AMMRL Community,
Here is a belated summary of my Nov. 2 query regarding the shutdown of
our lab's HVAC system over two holiday weekends. Please recall that
the chief concern was how the instruments would be affected by a
drastic reduction in lab temperature - perhaps below freezing.
In the end, we took no special precautions, our lab temperature
dropped only ~ 10 °C, and we were operational the day after power and
room temperature control came back online. We simply gave the
instruments ~24 hours to equilibrate before shimming them and resuming
experiments.
As seen in the compendium of responses below, there were few major
concerns about instrument health in this situation. The consensus
seems to hold that the major worry is the potential for bursting
pipes, causing flooding.
Perhaps the most convincing evidence that NMR systems can survive cold
temperatures came from two people who told of their experiences during
the Ice Storm of 1998, which shut down much of the northeastern U.S.
and southeastern Canada for many days, but did not cause any direct
damage to the equipment.
Several respondents shared three points:
First, it was noted that magnet seals might be leak if the temperature
changed too much or too rapidly. This concern applies primarily to
older magnets with seals that have not been replaced in a long time,
say more than 10 years. One person noted that many magnets are now
shipped under vacuum, and must survive the temperature fluctuations of
a freight compartment for many hours, so a few hours without HVAC
shouldn't pose a significant threat.
Second, it was recognized that the console, probe, magnet leg, and
other hardware should be unaffected, especially if they are turned off.
Third, it was pointed out that there is potential for condensation
inside the instrument and inside the computer (notably the hard drive)
if the room temperature and humidity rise faster than the temperature
of the equipment.
On the whole, it seems there's nothing much to worry about, except for
bursting pipes. One simply needs to take care to ease the systems back
up once power and heat are restored, and recognize that shims are
going to fluctuate while the room temperature is changing.
Thanks to the many people in this fabulous community who responded,
and thanks to you for reading.
- Josh
RESPONSES:
****
I think your biggest worry might be the magnet "O" rings. Remember
the Challenger? I wouldn't think that storing the electronics at that
temp. would be a problem.
****
Considering that the systems often ship in trucks that aren't heated or
cooled, it's not hard to imagine NMR spectrometers experience
temperature extremes when not in use. The only concerns would be the
equipment in your lab that uses any water, say, for cooling. If you have
a water-cooled Helium compressor on a CryoProbe, then you could, in
theory, throw in some antifreeze (most compressor manufacturers tolerate
some Ethylene Glycol). Magnets certainly don't care about it getting
cold!
****
Don't let the lab temp go below 37 F. The o-rings in the magnets are
the
same type used on the space shuttle, you know ...
****
There should be no problems with any electronics getting cold. Most
all electronic assemblies get air shipped at one point, so they
withstand high altitude and freezing temps just by shipping them from
the factory.
Even whole magnets are air-shipped under vacuum now. Their seals
don't fail in the freezing cargo hold.
Mostly you might worry about the magnet seals, but there's not much
you can do about it. Seals should have no problem maintaining
elasticity below freezing, unless they are very old. Old seals
eventually get brittle and crack, but if that is the case, you need to
replace them anyway and a quench would do you a favor. I would just be
sure that there is adequate LHe in the magnets to survive a quench.
There is a higher specified minimum level for when the E-stick is in
the magnet, and I would have the magnets full to at least that level.
****
We've never gotten that cold (0 Celsius), but occasionally our plenum
air supply gets very cold if outside temperatures really drop, and we
have found the lab at ~10 Celsius early in the mornings and it's
probably gotten a degree or two lower without us being aware overnight.
This level of cooling hasn't caused any problems with the magnets of
which we have a mixture of new Varian/Agilent/Magnex magnets and 15-20
year old Oxford magnets. Our spectrometers probably kept themselves
near optimal anyway, albeit sucking in cold air!
Not much help I guess, but figured it might give you a little feel for
it.
(My gut-feeling would be to get some gas heaters in........ although
then you just fill the room with CO2!)
****
Our NMR lab is located in a pole barn in North Eastern Bucks County
Pennsylvania. Our electrical supply is provided by a local company. We
have unexpected power outages throughout the year and do not have a
Generac
(or other) on-site power supply.
For the cold winter months, we have portable KeroSun kerosene fueled and
portable propane fueled heaters to keep the lab from freezing. Perhaps
something similar can be rented in your area during the two upcoming
holiday
weekends.
****
Your lab can't go much under 32 F or the water pipes will burst. The
magnets should be fine. O-Rings won't become rigid at these
temperatures. The shims will need to warm up for at least an hour
before I would attempt to shim. I've been through air outages and
steam pipe shut down, but never for this long. Brick buildings hold
heat for a long time. Your probably won't see a drop in room
temperature passed 50 F. Burst pipes would be your worst case scenario.
****
Do they not sell kerosene space heaters in Evanston? No power needed.
****
The main problem you may have is moisture. So try to cover/close
anything that is susceptible to it and, after the power and heat are
up, give the equipment 24 hours to stabilize. The magnets should be
fine.
****
The shimming may be a concern, but give it a day i.e. let it come back
to equilibrium after the heat is turned back on before you look at it.
My biggest worry in this case would be: Will the water pipes freeze
and cause havoc? But I'm guessing they've considered this as this is a
problem all labs would have, not just an NMR lab. Do you have water
cooled gradients?
****
Sorry, no firsthand experience. But I would rather be concerned about
condensing humidity, than about the magnets. I just would make sure
that, after the room temperature is back, all cabinets are open for a
while and dry before switching them on again. The seals of the magnets
are designed to hold much lower temperatures than just freezing.
Humidity, on the other hand, can kill all kinds of electronics.
****
We had a similar experience about a year ago. My advice would be "do
it slowly". If there has been a significant change in magnet
temperature over a period of some days it is likely to take a while
for the field to get back to being completely stable, so there is no
great rush. But it should come back to pretty much the same place
within a day or so of the room getting back to normal operating
temperature. Get the air con / room heating back on >>as soon as
possible<< and allow some time for the room to warm up. I don't know
what the humidity it like in Chicago in the winter, but one of the
things I worry about here if something gets really cold is
condensation. Surprising the amount of damage a small amount of
condensation can do to electronics - and it's hard to find/prove
afterwards. So let the room warm up / dry out for a couple of hours
first. Then when you do turn consoles back on, check that all the
cooling fans come back on. Old fans really do not like being switched
off and allowed to get cold. And old consoles really do not like
working without cooling fans!
If all seems OK, reload your last good set of shims. Then allow an
hour or so for things to stabilize (how long will depend on how cold
things got). If you can get a reasonable lineshape then you can start
doing short, less demanding experiments but I wouldn't look to start a
long 2D until at least the next day - and in my opinion it isn't worth
sitting in front of a spectrometer chasing shims that are drifting -
may as well wait until things settle and then do the job once.
****
We had the power go off (to much of the city) during the "Ice Storm"
of January 1998. I forget how long we had no electricity on the
campus, probably for a week. The heat was off long enough for it to
get cold enough in one lab for a water pipe near the outside wall to
burst. (It gets cold here in Montreal in winter too.) After the
electricity was restored and the water was cleaned up all the
equipment was powered back up and there were no problems. I wasn't in
charge of the facility at the time but I don't recall there being any
issues regarding the magnet either.
****
I don't know how the cold will affect the magnets. However, if you
needed to keep the temp elevated, at least above freezing, you could
always try heating with space heaters powered by deep cycle
batteries. Or, your Facilities Dept might be able to set up a backup
generator for you to do the same thing.
****
During the January 1998 ice storm in Quebec, we didn't have electricy
for
weeks. The temperature of the lab was fairly cold, I was mostly
concerned
about water condensation, but likely there was none noticeable. I do not
remember any problem when we restarted the systems. Be careful about the
hard drives but I did not wait until everything reached RT.
****
with temperatures only near freezing nothing bad should happen. If it
gets a lot colder, I would begin to worry.
It should take about an hour or two for the system to come back to
normal after the power and heat are back on. This is to warm the shims
back to their operating temperature. Once everything is back to normal
also the standard shimvalues should apply again.
****
I do not have experience of such extremely low temperatures, but I would
beware of condensation problems within the electronics when the building
is re-warmed. You should take all steps to ensure that the ambient air
is as dry as possible at that point and I would leave the instruments
for a few days after normal temperatures are re-established before
switching the units back on.
****
About the possible need for reshimming, we recently used an updated
3D-shimming program (gxyzshim) from Varian. That has taken a lot of
pain
out of the process and we did get pretty satisfactory results in a short
time.
****
I don't think the change in temperature will be any issue - the magnet
itself is in liquid He anyway so going below freezing point at the
outside does not matter at all. I also do not expect any significant
change in shims. Only advice for a shutdown of several days:
- remove any sample from the system (especially D2O samples)
- shut down console / VT / PC / (warm-up of cryo probe if applicable
and shut down cryo unit)
- put the lid on top of the magnet (shim barrel/stack) to prevent any
(magnetic) dust from getting into the bore of the magnet
For starting up the console, I would wait until the room temperature
is back in the right range.
Josh Kurutz, Ph.D.
Instructor and Senior Scientist for NMR
IMSERC, Chemistry Department
Northwestern University
2145 Sheridan Rd.
Evanston, IL 60208-3113
847-467-1949
www.chem.northwestern.edu/imserc
NMR Blog: www.imserc-nmr.org
Other: www.joshkurutz.com
Received on Wed Jan 12 2011 - 03:39:08 MST