Dear all:
Sorry it's a long email. You might want to delete it if you are not
interested in the topic.
Many thanks to all who replied including Todd Rappe, David VanderVelde, Sara
Kunz, Jerry Hirschinger, Marcus Wright, Roger Kautz, Wei G Wycoff, Vitaliy
Gorbatyuk, Patrick Tishmack, Andrew Fowler, Martha Morton, Eric Paulson, Ian
Vorster, Richard Shoemaker, Philip Dennison, and Torbjörn Astlind.
Many gave detailed information and some with attached diagram and
useful links. Below please find a very brief summary (I hope I
understand them correctly.) But I attached all the replies at the end
of this email. Sorry that it's very long. Many mentioned a setup
similar to the backup air supply design made by Richard Shoemaker. One
needs two one-way check(isolation) valves and N2 gas cylinder with
pressure regulator set about ~5psi below the house air pressure.
If N2 gas cylinder is used only in the line that goes to the cold probe, it
can keep the probe from warming up for a long time. It is the magnet
legs and sample eject that consume a lot of air. A high pressure liquid
nitrogen boiloff tank can replace the N2 gas cylinder if all the air
lines need backup.
We may use N2 gas cylinder to only backup the cold probe from warming up.
I don't expect the spectrometer being used during this short time period.
Fortunately, the air outage is scheduled and we are notified ahead of time.
In the case of prolonged air failure, which is very unlikely, I guess
I just warm up the probe anyway.
We don't plan to add an additional compressor. One 'silent' model of
air compressor from Werther Inc was mentioned (see link below) if
anyone is interested.
Thanks again!
Dejian
-----------original replies-----------------
1
I got the attached document from Rich Shoemaker. The system provides enough
backup air to keep the cryoprobe from going into warmup for short
outages/pressure drops. Here is more info on the Panther "silent"
compressors they use:
http://www.werther.com/product/Panther-Silent-Compressors-P5024-AL/ProductDe
tails.html?search=1&ProductID=4&Category=4&HCatid=4
2
I did that. You also want, in my opinion, a couple of check valves. One
should be put facing backwards toward the compressed air line such that
your nitrogen tank will not try to repressurize your entire building's
compressed air system when the compressor goes down. The other I put in
going forwards from the tank such that the valve was closed as long as the
air is on. This allows you to change out the nitrogen tank without having
air blowing back in your face when you disconnect the regulator. (A
manual valve which you leave open except when changing the tank would do
the same thing).
The tank might last an hour give or take, the time being very dependent on
the use of sample eject during the outage. I had a compressed air dryer
in the lab, but made sure to plumb in the backup after the dryer. The
nitrogen itself was dry, of course, but the dryer vents a lot of air as
part of its normal operation, to carry away the moisture being removed
from the downstream air.
For planned outages for maintenance, we would normally get a high pressure
liquid nitrogen tank (a "gas pack") which is enough to run the NMR for a
few days. I put this in through the same plumbing as the cylinder. Use
of a tank like that in the NMR does mean a lot of frost will build up on
the outside of the tank, and when it melts, a lot of water on the floor.
3
Perhaps there is an upside to locating spectrometers in old buildings. We
have always had our own compressors running in pairs (they alternate and
can be maintained separately). This makes us independent of building
air. We have had our own dryer so that we can control the dewpoint as
well.... Our newest spectrometer is in a building with a great air system
which has multiple compressors (and a low dewpoint) so they can maintain
the compressed air system without bringing research to a halt.
Perhaps a Nitrogen gaspack would be a better choice of Nitrogen source.
They last longer.
4
For the short duration outages you're experiencing, a N2 gas cylinder should
provide adequate backup. You will need to use more than a simple Tee
junction, though. You'll need a pressure regulator on the cylinder and also
should place a one-way isolation valve on each of the cylinder regulated
supply and the house air supply to keep one from feeding back into the other.
Then you can set the cylinder regulated pressure to be ~5 psi below the normal
house supply pressure, and it will feed the system whenever the house pressure
drops more than 5 psi.
5
What's your flow rate at 90psi?
6
We get at least a day out of an N2 gas cylinder, running our inova 500, you
should have no problem for an hour or two. But do you get warning before they
shut off the house air, or do you need it to kick in unattended, an emergency
backup system to avoid a cold probe warmup? If you need a gas supply with
shelf life, I understand a J cylinder would last an hour or two. You could
plumb it in so it would supply just the cold probe, to keep it from turning
off and warming up if the house air dies. (Let the TMC legs sink and the
eject stop working). And maybe sound an alarm to alert you you have an hour
to hook up a gas pack.
7
Our air supply comes from a campus-maintained compressed air source located
in a different building. It supplies air to several buildings. It comes to
chemistry building and then gets divided to high pressure and low pressure
lines. The NMR lab is connected to the high pressure line. There is also a
compressor in the chemistry building which serves as a backup. If the
pressure of the campus line drops below 60psi the chemistry building's
compressor kicks in until campus line is back to normal. The switching is
automatic.
We also installed our own air-dryer in the NMR lab and a bypass line in case
we need to replace the filter and desiccant of the dryer. This reduced the
amount of warm-ups of the cold probe.
8
We have N2 gas cylinder as a backup for the cold probe and it last forever
since the cold probe does not consume the gas too much. However you need to
insert it in the line which goes to the probe only not to the entire
instrument. You may use the cylinder for the entire instrument as well but
then it will last for a day or so because the magnet legs consume a lot. Also
have a check valve in N2 gas line.
9
We T'd in a line for a high pressure liquid nitrogen boiloff tank to back up
our dry air. I does not go through any additional drying, but it does go
through the particle and oil filters before the spectrometer. The tanks do
not last long especially for VT use. I got maybe 6-8 hours from a tank for VT
and possibly 12 hours for normal use. This was for a single 400 MHz liquids
system. With a more sophisticated valve manifold, you can hook tanks up in
parallel and get more time. I would suggest a trial run for your system just
to check out how much time you can get from a tank.
The T and the gas line are not very expensive; the line is just like your
standard liquid nitrogen fill line (braided stainless) with a pipe T or Swage
connection into the air line on one end and a high pressure liquid tank
connection on the other end. You need the necessary valves or removable
connectors to prevent cross contamination of the two gases, but it's not
difficult to do.
I don't recommend an air compressor and dryer in your NMR lab unless you
have a way to isolate the noise and heat. Maybe there are some quieter and
cooler running systems available, but it might be overkill (and expensive) for
your needs.
We now use the boiloff from our bulk nitrogen for VT use and as a back up
when we service the air dryer. Presumably you don't have that available.
10
First, when I started this position, we were running 3 magnets (one cold
probe) off high pressure LN2 tanks only. We've since moved to a dedicated
air compressor + N2 generator, but once we got things optimized we could
hold everything for a bit more than one day/tank. To maximize the time,
you have to make sure the pressure builder is set to a lower pressure than
the safety/relief valve on the tank. You also need a regulator to keep the
output at a reasonable pressure. I would use this option if your outages
tend to be scheduled since it's pretty cheap.
A second option, in my previous position we got a small backup air
compressor. We used a 6 gallon "pancake" style compressor, specifically
see:
http://www.werther.com/product/Panther-Silent-Compressors-P5024-AL/ProductD
etails.html?search=1&ProductID=4&Category=4&HCatid=4
It's VERY quiet (less noise than a normal conversation), can get to about
120 psi, and definitely provides more than enough air for the cold probe,
and probably even to maintain VT airflow (although you'd need a second
regulator). It's probably slightly larger than a typical solvent box (4 x
4L bottles), so pretty easy to place. I saved the old quote, and in 2005
it was around $1000 delivered. We added a check valve so it would only run
when needed and attached it to a "T" so it was always available. The only
downside is that it's not dry air, but you could get a desiccant cartridge
to put inline if you expect to use it infrequently. I was in Boulder at the
time, so the air was dry to start with, and you'll probably be fine for an
hour here and there in any event.
The final option, which you mentioned, would be to have a standard
nitrogen gas cylinder if you're only concerned about the cold probe. Since
the probe only uses the air/nitrogen when it switches the pneumatics, one
tank should last at least several hours and I'd guess more than a day. We
used these to back up our LN2 tanks and would get 30-60 minutes out of
one, but that's running every bit of air for 3 spectrometers - figure
about 2 SCFM and 40+ psi continuously plus occasional spikes.
11
We used a cylinder with a check valve incorporated in two places.
Compressed air -> check valve -> T -> Line to instument
At the T <-check valve <- Cylinder with regulator set to just below normal air
compressor PSI (75 PSI)
This works and you can keep everything running for about 8 hours on the
cylinder. The cylinder should feed only the cryoprobe, VT unit, and maybe body
air. NOT the TMC legs! Eject air is an option that is worth while for longer
term air outages.
12
I have always argued for dedicated air compressors for any instrument
facility. However, if this is not an option I would consider putting a large
air receiver tank (maybe 200 gallons) with a one-way check valve on the inlet
as close to the lab as possible. This should keep you running through short
outages and have the added benefit of making the lab air pressure more stable.
13
With our cold probe we did as you prescribed using N2 gas cylinders however it
would not last if the compressor does not function for a few hours but will
give you enough time for a controlled shut down. Our N2 gas cylinders will
kick in via a one way valve as soon as the pressure drop below the necessary
pressure obtained from the compressor. It will only supply enough pressure for
the cold probe alone and not for the rest of the system. This way the cylinder
will last a bit longer.
On top of this we have a small dedicated compressor driving the air supply for
the NMR system. I also have a one way valve from the main lab air supply which
will kick in at pressure below 80 PSI if the dedicated compressor fails, which
has happened during a weekend. Thus 2 compressors have to fail before the N2
gas are used. (A bit of an overkill I guess) My thought was the costs involved
comparing to damage to the cold probe.
14
I have attached a document that shows the compressed-air backup system I
designed back in 2004 when we had issues with loss of compressed air causing
our Cold Probe to warm-up on our 600MHz ColdProbe system.
As you can see, it maintains pressure in the pneumatic valves in the cryogenic
system, even if the instrument air pressure drops below 80 PSI. This setup
has been duplicated since on our 800 and on our 900MHz NMR systems, and has
been very successful for us,
15
Could you just install a large pressure tank at the input to your lab, with a
non-return valve on the input? That would protect you from pressure
fluctuations and give some reserve for the cold probe.
16 Torbjörn Astlind
I take it from using "Cold probe" that it is a Varian system. We have serveral
Bruker systems and the two Cryoprobes has its normal gas supply from
compressed air from the building.
Additional to that there is a supply of gas from, in one case, the building's
N2 gas supply, and in the other case a N2 gas cylinder. The changeover is
controlled by an extra board in the console. The extra supply will provide for
VT gas, and gas for emergency eject in case the probe warms up, in case the
normal supply is absent.
The VT flow, and eject air might add up to 1000 -2000 lit/hour, so our
standard gas cylinder will last 10 to 20 hours.
I do not think, for our operating situation, an extra compressor would improve
anything.
The emergency gas supply is well worth having, not only to warrant
uninterrupted operation, but also to prevent the sample from freezing, and
possibly brake, if VT air is suddenly gone.
I would guess there has to be a similar option for Varian/Agilent systems.
---------------
Received on Tue May 01 2012 - 11:15:13 MST