Dear All,
Thank you so much to everyone who responded with experience, expertise,
anecdotes, and advice regarding my posting about interference from
high-voltage/current AC power lines near NMR spectrometers. I can't even
try to name everyone, so I'll just thank everyone who responded and express
my sincere appreciation. I will summarize what I learned as briefly as
possible.
1) There is a big distinction between high-current AC lines that supply
a fairly steady load, vs. installations where there are transient surges in
power usage. The latter is a much bigger worry, and there were many
stories of commuter trains, elevators, .etc. that caused transient problems
that shouldn't be an issue with continuous, steady AC current (as is
expected in our case here with the super-computing facility).
2) If the power is 3-phase (almost certain in our case), and if they
use twisted bundles w/ ground, the magnetic fields should effectively cancel
out and stray 60-cycle EMF should be minimized. Several people pointed this
out, and it makes sense. I haven't verified the accuracy or completeness of
this cancellation, but I think it is correct (at least to a large degree).
Some also pointed out that it is at junction boxes, or cable-splices where
the stray fields could be a problem.
3) Considering (2) above, if one also uses high-permeability iron (or
other magnetic soft-metal) as the conduit, one can contain any residual
stray AC magnetic fields very effectively. Different ideas like iron
drain-pipe, or other soft, high-permeability metals, have worked in many
places. One site has a 1600 Amp AC line running through such a conduit in a
lab between two high-field systems and they experience no ill-effects.
4) We DO want to find a currently operating power-line with similar
voltage/current load, and make careful measurements ourselves to verify the
distance needed for A/C EMF noise (60/120 Hz) to become within background
levels. People from Varian have responded, and are standing by to assist us
if/when we can locate such a place to make quantitative measurements.
Varian assisted us well in site-planning measurements for our 800 facility
pre-installation, and they are willing to help us again. The difficult part
is finding a suitable site, but we are working on it. Many, many people
strongly suggested that we do this, and we too feel that it is necessary.
5) Some people brought up something that I didn't initially think of,
which is Rf emission from the super-computers themselves. I have no idea
what kinds of clock-speeds/frequencies their CPUs will be using, but with
all of the ultra high-speed computing, there could even be concern for Rf
interference that might creep in on us (it will be just outside). We may
need to beware of Rf interference (not from the power-lines, but possibly
from the massive banks of super-computing processors) to the design
engineers. I was only thinking of low-frequency magnetic field modulation,
and didn't think about what was being "plugged-in" on the other end.
Hopefully, the new building will be well-shielded and this won't be an
issue. but it's worth thinking about.
I hope this summary does justice to all of the wonderful colleagues who took
time to respond to my posting. I am ever appreciative, and (as always)
humbled to be part of this group of talented professionals.
Sincerely,
-Rich Shoemaker
---
Richard K. Shoemaker, Ph.D., Director, NMR Spectroscopy Facility
University of Colorado at Boulder
Phone: (303) 492-7062 Fax: (303) 492-5894
E-Mail: Richard.Shoemaker_at_Colorado.edu
Web: http://chemnmr.colorado.edu/rshoe
From: Richard Shoemaker [mailto:Richard.Shoemaker_at_colorado.edu]
Sent: Thursday, January 21, 2010 6:01 PM
To: 'ammrl_at_ammrl.org'
Subject: AMMRL: Interference from High Voltage Power Lines?
Dear All:
I know certain issues have been addressed about the impact of new
utilities/services near existing NMR laboratories, but I'm not sure this one
has been discussed exactly. If anyone has any real-world experience that
might help us to understand the possible impact of this I would appreciate
hearing from you. I will, of course, summarize any useful information that
I receive.
Here is the issue:
Just outside our 800MHz NMR facility, they plan to install a new
supercomputing facility that demands a lot of power. Due to the location,
the current proposed plans have them running the AC power feed just outside
the NMR lab. My concern is EMI interference that would be picked up in
various ways by the NMR console, probe, cables, .etc. The power lines will
be carrying 480 VAC, 1000kVa, 1200 Amps. One of the plans has the power
lines going through the building just behind the wall where our console is
located, and another would site them on the opposite side (maybe 50 feet
from the magnet, and 65 feet from the console).
1) Does anyone have any experience that might suggest whether or not
this will be a problem? I have visions of all kinds of 60/120 cycle
interference showing up in strange ways in our data.
2) Does anyone have specific ideas how we could somehow test the
impact, in terms of interference, before they actually build this thing
because after the fact it will be too late?
The combined body of knowledge of this group is mind boggling, so I look
forward to hearing from you. I have already passed this along to Varian,
hoping that they might be also able to assist us in dealing with this.
Thank you in advance,
-Rich Shoemaker
---
Richard K. Shoemaker, Ph.D., Director, NMR Spectroscopy Facility
University of Colorado at Boulder
Phone: (303) 492-7062 Fax: (303) 492-5894
E-Mail: Richard.Shoemaker_at_Colorado.edu
Web: http://chemnmr.colorado.edu/rshoe
Received on Fri Feb 05 2010 - 11:07:29 MST