34
NuclearPlantJournal.com Nuclear Plant Journal, September-October 2016
Better Plant...
You say hey, we’re seeing this issue.
We’re seeing this problem. Have you
seen this before? And you can get an
instant feedback via a website or search
the knowledge base. You don’t even have
to go to the vendor, necessarily.
You can get a lot of information.
That’s a very knowledgeable and very
important database resource that exists
for us. That’s one of the advantages of
standardizing. We’ve standardized in
other areas. The Rosemount transmitter
is kind of becoming a standard by
default because we don’t have a lot
of choices. We love Rosemount, but
the Appendix B vendors seem to be
shrinking, less and less. Vendors that
are dedicated to supporting nuclear are
getting out of the nuclear business due
to regulations, low volume sales, costly
internal O&M required to maintain an
Appendix B program. The challenge
for us, is maintaining vendors that can
support us and addressing new vendors.
That’s why commercial dedication is
very important to us as we have to address
obsolescence of aging equipment. There
are a lot of vendors that are working on
protection systems and getting them
reviewed and approved by the NRC. We
as an industry do not upgrade protection
systems frequently. Most of our digital
upgrades are like digital recorders, smart
transmitters, control systems, safety/non
safety indication systems, embedded
digital showing up in breaker protection
such as digital trip units, generator
protection type relays, etc. That is really
where our focus is.
5.
Are you seeing the obsolescence
issue in the digital cards? The second
thing is how are you are improvising so
that at least for the next 30 years, you will
not have the obsolescence issue?
As far as obsolescence, we see a
faster obsolescence in the digital. The I/A
series 100 system at Browns Ferry, we’re
going through our project right now.
We’re upgrading to the latest Foxboro
I/A, with the 200 or EVO series. Foxboro
provided a migration strategy bridging
the 100 and 200 series evolution.
Westinghouse has been great in
supporting us with a very old digital
Eagle-21 system. They built us brand-
new Eagle-21 boards for Watts Bar
Unit 2. Those boards are still available and
manufactured at request. Westinghouse is
supporting those after 25+ years. They’ve
bought up spare parts to support our plants
along with Sizewell, Diablo Canyon, and
Turkey Point.
We have a users group that meets
every six months at the PWROG and we
talk about obsolescence, what specific
components are going obsolete, and what
is the path forward. There are parts on the
board that go obsolete that Westinghouse
has to address through commercial
grade dedication. At some point, we’re
going to have to significantly upgrade a
board due to a processor change driving
a software change. You need a vendor
that has a track record or past history of
supporting the users. Now, in commercial
grade dedication, that becomes more of
a problem because you have dedicated
that specific board or multiple boards.
Commercial products get upgraded at a
lot faster rate forcing you to go rededicate
the new revision again. That’s a major
problem because it’s done once, and
you have to stay at that revision. The
commercial vendor may work with you
on how to freeze the hardware board and
associated software for that design. But
at some point, the processors go out of
production driving a new hardware with a
probable software revision. The hardware
portion really drives the obsolescence
in the digital area. Can you get the parts
anymore to manufacture new boards?
Commercial grade dedication has other
problems that we have been seeing more
recently. We’ve got obsolete dedicated
chiller controls at a plant and no spare
parts but we were able to get the last 10
available boards. At some point in the
future, we’re going to have to have a
replacement option which is a significant
safety related upgrade.
6.
Concluding comments.
A structured digital design approach
is important to have within a utility. This
structured controlled design process must
address frontend requirements related
to the multifaceted levels of testing that
occur at different phases of the project.
Our program is a graded approach based
off of IEEE 1012, “IEEE Standard for
System and Software Verification and
Validation” to align with our Nuclear
Quality Assurance Plan.
It’s a graded approach where for
Safety related, we’ll meet the regulatory
requirements. For Non safety upgrades,
we have adapted our approach based upon
different grading and adapted the IEEE
1012, "Standard for Software Verification
and Validation", structured process using
the areas that add value based upon our
25 years of lessons learned. There are
some things in IEEE 1012 that we haven’t
seen the value in it for across our different
categories of grading. We will take pieces
and parts from IEEE 1012 that have
worked for us in the past. We also design
conservatively and get to know our
vendor and their products. We perform
critical digital reviews (CDR) to learn
about our vendors and their products.
I’ve been on five CDRs in the past six
months, looking at our commercial
vendors for non safety applications. A lot
of our vendors have very good structured
development processes.
Some of them have processes
which are better than some Appendix B
processes I’ve seen in the past. Because
they’ve gone to the electronic era, having
parts traceability down to a little surface
mount resistor manufacturing history.
They know exactly where a lot comes
from and this information is automatically
stored in their information systems. But
getting back on things that help us, in our
program, we’re very conservative. We
have digital defense measures that we use
todesignour systemtohelpprotect us from
unforeseen type of failures because some
of the failures will not be predictable such
as failure downscale (i.e., slowly drifting
input). That’s why medium signal select
will select the middle input, even if it was
the one that’s drifting now. It will control
for a few moments, but once it drifts
through the region between the highest
and lowest signals, it’s kicked out and the
previous lowest good signal comes in and
takes control. We try to avoid any type of
bumps for the operators.
From operations feedback, they
love the new systems. It provides less
challenges for them and better plant
response.
Contact: Ron Jarrett, TVA, telephone:
(423) 751-8298 email:
.
