Hi,
Thanks to all for your feedback. Here are the replies - thanks to
all. The risk of asphyxiation is well documented and very real - we had
good advice and help from our cryogen supplier when we did this risk
assessment and wrote Standard Procedures. There's also a European Pressure
Equipment Directive (97/27/EC), and Oxford Instruments have a policy
statement here: http://www.oxinst.com/SCNNWP524.htm.
Cheers,
Mike
Original message:
Our safety people have started a discussion on the safety of NMR
cryomagnets. The emphasis is on these as pressurised vessels in the case of
a quench. The suppliers can provide information on cryogen volumes,
bursting disk performances, etc., but I wondered if you know of anyone
writing on the subject. Or perhaps you have been through this exercise and
could comment on the the most important issues.
Replies:
I can't help you with references, but the most important issue in the
discussions I've had with safety professionals is dilution of the room air
during a quench. Depending on the size of the room and the cryostat, oxygen
levels may be low enough to be a hazard. Usually the calculation is based on
the immediate vaporization of all the cryogens in all of the magnets in a
room when they are full. I've been fortunate that my labs have been large
enough that the hazard was minimal, but in a small room with a high-field
magnet it can be a major issue. Cures include oxygen-level meters/alarms,
quench ducting, and/or special ventilation arrangements (full-time or during
a quench). Usually a SOP is required for cryogen fills that addresses
ventilation issues. All of my experience is with USA regulations, but I
assume the issues will be similar for you
=======================================
Liquid nitrogen (b.p. 1960 C) contains some liquid oxygen (b.p. - 1830 C) as
an impurity and therefore evaporation leads to an increasing proportion of
liquid oxygen, so that before complete evaporation occurs the residual
liquid may contain up to 80 per cent of liquid oxygen. Contact of this
residue with organic or combustible material of almost any sort is likely to
cause an explosion. If Dewar flasks containing liquid nitrogen have been
used us cooling baths, great care must be taken to ensure that all liquid
nitrogen and oxygen has evaporated completely before the Dewar flask is used
for another purpose.
======================================
I worry mostly about the vacuum part of the dewar. You never know how
much air has gotten sucked in there over the years and when the magnet
quenches the helium goes away. When that air melts it has to go somewhere.
I've had our magnets checked by reactor engineers certified in pressure
vessel design. If you have access to such people I would suggest having
them take a look at your magnets. Make sure they know about the potential
air leakage into the vacuum and the cryo-trapping that occurs. I know some
old magnets have no pressure relief on the vacuum space. These are clearly
not safe unless all of the flange bolts are removed so that the cryostat's
vacuum space can "burp" if it needs to.
==========================================
Generally, a magnet is designed to vent all quench gases with less than
~5psi backpressure buildup, and that vent capacity is redundant in the main
and auxiliary vents. In other words, the quench valve(s) alone should vent
the total volume, and the burst disk backup can also take the full volume.
This is to prevent ANY damage to the magnet, and therefore there is no
danger of explosion. The only exception to this is if ALL of the cryogen
access ports are closed with ice, so the gas cannot reach the protection
valves. Ice builds very slowly if at all, and can be avoided easily through
good fill practices and occasional inspections. Inspection can be performed
by placing a clear cover over the port and looking into the port with a
flashlight. A pressure gage face works well as a clear cover. Inspection
is only necessary in extreme cases. The constant presence of boiloff gas
flow is also an indication that there is no ice blockage, without
inspection. If there is not a boiloff gage to demonstrate the boiloff flow,
then you can put the end of the tube into a cup of water to see the bubbles
which indicate flow.
=========================================
The main concern with a quench should be asphyxiation, not explosion. The
volume expansion ratio
of LHe, combined with increasing cryogen volumes to get longer hold times,
makes asphyxiation even more important today than in the past.
There is just as much overpressure protection on the magnet dewar as there
is on the transported supply dewars. The supply dewars, however are MUCH
more dangerous for explosion than a magnet. If they persist pestering you
about the magnets, ask Safety Sam to investigate the possibility of the
supply dewars exploding, considering the damaged condition of most of them.
=============================================
Our safety people have not commented on this, however, I do remember
running into someone at an ENC who said that their safety
people happened to walk by while she was filling nitrogen and threw a fit
about how much nitrogen gas was going into the room, potentially
displacing the oxygen.
They made her immediately cease all filling operations until further notice
until they could meet to discuss how to deal with this (not caring that the
magnets would quench without proper maintenance). She finally got around
them by saying that she would install an oxygen monitor
on the wall. I think she said that the monitor never went off during a
filling operation, which was enough to satisfy the safety people.
During a quench, it is possible to displace the oxygen in the room
depending on how good the ventilation is and how big the room is.
My guess is that is probably the biggest worry and a relatively cheap
oxygen monitor might be enough.
===============================================
Contact this person. He worked for Oxford and is still installing and
repairing NMR dewars.
dickmrpf_at_ix.netcom.com. Phone number is 508-517-8341. He lives in New
England.
Received on Tue Jan 25 2005 - 13:19:56 MST