30
NuclearPlantJournal.com Nuclear Plant Journal, March-April 2015
Plant Process...
maintaining the PPC today. Add to this
the limited knowledge of the younger
engineering staff of “Middle Ages”
hardware and software technology and
the often haphazard documentation of
past changes and the recipe for significant
problems is clear. In the near term, the
industry will continue to benefit from
recent retirees acting as consultants.
However, as time passes so will the
availability of their considerable “tribal”
knowledge. It must be anticipated that
the ability of the younger workforce
to maintain current PPC operability be
reduced, perhaps even exponentially,
over the next 5 to 10 years.
Day-to-day plant operations rely
on precise monitoring of the plant
components and processes to maintain
efficiency. An example of the cascading
impact of the loss of tribal knowledge
is seen when the new engineering staff
requires more information from the PPC
to compensate for the loss of “retired”
operating experience.When the PPC is not
able to support that requirement because of
its age and limited resources, engineering
loses the ability to comprehensively track
and trend equipment performance. The
result is missed or delayed surveillances,
delays in equipment troubleshooting and
unrealized degradation in equipment
performance.
What the available data demonstrates
are four separate problems that are so
intimately linked that they are often
considered as a single entity. Each has its
own independent set of issues that must
be managed with different resources and
skill sets to maintain system performance.
They are:
Technological age
Hardware obsolescence
Software support
System reliability
Technological age is not defined
simply by component availability. Its
effects range from the ability to simply
expand or enhance the existing systems
to hard barriers and limits imposed on
the user. A common complaint is that in
order to add some new function, other
important functions need to be taken out
of service.
Hardware obsolescence is well
understood by all utilities. There remains,
however, a host of details that bedevil even
the best supply chain efforts to maintain
the spares inventory. One such detail
is the issue of hardware compatibility
between identical parts and part numbers
that should work together as the original
did, but do not.
Software support goes well beyond
the ability of engineering to support and
read code that may have been developed
more than 30 years ago. It can range
from making new revisions of software
work with old operating systems and
compilers to getting a new work station
to communicate with an older network or
computer to the ability to find or extract
design basis information from poorly
documented source code.
System reliability is the face of the
issue. It can impact everything from
generation and plant performance to the
confidence of the operators in the data
the system is providing. In fact, an all too
often heard complaint was that the age of
the PPC and lack of on line diagnostics
often conspired to mask or at least conceal
less obvious system failures. Perhaps
more importantly, underperforming PPCs
were reported to have complicated the
same events they were designed to help
manage.
PPC Life-Cycle
The hardware, operating systems,
and supporting software in general use
by the industry are robust and reliable. In
most cases, the hardware is designed to
be suitable for industrial applications and
the software applications are purpose-
built for use in the nuclear industry. If
then, reliability and robustness are not
the underlying issue, what is? There are
three common denominators and a host
of lesser complications. Simply stated
they are nuclear commitment, component
aging, and technological growth.
All utilities have suffered from short
term nuclear commitment issues. Vendors
aggressively pursue the short term
opportunities, but lose interest and cease
investment, when additional business
doesn’t materialize. The underlying
problem is typically that either the design
of a product that is so specialized that it
does not have a natural next generation
device or the product is wrapped so
tightly around the specifications that it
becomes a “few and done” design.
Certainly not a surprise, components
age and fail. Capacitor dielectric,
integrated circuit seals, and component
drift over time are but a few of the root
causes of failure. These are aggravated by
power cycling, thermal stress, humidity,
vibration and the occasional electrical
transient that occur every day of use. The
components in the operator workstation,
server, network switch, and I/O system
are not all that different from one another.
What differs is the time or life cycle over
which they are designed to function. A
well designed system uses higher grade
components, possessing, typically, an
extended temperature range and lower
drift and bias ratings. The net result,
along with judicious redundancy, is a
longer effective life cycle.
Technological growth is a particu-
larly difficult issue for a vendor to man-
age. Continuing design innovations de-
mand different supporting components.
The result is a trail of obsolete products
and parts. Another complicating issue is
miniaturization and increased compo-
nent density/functionality. What used to
be testable and repairable by experienced
technicians has now been reduced to re-
placing the offending circuit board. Of
course, that circuit board now also has
a shorter life cycle based on the rapidly
changing component availability.
Exhibit 2 (on page 31) combines
operational records and anecdotal data
from two plants, one shown in red and
one shown in blue, with relatively similar
periods of operation. By the tenth year of
service, operational reliability becomes
an issue.
There is little question that system
failure rates increase with the age of the
system, however, frequency of failure and
cost of failure do not necessarily correlate.
The most notable costs of failure are
during plant power ascension and, where
applicable, during technical specification
mandated Limited Conditions for
Operations that require power reduction.
Time to repair and the resulting cost of
replacement power is the most crucial
visible economic driver.
