Lots of opinions on this one. Most people don't think I need one with
the large room and only a 400. I agree, but my safety people are
leaning the other way, and I may not get a choice. The solid state IR
Oxigraf O2 sensor advice is greatly appreciated, as I didn't know
about this.
---------
We're considering installing an emergency exhaust system for a 400.
Any thoughts or recommendations? Thanks. Will summarize.
Robert V Honeychuck
George Mason U
Fairfax VA
rhoneych_at_gmu.edu
---------
Safety manual:
http://ccc.chem.pitt.edu/wipf/Former%20Group%20Members/Safety/NMR%20Safety.pdf
---------
Do you have a tall ceiling? Is the room reasonably large? Is this
a legal requirement? Do the people doing it know what they are
doing? I have seen an unshielded 500 quench in a small room - we
left. It was not a big deal. In the larger room where our 400
lived for many years we never had a vent and really did not think one
was necessary. In our new basement instrument space with 3 NMRs and
a 16' ceiling we did have to put one in because it was required by
code. It is built into the air handler. It was VERY expensive to
do. We also have one in our 800 building. And we have O2 sensors as
well.
Like the person earlier with the wet air I am sure you will get lots
of responses. People have a lot to say about O2 sensors but we have
to grin and bear the expense of calibrating them and replacing the
sensors. They are located so that they will not set off an alarm
during fills. In fact the new room is so big that they did not go off
when the 500 quenched. I did push the panic button for a full air
exchange. Our facilities director was present and it made him feel
better about the drama.
---------
My neighbors in the 9T small bore MRI lab had to delay installation of
their system more than a year because it was going to cost over
$100,000 to build a helium quench pipe. Truly outrageous for what
basically is a 20' stainless chimney requiring minor excavation
to install.
It sucked because the instrumentation grant wouldn't cover a
"facilities" expense, and the Univ. didn't want to pay for a fancy-
pants pipe.
I wouldn't think a 400 would hold enough helium to make such an
investment worth anyone's while... Unless you were building a brand
new facility anyway, in which case a quench pipe could be a cheap add-
on.
I hope they give you a raise instead of a pipe.
---------
Hi Robert,
there were some previous somewhat related posts
February-2006/34.html
December-2005/7.html
May-99/0022.html
---------
I have had labs with and without. For a 400, I doubt it is really
necessary. But if you do it, and have the fan automatically triggered
by
an oxygen sensor, the system can become the bane of your existence if
you
choose a crappy sensor. Our architect chose a sealed type chemical
sensor, all chemical sensors fail after 1-2 years, and they all fail
reading a low oxygen level. That can turn your fan on permanently. We
weren't very happy with our system until we trashed the original oxygen
sensors and replaced them with solid state Oxigraf sensors, which work
through an infrared sensor. They can be set up for regular
autocalibration against known good air (like outside the lab) and they
are
just great, and last for many years. We also engineered in both manual
fan on and manual fan off switches.
---------
Can you tell me who the magnet manufacturer is, and whether you wish to
exhaust both helium and nitrogen, or just the helium?
The magnet manufacturer is relevant for the amount of gas being
exhausted and the quench pressures involved for safety reasons and the
Pressure Equipment Directive. This information is needed if you decide
to go the quench ducting route. If you wish to go for an indirect
exhaust system (which is usually more costly), then the room dimensions
and the volumes of gases in the magnet are required in order to
calculate adequate exhaust rates to keep the oxygen level in the room at
a safe level.
---------
I imagine you are thinking of connecting the quench valve ports to an
exhaust pipe to route the helium gas outdoors? I how no experience of
this although I have a colleague at the University of Santiago de
Compostela in NW Spain who has done this. It makes most sense when
several magnets share a lab. I vent my regular boil off helium out of
the lab but indoors since pressure differences between the lab and out-
doors due to wind can cause fluctuations in the boil-off rate and
possibly even disturbances in the magnet dewar. The idea is to avoid
getting helium gas loose in the lab where it can penetrate the
magnet's vacuum seals. This would be worse in the case of a quench.
---------
We were forced to install one for a 600 WB for safety reasons.
I'd recommend solid state O2 sensors to activate the system.
The chemical sensors need frequent calibration to avoid false
positives. Ours have dropped about 0.1% in three years with
no maintenance. I'm surprised a 400 would be dangerous.
It must be in a small room.
---------
Hi Robert,
A 400 is not very big to need an exhaust, unless it is in a small
space. I
would do the volume calculations to see how much space will fill with
helium
in a quench. You might find that only half of the space will fill up,
which
would not need an exhaust, IMO.
The one concern for an exhaust which avoids a lurking problem, is to
be sure
the exhaust is separated from the normal ventilation/heating/cooling
ductwork. This is not only to avoid pumping the helium to other people.
The lurking problem occurs when a quench chills the ductwork, which in
turn
condenses water, freezes or floods elsewhere along the duct.
Another concern would be to avoid rain, dirt, or birds coming into the
duct
from the outside.
Also, be sure to isolate the magnet from duct vibration with a flexible
section.
---------
I have such a thing for my room housing a 400MHz and a 700MHz
spectrometer. While it is the recommended thing to do, there are quite
some pit falls to be aware of and deal with. Here are some aspects I
learned along the way explained on my system:
My exhaust system is automatically switched by the oxygen sensors. I
have two oxygen sensors, one low, one high. The trouble is, the life
span of these oxygen sensors is around 1 year. So I sat in a wind
tunnel when one sensor went bad, until the new sensor was purchased,
delivered and installed. Now we have it set such that both sensors
have to fail. Also, I finally got an overwrite mechanism which
supposedly lets me turn off the fans, when I know that none of my
magnets quenched (manual oxygen sensors additionally?)!
Also, when we have a power outage, the fans automatically open (i.e.
they need power to stay shut). So I started to accumulate all kinds of
leafs and other dirt under the opened but not running fans. Also, we
had construction of metal towers outside on the other side of the
building. Still, my magnet was a good vacuum cleaner for steal
particles. By the direction of the accumulation of the iron filings on
the magnet, I could tell that it had come through the vent. I finally
got a filter put on the inside of the fans to avoid some of it.
I also finally got a panic button, where I can actually turn on the
fans. When I had a smoke incident from smoke coming in through cracks
in the wall, I ended up blowing nitrogen gas from a liquid nitrogen
tank onto the sensor to turn on the fans (a real emergency response!)
Finally, my system is exclusively monitored by facility management
somewhere on Campus. Except for the features described above I have no
control over it. The only good part about it is: They have to purchase
the oxygen sensor replacement parts.
Make your own decision, whether it is worth it. Gratefully, I had not
yet had an incident here, where I needed it. I my previous positions,
I have been around quenching old magnets (during charging and one just
spontaneous over night), but it was very obvious and we got out. I was
told that the more dangerous situation actually occurs during nitrogen
fills.
---------
How big is the room? What type of magnet? If you are in a closet
with a shielded magnet, a simple hood with a temperature switch or
oxygen sensor, might be enough to exhaust a quench. These magnets hold
about 80 L of liquid helium full X 740 Lgas/Lliquid. That would fill
a small room quickly, but a hood should move that gas out at about 80
CFM. If you have a unshielded, long hold 400, you have about 150L of
liquid, when full (111000L of gas). An exhaust hood, should also work
in this case as well.
I have a unshielded, long hold 400. My room is roughly
40'X20'X16' (roughly 27L/cubic ft). I ask students to evacuate during
a quench. The building that I'm in turns over the total air in the
building 10 times/ hour - so I don't have to worry about the problem.
During training, I tell each student, "If there is a sound like a
freight training or jet, get out of the room for a couple of minutes.
Essentially anything loud and noisy above normal sounds is not good!"
---------
We also have an emergency exhaust for our 800MHz. Some things to
consider are;
1. Make up air. When the fan is on it will draw the He gas from the
room, but depending on the size of the fan there must be a supply of
fresh makeup air that can be drawn into the room. If makeup air is
not provided then air will be drawn from the most available sources
which could mean the introduction of dirt, dust and other debris to
the magnet room.
2. Triggering mechanisms. Our emergency exhaust fan can be triggered
3 different ways. The first way is via a panic button that a user can
push in the host computer room. The second is an oxygen sensor which
will activate the fan if the oxygen level is reduced during a quench.
The third is a temperature sensor located near the emergency exhaust
port. Cold helium exhaust will trigger this sensor during a quench.
These sensors allow for emergency exhausting whether an individual is
present or not. Also, an alarm is sounded and we are notified by
phone both by a monitoring computer and the University Central Control
to which all our alarms are sent.
3. Regular testing. Whether your system is manually or automatically
triggered. The mechanisms of activating the fan should be tested
every 6-8 months a) so everyone knows how to do it, and b) everyone
experiences the sound and affect of the fan being on. We don't want
someone going into shock when that big fan starts howling.
These items may be over the top for a 400MHz installation. I believe
makeup air is still a concern, but if your environment is relatively
clean then there should be no problem. Also, less expensive
individual portable oxygen sensors placed near the magnet will detect
a quench or significant leak. These sensors have a good loud alarm.
Be aware that oxygen sensors need regular replacement which is an
added expense.
---------
Hi Robert,
The questions to ask are:
How big is the room compared with the volume of gas that will be
produced in a quench (LHe only) or if the magnet falls over (LHe and
LN2) and what would be the resulting oxygen content? The expansion
ratio is between 700 and 800 to 1.
What is the air turn over time in the room?
If there is a quench and the pressure in the room goes up, will it
prevent doors from being opened?
---------
I'm sure that you'll get many responses like this. I think it would be
rather unusual to consider an exhaust system for such a small magnet.
Unless you have very poor ventilation in your lab.
I have only seen exhaust systems installed for large magnets, 800MHz
upwards, or perhaps 600MHz upwards for wide bore magnets.
Particularly if you have a shielded 400, the stray field outside the
magnet can is so small that a quench is very unlikely to be triggered
by moving equipment in the lab.
If people are present and the magnet quenches, it is such a noisy
process that they should have time to exit, unless there is something
very unusual about your lab. The dangerous scenario is if the magnet
quenches, replaces all the air in the lab with N2/He, and then warms
up while no-one is around. Normally ventilation systems will refresh
the air in that time, but if there is very bad ventilation and the lab
is very small, there could be danger of asphyxiation.
---------
Received on Mon Feb 23 2009 - 04:28:54 MST