Re: US nuke regulators criticized for false safety methods

["dkosloff1" <dkosloff1@email.msn.com>]

Hank and Sandy (beware, long post),

Since I was a supporter of UCS for about nine years and an NRC resident
inspector for 14 years I would like to provide the following information
from personal experience:

The UCS is not a "membership" organization.  The vast majority of people who
contribute to the UCS are "supporters" although probably many of them think
they are "members".   For the first few years that I supported UCS, I
thought I was a member.  Then I looked up "UCS" in the "Encyclopedia of
Associations" to see how many members it had.  Under that category the
encyclopedia stated "Not a membership organization".  When I looked at the
next fundraising letter I received, I noticed that it said "continue your
support" rather than "renew your membership".  The title of the organization
has been misleading for many years.  Based on my direct observations of
issues that UCS has commented on, I believe that Hank was being kind in his
assessment of their tactics.  The UCS participated in the NRC process to
develop the new inspection and enforcement process, and a UCS representative
was quoted as agreeing that elimination of low-severity-level violations
would enhance safety.   Based on my experience as an inspector, I am not so
sure.

Although I agree with the concept of Sandy's following statement, I have to
disagree with some of the details.  Some of the difference in details may be
because of the difference in inspection practices between power reactors and
materials licensees.

>There are many levels of violations.
>Many in fact are derived from self-
> assessments at the plants. No mention is made that the numbers they
> quote (and I don't know over what time frame they were collected and
> reported) that there were an insignificant number that really
> reflected on any issues requiring significant attention. Going back
> to 1982 and using a vertical slice assessment based on that time
> frame is not a valid comparison. Any deviation from "verbatim
> compliance" in the past has been deemed to be a violation, regardless
> of the safety significance.

Going back to 1982 would be dishonest without significant disclaimers,
because the UCS if familiar with the enforcement process and its history.
When I started with the NRC in 1984, the enforcement process was considered
one of many tools available to enhance licensee performance.  There were
five severity levels (SL) , "one" being of the highest safety significance.
There were also "deviations", wherein the licensee did not do something that
they had promised to do, even though it was not required by the regulations.
Inspectors seldom wrote SL-5 violations or deviations, they were more
trouble to everyone than they were worth.  The regional administrator
eventually said he did not want to see any deviations or SL-5 violations.
There was one exception to this, issues that arose from allegation
inspections.   NRC policy was to encourage allegations and be responsive to
the concerns of allegers.  Therefore, if deviations or SL-4 or 5 violations
were discovered during an allegation inspection, it was written.  SL-4
violations were a matter of inspector discretion and, sometimes, discussion
with a section chief.  This informal process eliminated most SL-4
violations.  If a violation was identified by a licensee's self assessment
process or self-identified in its corrective action program it generally was
not even mentioned in an inspection report.  Inspectors used SL-4 violations
to "send a message", usually after several cases of the same problem had
occurred or corrective actions had been ineffective.  A few years later, the
concept became more formal and more cumbersome with the introduction of
"noncited violations" (NCV).   The main change was to require that some NCV
(some SL-4 and a few SL-5) be written about in an inspection report with an
explanation of why it was an NCV.  This had no safety benefit but if
increased the regulatory burden for everybody and placed every inspector at
risk.  This was because every inspector was probably aware of hundreds of
NCVs each month, but it was impossible to write about every one.  So if a
problem initially identified as an NCV later was linked with a more severe
problem, the inspector could be blamed for not writing about the NCV.  At a
resident inspector meeting, I asked someone from the HQ Enforcement Office
about this and he said "Don't tell anyone I told you this, but we know you
cannot write about every NCV."   Over the last five years, SL-4 and 5
violations have been essentially eliminated, even if they are identified by
the inspector and the licensee didn't have a clue that a problem existed.

As for "verbatim compliance", about five years ago the NRC officially
directed all inspectors to cease using the term.  Prior to that, every case
of failure to follow a procedure was not identified as a violation.  A power
reactor has thousands of procedures and hundreds of people using them daily.
Two cases come to mind where I would have written violations for "failure to
follow procedure" or "inadequate procedure".  In each case I would have
considered several factors, including overall quality of procedures,
previous problems, ease of discovery, and corrective actions in progress.
One case would have been where an "inadequate procedure" had set a trap for
an operator to violate a regulation or cause a plant event, and it actually
happened.  Another would have been where I observed someone performing a
procedure and they failed to comply with the procedure.  I also have to
admit that there had been cases where I asked people questions that I
thought might cause them to recognize that they had missed some step.  Also,
if the situation might have had an immediate impact or the error were
significant, I would have mentioned it to licensee personnel, even if it
would have meant that I would have "lost" the opportunity to write a
violation.  That is also a written NRC policy.

Don Kosloff dkosloff1@email.msn.com
2910 Main St. Perry OH 44081




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