Hi,
I recently asked this question to the ammrl community:
Our group is working on a grant application for a new nmr system. We are
> hoping to get a 600 MHz with a TCI cryogenic probe. The big point on which
> we hesitate is about the new aeon magnet. Their system is very attractive,
> especially since we don't know the helium price in the future. However, if
> the He compressor system go down, we have one week to get it back to work
> before a quench. Also, i have been told that, since the compressor is just
> beside the magnet, it generates a lot of vibration which is not good for
> the SNR. I would be very happy to ear all of you (opinion or aeon user
> feedback) on that topics.
I would like to post a summary of the answer i received. Also, i will give
my own thoughts about the question
But before, let me thank all of those who send me their thoughts, i really
appreciate!
=============================
I know, I am not independent here, but we had a 600 MHz Aeon Magnet in the
demo lab in Massachusetts for over a year now.
If you go looking for it you will find small artefacts that come from the
compressor.
But it does not affect SNR overall as these artefacts are limited to just a
few Hz around signals.
The compressor can be up to 60 m away from the magnet. This reduces the
vibrations and the noise in the lab.
To be honest one has to consider compressor failures and also cooling
failures, when using a watercooled compressor. We do recommend a
auxilliary colling such as tap water. Electricity failures will also
interrupt the cooling.
=============================
I don't own an Aeon, but it's almost $200k more than for a conventional
magnet. Before paying this amount for liq. helium... Remember that a 600
MHz needs a maximum of five annual refill of 100L.
Adding that the maintening cost for a cryoprobe will be soon a pain in the
a.., i don't think that's worth the cost
=============================
I don't have an Aeon magnet, but I think that they are the future. At the
moment they are still expensive compared to buying helium, but it is the
best way to recover helium.
Bruker claim to have solved any vibration issues, I haven't heard of any
problem. But then I don't know how many Aeons they have actually sold. When
I last got a quote, the price was $170,000 extra for the Aeon magnet. That
buys a lot of helium!
Here's a question you can ask Bruker - can you manually add liquid helium
to extend the time before a quench if the compressor fails?
=============================
The catch behind the Aeon, particularly at 600MHz, is that you have to
maintain two pulse tubes. That requires an extra investment for the
maintenance of the pulse tube (Bruker has to do it) of $25000 USD a
year. To my understanding you have to pay that money upfront. On the
other side, it is much more expensive than the 600 Ascend ($200000
more?). Even you pay a very high price for LHe, it is much cheaper the
cost in LHe than the difference in money you have to pay for having
zero LHe boil-off rate.
I would go for an Ascend.
=============================
Do you have other NMR magnets now? Your money might be better spent on a
helium recycling system if you have a helium re-liquification facility on
your campus.
=============================
I cannot comment on the issue with the pulse tube cooler going down, but
the cost of the required maintenance on this cooling system must be weighed
against the cost saving of helium fills. Regarding the vibrations, the
lineshape standard spectrum will tell you everything you need to know I
think. If the artifacts from vibrations are within spec for this test then
you should be fine as they should not cause any issue.
=============================
I did some investigation down this avenue a while back myself, for lower
field magnets (400/500) there is only a single cryocooler for the helium
dewar, and no liquid nitrogen jacket exists, leading to a very short time
to quench (about 1 week). Because of the lack of a outer jacket, you still
have to do helium fills, but only about once every 2-3 years.
In the case of the 600/700, there are 2 cryocoolers that cool two separate
jackets in the cryostat, and helium hold time is effectively infinite, with
only minor top ups very infrequently, or due to compressor failure. Both
systems require yearly compressor maintenance like a cryoprobe, roughly the
same cost as well. My opinion is that there are better options for helium
recovery that mitigate a single-point-of-failure. I'd personally rather opt
for a regular 600MHz cryostat, and get a Quantum Designs helium liquefier.
The helium recovery capacity of their ATL160 has the capacity to service
multiple magnets easily with the appropriate piping, and with their
purifier even take helium from high pressure cylinders and liquefy it as
required, completely freeing you from liquid helium supply. The only helium
you would need to bring in, is to compensate for boiloff during a helium
fill (which, with further infrastructure with a traditional gas bag and
booster compressor system) would get you to near 100% recovery (at current
prices, overkill). The bonus is that the ATL160 also acts as the helium
transfer dewar and is magnet safe, allowing you to run your fills on site
as required without supplier lead times.
=============================
The vibrations: We have just finished installing an Aeon 400 MHz in
January. This is a slightly different configuration as the 500 and above
magnets come with a two-stage refrigeration system whereas ours is
single-stage. As our system is so new, I have few cautionary comments to
make. We did experience some issues with the vibration of the pulse-tube
cooler's effect on the SNR. The performance qualification tech actually
recommended we turn the pneumatic legs off to reduce the amount of
vibration. However, it is noteworthy that our instrument is placed
slab-on-grade, and the accelerometer measurements at the site survey were
extremely small. This, despite heavy construction equipment operating
within 200 meters from the accelerometer. Also, the technician was back
out for an unrelated issue with instructions to measure SNR with both legs
up and legs down for the Manager of Magnetics Division (N. America). I
don't know what is happening with that data. While I have not yet
conducted any experiments that will push the SNR limit, the performance
qualifications stated the sino=296:1 for 0.1% EB in CDCl3 with the legs
down and is currently at 226 with the legs up and 222 with them down. No
difference in my opinion.
The cost: I made simple calculations as to the cost of cryogens for a
conventional magnet system vs. the additional cost of the Aeon, and I
recommend doing the same for your quote. Use the difference as a
bargaining point with Bruker, but factor in the amount of time you'll gain
by not filling instruments with LN2. What a relief! Our helium is holding
at least as good as they predicted; I expect to buy helium no more than
once per year.
The worries: Finally, at the install I spoke with several of the magnet
engineers who are telling me stories of failure testing at least two He
compressor units at the Billerica site. Rumor is that one has been running
for > 5yrs without servicing. That is reassuring, but having to replace
one in less than a week if it does fail is a little unnerving.
Best of luck with your decision, and don't hesitate to send me email again.
=============================
The Aeon magnet does seem to be very robust technology, so from a
spectroscopy point of view, I wouldn't worry about compromising the quality
of the data.
The big advantage is that you will be much less reliant on a supply that
you cannot control. I have been quite lucky in that whenever I NEED liquid
helium, I've been able to get it. But I have heard far too many horror
stories from people who have not been so lucky.
The concern you have concerning a breakdown/power failure are in many ways
no different than anyone that has a sub-cooled magnet (generally > 800MHz).
If the pumps fail, you have only so many days before the magnet quenches.
If I were to get an Aeon, I would make sure the system was on UPS/Emergency
Power.
I like the idea of an Aeon, but you will have to weigh the following
factors:
Cost of helium vs maintenance cost of helium compressor
Reliability of helium supply
Upfront capital costs: 500MHz Aeon vs 600MHz Ascend.
Upfront infrastructure costs (additional construction, space, utilities).
Noise (we have two cryoprobes, and I have sited the compressors in an
adjacent room to keep the noise down).
You might be able to build a sort of sarcophagus around the compressor, if
you don't have a convenient adjacent room.
How much more efficient are the new magnets than what you have currently?
We have a five year old 700MHz that uses one/third the amount of helium
that our old 700MHz uses. And the new ASCEND magnets are even better!
Anyway, I think from a technical point of view, I believe this is the wave
of the future. You will have to weigh out all of those different factors.
From a fiscal point of view, you have a pretty good idea how much the
maintenance costs will be.... can you predict how expensive the helium will
become in two - five - ten years? How long will you keep the magnet?
=============================
=============================
my comments
=============================
About the possibility to add liquid helium if the compressor fail. The
answer is yes, and it's the only way to avoid the quench, apart to restart
the compressor.
A lot of people talked about the additional cost for a Aeon magnet which is
$200K: it a good estimation, but a little high. I don't know for us, but
here in Canada, founding this kind of equipment is a matter of winning the
"graal" grant CFI. However, founding the maintenance of all scientific
equipment is really a pain in the ass. So, having to weight the investment
for an Aeon magnet at the purchase of the system (by the CFI), and the
helium during all lifetime (by who knows) would be an easy one. However, we
need to consider the maintenance cost of the compressor which is also very
high (about $20k/year). Because of that, we still hesitate.
Concerning the Quantum system, i tried to contact them, but i didn't get
any answer. If someone knows the contact there, i would appreciate the
information.
*Serge Lavoie*
Département des sciences fondamentales
Université du Québec à Chicoutimi
555 boul. Université
Chicoutimi, Québec (G7H 2B1)
Received on Mon Apr 07 2014 - 11:20:29 MST