.
If your facilities people are planning to use the Seimans system, make
sure they have a good template of a similar system that works, is my
advice. Pressure seems to be the only thing that constantly works on
mine.
****
****
ORIGINAL POST:
We've got some problems with chilled water in the fabulous new
building into which we've moved our AvanceIII system with a cryoprobe,
and I'd like to hear how you all with cryoprobes are configuring your
cooling water.
We had been set up to run on a small chilled water loop that runs
through part of our building, but this is giving us two problems. 1)
The pressure in the line sometimes fluctuates, and 2) some junk has
been clogging up the lines inside the cryoprobe. We're temporarily off
that system, on a closed-loop Haskris chiller system that exchanges
the cryoprobe's heat with the room's air, wafting a hot scirocco
breeze at the spectrometer - clearly not a permanent solution.
On point 2: It may very well be that some of the anticorrosion/
antibacterial stuff in the building's lines is reacting with our black
rubber hoses, causing them to degrade and plug up our lines. Any of
you have this experience? Do you have any specific recommendations for
types of rubber hose?
Our building guys have installed a pair of filters in the line, and
now they're talking about a system with an additional pump and some
temperature-controlled valve and recirculating the heated water back
into the cryprobe because that'll somehow make the pressures work out,
and that just seems too complicated.
How do you do this? If you use house chilled water, please tell me if
it's on a local or a main loop, and whether you have any extra filters
or pumps in the line.
Josh Kurutz, Ph.D.
Instructor and Senior Scientist for NMR
IMSERC, Chemistry Department
Northwestern University
2145 Sheridan Rd.
Evanston, IL 60208-3113
847-467-1681
fax: 208-978-2599
Facility: www.chem.northwestern.edu/imserc
NMR Blog: www.imserc-nmr.org
Twitter feed: twitter.com/imserc_nmr
Other: www.joshkurutz.com
Received on Thu Jul 08 2010 - 13:29:54 MST