EMI and safety

["Marco Caceci" <radal@ptv.es>]

Dear Radsafers:

I would like to comment on the recent thread about instrument failures due to EM (electromagnetic) interference, which may be
relevant to the ongoing discussion on safety and safety culture or lack thereof...

Some measuring instruments were reported here on RADSAFE to fail when subject to 'strong' EM fields. I am surprised that recent
posts indicate implicit acceptance of this as a fact of life, to be corrected by extra shielding (aluminium foil, for heaven's
sake...) or whatever tricks our creativity can come up with.

In the US, EU and other civilised countries, the law sets clear limits on the amount of EMI (interference generated) and EMS
(susceptibility) allowed in any electronic device sold on the market.
In the US, the law is set by part 15 of the FCC rules, which claim 'jurisdiction' on all electronic devices: check the back of your
mouse and you will see the FCC sticker there.
More on this at http://www.fcc.gov: the complete text of the law is at address
http://www.fcc.gov/Bureaus/Engineering_Technology/Documents/cfr/1998/47cfr15.pdf for all to enjoy.
Compulsory compliance with well defined limits on EM emission and susceptibility levels (EMC: ElectoMagnetic Compatibility) is
indeed an excellent idea. If a device fails in the field but complies with the norms, it's your environment's fault. But if it does
not comply with spec'ed norms, it is the manufacturer's fault. Also, if you can trust compliance with definite EMC norms,
calculations can be made about failure risks etc.

That's the theory. Now, unlike the EC in the EU, the US FCC allows (paragraph 103.c of part 15 of CFR 47) exceptions to its own
rules, specifically in the case of 'test' devices. And lo and behold: nuclear instruments such as ratemeters, dosimeters, and even
safety related and mission critical gamma and neutron meters automatically controlling pumps and valves... have been ruled, by the
FCC with acquiescence of the NRC, exempt from the requirement to comply with FCC EMC rules, having been decreed to be 'test'
devices. What they test is open to discussion.
In proof of this, I post at address http://chemitech.com/NRC/FCC.GIF copy of a letter by Mr Art Wall, Chief of the Customer Service
Branch of the FCC, to Mr Earl Pollock, the (senile) President of Nuclear Research Corporation (letter kindly passed on to me by Mr
Theodore R Quay, Chief of Quality Assurance in a long-named office of the USNRC) in which the FCC, having obtained a sample
instrument for testing, following reports of non-compliance... ruled that compliance was not required! (I wonder what were the test
results...).
To my knowledge no stock instrument sold in the US by Nuclear Research Corporation (whose SNOOPY instruments were reported in a
previous post to fail in the field due to EMS) claims compliance with FCC rules. Many but not all other manufacturers' products do
the same. I have little doubt that failure to state compliance is generally due to failure to comply; in my opinion failure to
comply is a clear sign of bad electronic design, and bad electronic design in any instrumentation is symptom and result of
complacency if not irresponsibility, and a reliable predictor of trouble to come.
The FCC wouldn't let across your border one single Japanese toy that did not comply, but gives up on its own jurisdiction on nuclear
electronic instrumentation. The NRC sees nothing wrong in this. Let's wait and see what they'll all say next time a plant's
monitoring net goes ballistic. Maybe it will be your fault.

This, in my opinion, is serious matter, with definite safety implications. I can envision scenarios in which, following a minor
incident, a team intervenes, equipped with walkie-talkies... and the meters go berserk shutting valves up and down. Alternatively,
operation of high-current emergency systems may generate EM fields at levels never tested before, which may cause failures in a
plant's safety gamma and neutron monitors...
There are interesting pages on the web on what terrorists could do with home-made EMP devices, but really the implications here are
that any kid with a Hertz coil could break havoc in a power plant. Prove that that is not the case.

Nuclear instrumentation can be produced (and is produced by many vendors) in compliance with FCC norms. I cannot understand then why
the FCC requires EMC compliance of all electronic devices... except nuclear instruments, nor why the NRC chooses to ignore the
issue. I would not hesitate in calling this another case of regulatory complacency. No wonder then that the NRC sends out warnings
against the use of radios insides power plants... Indeed, I question what is the point of the NRC supervising the quality of safety
related instrumentation, if EMC is not a part of their quality criteria (I have statements from NRC authorities to that effect in
correspondence I had with them).

Be advised: if an instrument does not carry a sticker claiming compliance with part 15 of FCC rules, that instrument - by definition
really - may fail under EMI and may cause others to fail because of its emitted EMI. You sure could do yourself a favour by
specifying CE or FCC compliance in your electronic purchases, whether the FCC and the NRC require that or not.

Of course, it would be nice if manufacturers of nuclear instrumentation were required by the law and its representatives to produce
at least to the same degree of quality and compliance as the manufacturers of pocket voltmeters. Certainly some do. Others don't and
get away with that.

As usual: caveat emptor.

/s/

Marco Caceci, Ph.D.
Principal
LQC s.l.
CENT, Parc Tecnològic del Vallès
08290 Cerdanyola, Spain
Tel + 34 93 5820171 fax + 34 93 5820191
ICQ#: 23469430
Netmeeting: uls.ruhr-net.de/cacio@hotmail.com
http://radal.com

my opinions. the facts. sue me.

************************************************************************
The RADSAFE Frequently Asked Questions list, archives and subscription
information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html