Dear AMMRL members,
Thank you all for participating in AMMRL. This group is so very helpful
in providing input from a broad range of experts! I am very
appreciative to those -- listed below -- that took time to provide
information about your experiences. This was particularly helpful, as
we have four unshielded magnets, relatively low ceilings, and time
pressure with a construction crew and faculty that are difficult to
interact with due to coronavirus. The following summary has been very
useful in helping to focus those interactions.
My best wishes that you and yours are safe and healthy,
Charlie
*
*
*
------------- De-energizing / re-energizing magnets -------------*
Comments are scattered throughout the AMMRL responses provided below.
There are three key factors involved:
*Safety to the magnet: *De-energizing and re-energizing high-field NMR
magnets can be done safely, if everything goes as one expects. If the
magnet rod inserts properly, the magnet book notes on last energization
are accurate, the power supply works properly, and the engineer does
everything else correctly, then the magnet will likely come back to
field and all will be right with the world. If there is a problem at
any step, however, trouble. My guess after asking many people about
this over many years is that about 5% of the time the magnet is lost.
*Expense: *The process will cost many tens of thousands of dollars. For
a newer Bruker ASCEND magnets, the cost for a 500 is > $50K.
*
Downtime: *The time for getting the re-energization done along with
letting the magnet settle in would normally be about 2 months. Right
now, in the middle of the current crisis, with who knows how many
magnets will need re-energization once we're all getting back into our
labs, and the delay in getting an engineer to your site with the proper
equipment might be very long indeed. Research will be on hold during
this time. Since spectrometers have unique capabilities in our
facility, some experiments might not be accessible during the wait.
*Conclusions: *All these reasons recommend very strongly that if
cryogens stay available, keep the magnet at field!
*------------- Fire suppression/sprinkler installations-------------*
*
**piping and sprinkler head materials*
-- Due to the very strong magnetic fields and high magnetic homogeneity
requirements of NMR, only non-magnetic materials should be placed within
the magnets 5 Gauss field lines. Metal dust is a very big concern in NMR
labs. Once in the magnet bore, the dust will ruin field homogeneity.
-- U IL-Urbana used plastic plumbing
(
https://www.blazemaster.com/en-us/); might be the best option.
-- others strongly recommend copper plumbing; many strongly recommend
against black steel within the lab
-- do not use alcohol bulbs for heads; use instead bimetal-triggered
heads (have info on Viking heads)
-- doubly actuated systems are strongly preferred
-- some recommend pre-action system where no water is in line until
first alarm goes off; then heat-actuation is required to spray
-- optical alarms are problematic due to possibility of cryogen clouds
being confused for smoke
*location of plumbing and heads*
-- our low ceilings and unshielded magnets exacerbate issues of
placement; we have to be more careful to prevent issues during and after
installations
-- steel (stainless would be similar) piping within 5 gauss line can
make magnet unshimmable due to non-symmetric magnetization of this pipe
(one example where this happened is listed below)
-- piping too close to cryogen fill venting and quenches can freeze;
insulate such pipes
-- if at all possible, do not locate sprinklers within 5 gauss field; if
not possible, get them as far away from magnet center as possible
-- magnets don't catch fire, so some have used that to get sprinklers
oriented to not do coverage spray of the upper parts of the magnet
>> water in the gap of the magnet is problematic; junk in the water
gets into the very strong field and the magnet then may have to be
de-energized to clean up
>> important to not spray cryogen vent ports which might then freeze;
magnet then becomes a high-pressure bomb dangerous to fire fighters and
everyone else working in the lab
*safety during installation*
-- many recommend using plastic bag surrounding magnet with a blower
providing overpressure to keep out dust
-- magnetic dust from pipe work can be _very_ problematic; all work
cutting threading etc. of the pipes done out of the lab, and shoes cleaned!
-- wooden crates are best to protect from tools (we have used these in
our lab in the past; unfortunately the crates were discarded); can also
use large styrofoam enclosures
>> one lab used strong tarp suspensions above magnets to protect
again drops during work above (that work was replacements of heads, and
was quickly done)
-- direct supervision of work in lab is very strongly recommended by many
>> many stated clearly that they would not let such work happen
without direct supervision at all times if at all possible
>> a couple labs marked off areas where the lab staff brought tools
to workers (i.e., workers could not take tools through that area themselves)
-- provide non-magnetic tools and aluminum ladders if at all possible
-- get warnings about magnets to fire dept; keep up-to-date after install
*-------------original email to AMMRL**-------------
*
Hi AMMRL,
Hope you are all staying safe and healthy, and finding good ways to keep
your magnets safe and healthy and alive. Fire sprinkler installation is
sideways from coronavirus, but I am hoping some of you can help with the
following.
*First a generally useful question to many due to coronavirus is how
dangerous and problematic de-energizing and re-energizing is for
superconducting magnets? *One component of the question is what
percentage of magnets don't survive such work? But there are other
issues including costs and length of time to re-energize. I have my own
views on this, and will add that to responses here when making up a
summary to post.
Now to fire sprinkler questions. We have been looking at having fire
sprinklers put into our facility as part of new building construction,
and the coronavirus shutdown has accelerated the timeline for us. Why
not get them in while no one else is in the building? Doing it soon
makes sense. So we saw the first set of plans for the plumbing and
sprinkler head locations this morning. Many of the current locations are
very close to the magnets, some directly over the cans. So we are
working to negotiate better locations further from the magnets. (Note
that everything is happening for staff and dept faculty remotely via
email and Zoom sessions. But the construction crews are still working
full speed ahead in our building complex.) Perhaps it's not surprising
that we're being asked about de-energizing and re-energizing our magnets
to ease the work around them.
*Have any of you that have had fire suppression systems put into your
facility with live magnets already present actually de-energized magnets
just for the sprinkler installations? *We absolutely do not want to do
this, but it will be useful to see how much of this might have happened
in other places. Please provide a bit of context as to why, and of
course your comments on the outcome (good, bad...).
*For those that have had fire sprinklers installed in the presence of
live magnets, we'd very much appreciate any sage advice you might
provide. In particular, how far away were you able to keep the plumbing
and sprinklers? *It would be great to hear any comments you made to the
construction planning group that seemed to be particularly effective in
getting them to understand about our lovely invisible-stray-field
superconducting magnets? We have used plywood crates to protect magnets
during other renovations we've had. In this case, the close proximity
of the current planning would not allow that, and the accelerated
timeline might not enable time to construct them in any event.
Much thanks in advance for any input you can provide.
Best wishes to everyone during these trying times,
Charlie
**-------------*list of AMMRL emails**------------- *
------- 1. Russ Hopson, Brown U
We had sprinklers installed in our entire building in the last couple
years. It was a long process. My facility houses 1x600 MHz and 2x400
MHz spectrometers. The magnets are shielded within their respective
footprints and were not de-energized. The work crew was made aware of
the dangers involved. I believed they may have used some special
sprinkler head. See paragraph below. I think they may have used copper
tubing instead of steel as well. There was an accident and the room was
flooded during install. It only did about $30K of damage. Mainly
pre-amps/temp controllers sitting on floor. I had the shop make little
6" platforms to keep the equipment off the floor in the future.
Good luck. We certainly didn't want the sprinklers in our facility but
we had no say.
------- 2. Dani Rentsch, Swiss Federal Laboratories for Materials
Science and Technology (EMPA at ETH)
Even though we only have one magnet, once the whole laboratory (cables,
electricity etc.) has been renewed with the magnet "on the field". We
packed the magnet in a big plastic bag (at that time we had an
"unshielded one" to keep out dust...but filling of N2 once a week still
was possible) and installed all around a construction of wooden boards
(to prevent tools like hammers etc. from damaging the magnet). Of course
we have put the console in another room for safety.
No reason at all to de-energize your systems!
And: Keep all the fire fighting stuff as far away from the magnets as
possible these firestuff...
------- 3. Clemens Anklin, head of apps lab and VP of Bruker BioSpin
The main question is what material are the pipes?
The location of the heads is secondary. They are supposed to drench a
room in case of a fire so the only thing you can do is to prevent them
of filling the bore by either location of them or by installing some
deflectors later.
I am also not sure what happens if you basically hose down a console
with water.
Back to the material, I think most of the sprinkler pipes are black
steel. The main worry during installations are filings and general steel
and iron dust in the area. Magnets tend to grow beards during such work.
It would be preferable if the system would use copper or plastic pipes
in the NMR lab.
De and re energizing is always an expensive proposition. If it can be
avoided it should. In addition to plywood boxes one can also put the
magnet inside a huge plastic bag that is kept at slight overpressure by
a blower. This keeps the tent up and prevents dust from getting to the
magnet.
Bruker is working from home these days. Maintaining the magnets is work
that is exempt from the business closure order.
------- 4. Gary Strahan, USDA Maryland
We have standard fire sprinklers in our NMR lab. They are not over the
magnets. I have no idea when they were installed, but I suspect that my
predecessors steered them to relatively innocuous locations. The
reality is that they will only emit water if they are directly exposed
to heat, and since our lab is essentially all concrete, I don't see that
as a very likely probability. If a fire ever got so huge as to set one
off, the NMRs would almost certainly be doomed anyway.
We do have non-magnetic de-ionized water fire extinguishers in the lab,
same as used in MRI facilities.
------- 5. Juergen Schulte, Binghamton U
This may not be exactly what you are looking for, but somewhat related:
We moved into a new building two years ago and had sprinklers installed
before we moved in our 600. I made sure that they did not place a
sprinkler head where the 600 was going to go. They also added extra
insulation around the sprinkler pipes because they were worried about
the water freezing during a quench. Maybe a little overly cautious.
Although the 600 quenched twice during re-installation and the
sprinklers did not go off. So maybe it helped???
Now we are waiting for the installation of a 400 to begin. The sprinkler
head locations apparently aren't "set in stone", as the construction
crew was able to move one head 3-4 feet from where I wanted the new 400
to stand, without objections by the fire inspectors. No effect on the
live Ascend 600 a few feet away.
------- 6. Gary Martin, retired from Merck, now Seton Hall and Stevens
Institute in Hoboken
We moved a number of magnets while I was still at Merck, probably at
least 10. Only one refused to come back to life if that statistic helps
any.
The NMR labs at Merck had a Halon fire-suppression system built in the
labs before any magnets were ever installed, which isn't much help
either. We managed to convince safety that water sprinklers over NMR
magnets and electronics consoles were a no-no when we redid one lab for
3 x 600 MHz instruments.
------- 7. Markus Voehler, Vanderbilt
We have always tried to stay away from sprinklers in the NMR lab as much
as possible, in the Biomolecular building successfully with a variance
from codes. Where we were not successful to keep them out, we installed
a dry system, which I highly recommend for an NMR facility. In this
system, the pipes are dry and it requires a dual activation for the
sprinkler to actually go off. It is not only the alarm that triggers it,
it also requires that the sprinkle experiences a certain temperature
before the bi-metal opens the head for water to flow and reduces
“accidental” take off greatly.
I’d definitively would also make sure the sprinkler is placed as far as
possible away from the console and magnet, so not to dump an excessive
amount of water directly onto the equipment. There is an obvious risk of
that on top of the damage it might cause.
We never had to do this work in a “hot” environment.
------- 8. Jerry Hirschinger, Purdue
Regarding refurbishing magnets – a stable magnet should be returned to
field without much problem, so long as it is treated with the care it
deserves. Unhealthy drifters are never a good bet. Each time a magnet
is re-energized, it should theoretically become more stable.
Regarding fire sprinklers. The obvious solution to placement is to
require copper plumbing. The contractor might have to do some
searching, but copper drop plumbing is indeed available. Another
concern is the alcohol vial design of modern sprinkler heads. In a
quench, the vial of alcohol can be ruptured by the cold exhaust gas. Be
sure they use the old bi-metal type for sprinkler locations in the
quench or fill exhausts.
------- 9. Allan Kershaw, USC
Just to let you know, at USC, we have opted to keep the magnets up. We
have been assured by our LN2 and LHe suppliers that they will deliver as
needed. Another thing that works for us is that I have had an LHe
recycling system for the past 5 years, so LHe is a little less of a problem.
As for fire sprinklers, my NMR rooms have had them since 2004. They
were installed while the magnets were charged. While construction was
happening, the magnets and consoles were wrapped in plastic sheeting to
keep dust and debris out of them. I did work with the contractor in
placing the sprinkler heads as far as possible, about 3' or more, from
the magnets and consoles while still maintaining the code required
coverage. The feed pipes were run along the sides of the room far from
the magnets and consoles. The heads were then run to the required
locations with stub pipes. My room is fairly high and the sprinklers
are fairly close to that ceiling, so field distortion was not significant.
Another concern that should be brought up is what happens if the water
is released. The consoles will probably take a bigger hit when spraying
water on them than the magnets. I did bring up the possibility of other
suppressants. Powder release is just as bad as water and the safety
concerns with Freon made these options not viable nor cost effective.
What we do have is called a pre-action system. This system takes two
triggers to release the water. The pipes are actually empty and are
under a little vacuum. If something were to break a head, the water
would not release and an alarm would sound. As mentioned, it takes two
triggers. The first would come from the smoke detectors. When
triggered, the water would be released into the pipes. Then if there is
a real heat source as in a fire, the heads would open and release the
water. We have minimized the amount of flammables in the NMR rooms, so
it is deemed a minimal and safe risk to have the water system. So far,
in the 16 years of the system, we have had no triggers of any kind.
------- 10. Rainer Haessner, Technical University of Munich (TUM)
Concerning deenergizing/reenergizing...
About two years ago we moved altogether 6 magnets to a new building.
Between 400 and 800 MHz. No technical issue at all, no quench, but crazy
time consuming. And we needed 8000 litres of helium, which was no
challenge that time.
About 15 years ago a 500 MHz magnet didn't come back to field after 8
attempts.
Do you really believe, deenergizing magnets will become an option? At
least this time in Germany doing cryogen service is possible.
------- 11. Dean Olson, U IL-Urbana-Champaign
There is a great simplification you can bring to the fire sprinkler
installation scenario, and it is this:
Use plastic pipe.
See these web pages:
https://www.blazemaster.com/en-us/ (we used this exact product; it has
a metal film liner)
https://www.ppfahome.org/page/FireSprinklers
We had this plumbing put in by ordinary guys on aluminum ladders. We
covered bore holes, and supervised their work by direct, in-person
observation. Like sit in a chair and watch them. Be conscious to not
put the pipes or sprinkler heads where cryogen vapor will go, neither in
a routine situation, nor in a quench condition (intentional or not).
Also, keep sprinkler plumbing away from where helium transfer line
apparatus will go.
That’s it, really. Depending on the experience of your work crews, they
will either be surprised by the rated plastic pipe available
commercially, or not. I believe it is more expensive, but impress upon
them that this option is much, much less expensive and risky than any
manipulation directly involving the magnets.
You should be OK, and disruption to actual work can be minimized to a
few days.
Good luck with all the other virus kerfuffle. No talk of de-energizing
NMR magnets here.
------- 12. Martha Morton, U Nebraska
Don't let them de-energize the magnets. Box the magnets instead, with
a space for doing nitrogen fills rigged into the box.
Fires won't be near the magnets and the sprinkler heads can freeze in a
quench situation. Sprinkler heads should be closer to consoles or better
yet don't install sprinkler heads (water and electricity don't mix after
all!)
I have to install water lines for a 600 cryoprobe installation. I feel
your pain.
------- 13. John Synder, Bruker Midwest service engineer
Any sprinkler head near a magnet should be the metal type. The glass
bulb type can break during a quench
So to the planners — what items could start a fire?? Not likely to be a
cryomagnet
So consoles, cryoplatform, compressors, chilled water units, lasers for
your 600
Keep the mainlines outside your 5 gauss lines and then if they want a
head near the magnets— get them to use very short pipe segments.
I’ve seen some that have a deflector to shoot the water further in one
direction
Of course then you have to deal with the dust during install
And SUPERVISE all the work. No work without someone there.
------- 14. Phil Dennison, UC-Irvine
We had a new sprinkler system installed last year. The process was very
slow, but quite smooth. My lab has a pretty high ceiling and the magnets
were all energized. There was also a new alarm sensor and warning system.
The leaders of the contractors were reasonable people and consulted well
over the needs of working around magnets. Unfortunately they didn't
always pass that information on to the workers actually doing the
installation. One day I found a contractor sawing through electrical
ducting on the floor close to my unshielded 500.
Having said that, overall there were no problems with the instruments
during the process.
------- 15. Chris Barr, U Victoria
We had sprinklers installed a couple of years ago without shutting down
the magnets.
Initially they planned to run the line right down the middle of the lab,
but after discussions they agreed to run the main lines down the
hallway outside the lab and only the sprinkler heads enter the lab.
They drilled from the outside, and put tape and plastic over the inside
so no dust issues. Since the lab is narrow (15-16 ft) the sprinklers
they installed can spray right across and did not need to be in the
center of the lab. I’m not sure this is an option for your space, but
you might want to see if they can come up with options.
They also installed water-based A/C units, and again ran the main water
outside the lab, but had more extensive installations inside the lab.
We negotiated the positions in the lab, and the workers were amenable to
me delineating the space where they could work, and the no-go zones with
post and hazard tape. I was also the only one allowed to move tools
from the door to the work space, and the workers were 100% compliant
with this.
------- 16. Dave Russell
While I have no direct experience with just a sprinkler installation,
I’ve been involved with several lab renovations and the issues are very
similar, I think.
Non-shielded magnets? No advice, and it scares me to even think about it.
All shielded systems? Too many people worry too much. Move the
sprinkler heads away from the magnets as much as practicable so no one
has to work immediately above the bore, cover the magnets with
plastic/tarps/plywood/whatever, put up a physical reminder (like a
plastic chain or a cardboard wall) around the magnets giving a foot or
two of clearance, warn *everyone* about the systems, and then let them
do the work. They won’t want to be the person responsible for
destroying a $1M magnet, so they’ll be careful.
------- 17. Jeff Walton, UC-Davis
Talk about timely - we just went through this. We had them build plywood
boxes around the magnets. Sprinkler heads were not over the magnets
although plumbing runs were. Boxes went around 3 sides and went higher
than the top on the magnet. We also read them the riot act about
magnetic field safety. Ours survived although one of Matt Augustine’s
(down the hall) quenched during this. It was unclear if it was the
result of the installation but we have our suspicions.
We reused these boxes when they came in to plumb the Helium recovery
system.
And them we moved them to our other location and put them in place while
they did renovations in prep for moving a magnet in there. There are
still lights to replace and duct work in other rooms and so we will be
moving those boxes from room to room as the work is done.
------- 18. academic lab (asked to remain private)
"basically hopeless for construction workers to remember the safety
issues during their work"
------- 19. Gerd Gemmecker, TUM (Germany)
How are things going at UW Madison? Everything shut down? Except for
construction work - same here at TUM: we are
not even allowed to enter the buildings without being on the dean's list
of emergency personnel, but the construction
people seem to continue as normal.
Let me add that we managed to move most of the magnets cold (across the
street, from the Dept. bldg. to the new NMR centre).
Actually, discharging & re-charging those 5 magnets (400-600 MHz)
proceeded quite smoothly, and all magnets, even our 29 yr.
old Oxford 500 came back on field right away and are happy ever since
(knock on wood), a matter of a few days.
Since they stayed cold, He consumption was moderate: just the additional
boil-off during the charging.
The two 800s were a different story, they moved here from ca. 12 km
away, of course warm and completely dismantled.
Cooling them down again (& charging) was of course VERY time and He
consuming (ca. 2 months).
One alternative to discharging them might be protecting the magnets in a
makeshift plywood housing (or similar), to keep
workers, their tools and magnetic scrap away from the high fields. I
have seen this at several places where they had
unavoidable construction work going on. Of course that might not be a
workable solution for a large unshielded magnet
with several meters of tool-attracting stray field ...
I would strongly suggest to try to move all plumbing and sprinklers away
from the magnets. Keep in mind that, sooner or later,
someone will come and have to test those things, or replace them. You
can't discharge the magnets every X years ...
We don't have sprinklers in the NMR rooms, but just the electric
lighting - which of course needs replacement from time to time
(burnt out bulbs and, yes, even LEDs).We tried hard to move them away
from potential magnet positions - if they are located right
above a magnet, that will be very tricky (and of course that's the one
that always dies first).
Of course the new NMR hall is high enough (> 8 m) - but this means the
workers will have to use some sort of (iron?) platform
on wheels to reach there ... again a source of concern.
Also I would be careful with having the sprinklers mounted to close to
the magnets - what happens if they are engulfed by very cold
N2 or He gas during refills? Might they be activated? Hope the room is
high enough (I'm pretty sure no one knows what happens
to them at -100 degrees ...).
------- 20. Oliver Callies, AbbVie (Ludwigshafen, Germany)
I am writing as the head of the NMR lab at AbbVie in Ludwigshafen
(Germany), and I can tell you that we have a sprinkler installed in our
magnet room. However, as the installation was performed before my time
at AbbVie, I cannot provide information on how the installation was done.
In any case, for the sprinkler, which is just above the magnets, we have
an additional security valve outside of the magnet room, which must be
opened manually in the case of a fire outbreak. This installation has
been included in our fire emergency plan, so that the fire brigade knows
about this installation, as well. This avoids that a sprinkler
unpredictably gets activated.
Recently, as we are planning a new lab, Bruker has told me that another
possibility would be to use metallic caps for the sprinkler, as the
normally used caps might break more easily, e.g. in case they are
exposed to suddenly lower temperatures, such as during helium fillings.
However, Bruker insisted that the manual security valve was the better
solution.
------- 21. Bob Berno, McMaster (Montreal)
Regarding de-energizing magnets during this crisis, I think Ryan McKay
posted a thorough response. One additional point to consider is the
additional down time due to all of us needing to queue up to have our
magnets re-energized. There are only so many power supplies to go
around to charge these magnets, especially at the very high fields.
And then there is all that extra helium we will need. It’s one thing to
keep the magnets topped up with helium. It’s quite another to energize.
I think our respective administrations are simply going through the risk
management exercise: identify all possible contingencies, and then
weigh the pros and cons of each before making any decisions.
For me, I’m fortunate that our University President is a chemist who
understands the importance of core facilities. The delivery of
cryogenic liquids has been deemed an essential service, so I didn’t have
to expend much energy to justify keeping our magnets at field.
Regarding sprinklers:
I have not had to deal with installation/moving of sprinklers near live
magnets. About 10-years ago I was involved with a room renovation where
a new magnet was to be installed. At that time I was working with a
Bruker engineer names Thomas Buser (he used to be the Bruker Service
Manager for Canada, but he has since gone back to Switzerland) who was
very smart about how best to deal with these. He had them positioned
such that the entire flow would be covered by the spray from the
sprinklers, but the spray would not strike the magnet can at all. In
fact, spray would go just below the gasket/seam.
Yes, all of the electronics would be toast, and maybe the cryoprobe as
well. But the magnet would be mostly dry.
So, from your description where they want to install a sprinkler
directly above a magnet, that seems terribly scary. The magnet isn’t
going to catch fire, so it doesn’t need a suppressing spray.
------- 22. Dave Davidson, U Manitoba
We had a sprinkler system installed over our unshielded Bruker 400 MHz
widebore magnet in January. The physical plant framing shop built a
plywood box to surround the magnet from 5 sides, the overhead sprinkler
work was done, and the box was removed without incident. It was not my
preference to use piping overhead and I voiced this during the planning
stages but the work was deemed important enough to go ahead regardless.
We repeated this process when the room temperature shims on the
spectrometer were affected badly enough that adequate lineshape could
not be obtained with the newly installed overhead pipe in place, which
was then re-routed around the instrument.
Both times we had no incident (although I had a fair amount of
trepidation before the work in each case). The instrument is now back to
operational status with updated shims and a good lineshape.
------- 23. Joe Dumais, Boise State
*First a generally useful qu**estion to many due to coronavirus is how
dangerous and problematic de-energizing and re-energizing is for
superconducting magnets?*
I have experience from the vendor side of this endeavor as well as
research lab experience too. I have never heard of a properly
de-energized magnet not being able to be returned to field. I have seen
cases (dealt with one myself) where the shorting plug was frozen into
its connector, we successfully got it out but it was touch and go, as I
understand it sometimes a pin can break off during that process and you
cannot get the stick into the connector to ramp the magnet down. In my
case the magnet was being de-energized for disposal so I was able to act
pretty freely. Most sites do not have the sticks on site for new
magnets so that is something to be aware of. It used to be safe to
de-energize them with a diode pack but I am not sure it that is
compatible with current magnet design.
*Have any of you that have had fire suppression systems put into your
facility with live magnets already present actually de-energized magnets
just for the sprinkler installations?*
Many years ago we had a fire suppression system installed in an NMR lab
at Bell Labs with live magnets without clearing the work with us. It
was a deep bay setup with steel separating walls and they hole sawed
large diameter holes through each separating wall and fished pipe
sections through and joined them in each lab. Fortunately for us we
found out about it after they had only done this near a 4.7 T narrow
bore. We were able to coordinate and protect our other systems before
the work started but we kept them energized. Other chemistry research
labs found steel filings on active research benches. That really taught
me to make sure I am talking with the right person and make sure they
communicate the issue to their crew and given the opportunity I will
also talk with the crew.
*For those that have had fire sprinklers installed in the presence of
live magnets, we'd very much appreciate any sage advice you might provide*
Are they using cold steel for the pipes? Any chance they could do the
NMR lab in copper? The might not meet code. I have used wood to protect
magnets from tools at times and lots of plastic and duct tape to protect
against filings, remembering to vent the boil off gases out of the
"tent". Are these systems set up such that when one head triggers they
all go? I have heard folks say that once you get to the point of
sprinklers triggering then the magnet is the least of your worries.
Another question I would have is what is the failure rate of the heads
in terms of leaking and how well are they protected against being
bumped with the Helium transfer tubes and such?
------- 24. Hasan Celik, UC-Berkeley
We did not have a full blown installation project around live magnets
involving sprinkler systems, but the facilities services had to replace
all sprinkler heads in my lab space. This is a significantly smaller
project compared to what you have to do, but I wanted to share our
experience nonetheless with the hope that it might help.
The sprinkler heads in my lab (and in our building in general) were over
50 years old, so the fire marshall mandated a project involving full
replacement of all the heads. They were located all over the place with
some of them directly over the cans, similar to your situation. After
extensive discussion with guys from the facilities services, it was
determined that the heads were made out of bronze, and all the pipes
were copper. There was no direct threat to the magnets from the
components that would be replaced. However, bronze can apparently
magnetize when exposed to external magnetic fields over a long period of
time depending on its alloy content (I didn't know this!). We didn't
want to take any chances, so we had to use a secondary barrier between
the sprinkler heads and the magnets at the time of replacement in case
one of them dropped on the magnets. Due to lack of time for
preparations, as this whole project happened very quickly, we had to
stretch tarps above the magnets while the work was taking place.
Fortunately, each replacement would take ~10 min, so we could move on to
the next magnet location very quickly. Also, I provided non-magnetic
tools and an aluminum ladder for the crew, and asked them to leave any
other tool behind while working around the magnets. Luckily, no drilling
was involved, but I find the drills that come with a vacuum attached
next to the tip to be useful, and mostly safe to work around magnets,
assuming the tool itself doesn't get too close to the can.
As for training the crew to work around strong magnetic fields... Beyond
the usual spiel about the yellow lines around the magnets for 5G, the
following works best for me to make the crew to understand how strong
(and potentially dangerous) our fields are: I compare our magnets to the
electromagnets used for lifting junked cars. Those magnets have ~0.5 T
field strength, and when you say they will work around magnets that are
20x to 30x as strong, they tend to start respecting the space around the
cans. By the time they feel the pull on their tools, it will be too late...
------- 25. John Decatur, Columbia
I can’t speak to the sprinkler issue but we discussed the possibility of
de-energization. As you know, this is a move of desperation and would
only be carried out if there was a real chance of cryogen disruption or
absolute prohibition on personnel. So far, neither of these seem to be
on the horizon…..
I spoke with Brian Lord at Bruker about this and Bruker will not
guarantee that the magnets will come back to field nor that all
cryoshims function. I don’’t have any direct experience with the
likelihood of these negative scenarios actually occurring, but I would
think if the magnet is safely ramped down, then the chances of it coming
back perfectly are good - maybe 90% is my guess. But I don’t really know.
But that assumes the magnet ramps down smoothly which requires exact
knowledge of current in the main coil and shims and that it hasn’t
changed appreciably over time. When they open the cryoswitch, if the
current in the power supply doesn’t match the current in the coil, there
is a voltage that may exceed the diode and a quench will occur. Some of
my magnets are so old and there is no data on them, to my knowledge, so
ramping them down would be very risky.
Even riskier, however, is allowing magnets to quench because of no
cryogens. Then, the likelihood of damage would be much greater, is my
understanding, since all the energy is dissipated at once and
burning/breaking some part of the coil is a real possibility.
I have a current quote for moving a 300 magnet that involves
de-energization, and re-energization and it is $42K. Subtract maybe $5K
if you don’t move and I believe that is a reasonable cost for cycling a
lower-field magnet. Time is an issue and Bruker is busy. I would think
delays of up to 2-3 months could be expected. Then the magnet will drift
for weeks preventing useful data collection for several more weeks.
Maybe other vendors could do if faster…
------- 26. Monika Ivancic, U Vermont
Well, these are interesting times to be getting your fire sprinkler
systems installed. As long as the workers are still allowed in the
buildings...
I don't have experience having fire sprinklers installed while the
magnets were live. In our new building this was one 'last but not
least' item to deal with, and another headache with all the rest in
getting proper utilities for a new NMR lab.
You likely already know to not get alcohol based triggering for the
sprinklerheads, since a cold quench could rupture these, causing the
sprinklers to go off. Instead make sure they use bimetal-triggered
sprinkler heads. I'm attaching an info sheet on these Viking ones that
were used in my lab. They also used copper piping in my lab and
fortunately we didn't have to insulate the pipes, because they are so
high above the magnets.
The best solution for sprinklers in NMR labs are doubly actuated
systems, which normally only have air in the lines and water is allowed
down the pipes when the alarm goes off. After this point the sprinkler
heads work normally.
------- 27. Letitia Yao, U Minn (Chemistry)
They put sprinklers in our lab when we renovated in 2002. I didn't have
any say as to where they went. The magnets were all energized and they
were all unshielded. We put a plywood box around the 500, and big
styrofoam ones around all the others.
http://nmr.chem.umn.edu/renovation.html
Received on Mon Apr 06 2020 - 06:29:31 MST