Information Notice No. 96-23: Fires In Emergency Diesel Generator Exciters During Operation Following Undetected Fuse Blowing

                                 UNITED STATES
                         NUCLEAR REGULATORY COMMISSION
                     OFFICE OF NUCLEAR REACTOR REGULATION
                          WASHINGTON, DC  20555-0001

                                April 22, 1996


NRC INFORMATION NOTICE 96-23:  FIRES IN EMERGENCY DIESEL GENERATOR EXCITERS
                               DURING OPERATION FOLLOWING UNDETECTED FUSE      
                               BLOWING


Addressees

All holders of operating licenses or construction permits for nuclear power
reactors.

Purpose

The U.S. Nuclear Regulatory Commission (NRC) is issuing this information
notice to alert addressees to the potential for damage (possibly fire) to
emergency diesel generator (EDG) exciters resulting from sustained high-power
operation after undetected blowing of a secondary fuse.  It is expected that
recipients will review the information for applicability to their facilities
and consider actions, as appropriate, to avoid similar problems.  However,
suggestions contained in this information notice are not NRC requirements;
therefore, no specific action or written response is required.

Description of Circumstances

On September 30, 1994, during refueling at the Wolf Creek Generating Station,
the A train emergency diesel generator was undergoing post maintenance testing
and balancing.  After about one hour of sustained operation above full diesel
generator power as part of a routine prolonged full and above full-power run,
a fire occurred in the main power potential transformer of the static exciter-
voltage regulator (exciter).  The fire was extinguished quickly by deener-
gizing the exciter and using a portable carbon dioxide fire extinguisher. 
After the fire in the EDG-A exciter, the licensee performed electrical checks
on the EDG-B exciter and found no problems.  On October 11, 1994, again after
about an hour of above full power operation of EDG-B, its exciter potential
transformer also caught fire.  This fire was also quickly extinguished by
deenergizing the exciter and using a fire extinguisher.  After each fire, the
licensee found that one of the 100-ampere fuses in the secondary circuits of
the respective exciter potential transformer had blown.  It was later deter-
mined that the fuses had not blown as a result of the fires, but that the
blown fuses were a contributing cause of the fires.

The phase B windings of the potential transformer removed from the exciter in
EDG-A were severely charred, and the primary and secondary cables were
blistered from the terminal lug back several inches.  Some collateral damage
had occurred to the portions of the A- and C-phase windings closest to the
center, B-phase, windings.  The condition of the damaged windings in both
cases was consistent with progressive insulation breakdown caused by sustained 
 

9604170169.                                                            IN 96-23          
                                                            April 22, 1996
                                                            Page 2 of 4       


current well in excess of the secondary winding ampacity.  Even with lower
current in the primary, the primary and secondary windings being wound
together caused thermal damage to the primary insulation with its ultimate
electrical failure as well.  The A-phase windings of the potential transformer
for EDG-B were similarly damaged.

The blown fuse in the exciter for EDG-A was a power amplifier fuse in the
phase C line of the secondary branch circuit supplying one of the power
amplifiers.  The blown fuse in the exciter for EDG-B was the corresponding
fuse in its phase B line.  The licensee determined that the fuses had blown in
each case as the result of manual engine shutdown without exciter shutdown. 
This had occurred with the generators unloaded (output breakers open) at the
end of the EDG test run preceding the prolonged high-power runs during which
the exciter caught fire.  The blown fuses were not detected at the time
because these fuses had no blown-fuse indication.  The fires occurred in both
cases after about an hour of sustained operation above full power.  Because
there was no blown-fuse indication, the normal full and above full-power runs
for routine testing were conducted subsequently without knowing that the fuses
had blown and "single phased" the potential transformers.

Discussion

Single-phasing the secondary windings of the delta-connected potential
transformers, as a result of the blown fuses, left them at about 58 percent of
rated capacity with one phase winding carrying two-thirds of the total load. 
The other two windings were effectively put in series with each other;
together they were now in parallel with the first winding, but with twice the
total impedance and hence, only one-third total load.  This effect is
illustrated in Figure 2 of Attachment 1 to this notice.  Subsequent operation
with the undetected blown fuse for about 1 hour at or above full load severely
overloaded the affected windings.  It is expected that prolonged operation at
any power level significantly above the reduced capacity of the single-phased
potential transformers would eventually cause damage.  However, even with the
imbalanced load, total excitation current demand from the potential
transformer, being within expected limits, was not sufficient to blow either
or both of the other two fuses.  Operation continued until the overloaded
windings overheated and suffered progressive insulation breakdown, internal
short circuiting of the windings, and the resultant fire.

Although an undetected blown fuse for any reason could cause potential
transformer single-phasing and subsequent sustained high-power operation could
lead to fires, the reason for the blown fuses at Wolf Creek revealed another
apparent design deficiency.  In both cases, the exciters were not shut down
automatically when the engines shut down due to mechanical causes not
accompanied by any of the normal electrical signals that would have otherwise
automatically shut down the exciters.  On a normal manual shutdown from the
control panel, the exciter is turned off automatically as part of the normal
shutdown sequence of relay actuations.  Similarly, in the event of one of the
standard alarm conditions that require immediate EDG shutdown, such as low
lube oil pressure, the system automatically shuts down the exciter. .                                                            IN 96-23          
                                                            April 22, 1996
                                                            Page 3 of 4       


However, in both cases at Wolf Creek in which fuses were blown, the engines
had been shut down mechanically in a manner in which there were no attendant
electrical signals to the system to automatically shut down the exciters.  In
the first instance, after it appeared that control of EDG-A may have been lost
upon a failed attempt at a normal shutdown (the unit tried to resume speed due
to an unrelated malfunction), the unit was shut down manually at the engine by
the operator shifting the engine manual control lever to the stop position
which shuts the fuel racks.  In the second event, EDG-B shut itself down due
to a mechanical failure associated with the governor and the mechanical
overspeed trip device, but without an actual overspeed alarm condition
occurring.  In systems equipped with underfrequency and/or voltz-per-hertz
protective features, this contingency is provided for, but the system at Wolf
Creek had no such features.  Therefore, when the engines stopped in both cases
under the circumstances described, there was no electrical signal generated to
effect automatic exciter shutdown.

Engine shutdown without exciter shutdown caused a potential transformer
secondary fuse to blow in each case as follows:  The exciters at Wolf Creek
(and Callaway) are type WNR manufactured by the Applied Products Division of
Westinghouse Electric Corporation (Westinghouse).  Figure 1 of Attachment 1 to
this notice is a simplified functional block diagram of the affected static
exciter-voltage regulator design.  The potential transformers that caught fire
(designated "PT" in Attachment 1) are Model 26616, delta-connected, 3-phase,
step down 4160-Vac to 480-Vac, 45-kVA, 60-hz transformers, manufactured by the
NWL Transformer Company in Bordentown, New Jersey.  The primary windings of
the PT are connected to the 4160-Vac, 3-phase generator output busses.  The PT
secondary windings supply a portion of the generator DC field excitation
through power amplifiers.  A major portion of the generator field excitation
with the generator under load is supplied by current transformers (designated
"CT" in Attachment 1) from the generator output through rectifiers.  However,
with a generator unloaded and its output breaker open, only a minimal amount
of excitation is provided by the current transformers; most of it being
provided by the potential transformers.

Expanded safety parameter display system trace printouts of EDG voltage,
speed, and load for both events showed that as the generator slowed and
produced less output voltage (in both cases the generator was already unloaded
with its output circuit breaker open), the exciter, not having been turned
off, sensed this and demanded more generator field excitation current to
attempt to compensate.  In both cases, the exciters attempted to maintain
voltage for about 20 seconds during engine coastdown, after which voltage
decayed rapidly.  Attempting to maintain voltage with the generator slowing
caused excessive current in the potential transformer secondary (now the sole
source of excitation power with the generator breaker open) which blew one of
the fuses.  With the resultant reduction in generator field excitation,
coupled with the collapsing stator (armature) induced voltage as the machine
slowed, the available current was insufficient to blow either of the remaining
fuses.  Although these Westinghouse exciters are used at the Standard Nuclear
Unit Power Plant System (SNUPPS) plants, Wolf Creek and Callaway, their
general features are not uncommon and other designs including Basler Electric
Type SB (series boost) exciters (without underfrequency or volts-per-hertz .                                                            IN 96-23          
                                                            April 22, 1996
                                                            Page 4 of 4       


protection), are potentially susceptible to problems similar to those
described herein.

The licensee has installed blown fuse indication on the EDG exciter cabinets. 
Prior to the installation of the blown fuse indication, the licensee
established procedures and trained operators to verify the condition of the
potential transformer secondary fuses following all EDG shutdowns,
particularly those in which there are indications (such as no exciter shutdown
light) that the exciter was not turned off automatically.  Procedures and
training also covered the conditions under which the exciter shutdown
pushbutton is to be actuated and under which the alternate power amplifier is
to be selected to maintain operability, e.g., in case an EDG demand signal is
received during the fuse verification, or in case a fuse should blow during
operation with or without an attendant fault or overload condition.

In other designs, such as newer Basler equipment, underfrequency protection is
often available that will independently shut down the exciter upon loss of the
prime mover.  The licensee has installed volts-per-hertz protection to avoid
the conditions in question.  However, EDG exciter systems of other designs
that remain on, either through a system design flaw or malfunction, after
engine mechanical shutdown may behave in a similar manner.

No specific action or written response is required by this information notice. 
If you have any questions regarding this matter, please contact one of the
technical contacts listed below or the appropriate Office of Nuclear Reactor
Regulation (NRR) project manager.


                                          signed by

                                    Dennis M. Crutchfield, Director
                                    Division of Reactor Program Management
                                    Office of Nuclear Reactor Regulation

Technical contacts:  Stephen D. Alexander, NRR
                     (301) 415-2995
                     Internet:sda@nrc.gov

                     John Whittemore, RIV
                     (817) 860-8294
                     Internet:jcw@nrc.gov

Attachments:  
1.  Exciter Block Diagram and Blown Fuse Effect Diagram
2.  List of Recently Issued NRC Information Notices
Note:  Attachment 1 not included in electronic version