Information Notice No. 85-18: Failures of Undervoltage Output Circuit Boards in the Westinghouse-designed Solid State Protection System

                                                       SSINS No.:  6835 
                                                          IN 85-18  

                               UNITED STATES  
                       NUCLEAR REGULATORY COMMISSION 
                    OFFICE OF INSPECTION AND ENFORCEMENT 
                          WASHINGTON, D.C. 20555  

                               March 7, 1985 

Information Notice No. 85-18:   FAILURES OF UNDERVOLTAGE OUTPUT CIRCUIT 
                                   BOARDS IN THE WESTINGHOUSE-DESIGNED SOLID
                                   STATE PROTECTION SYSTEM 

Addressees: 

All Westinghouse-designed pressurized water reactor power facilities holding
an operating license (OL) or a construction permit (CP). 

Purpose: 

This information notice is provided to alert recipients of a potentially 
generic problem pertaining to short-circuit failures of the undervoltage 
(UV) output circuit boards in the Westinghouse-designed solid state 
protection system (SSPS). The failure of one UV output circuit would result 
in the loss of automatic reactor trip redundancy; the unavailability of both 
UV output circuits would result in the loss of the automatic trip function 
of the reactor protection system and hence make the plant susceptible to an 
anticipated transient without scram (ATWS) event. 

It is expected that recipients will review the information contained in this
notice for applicability to their facilities and consider actions, if appro-
priate, to preclude similar failures from occurring at their facilities. 
However, suggestions contained in this notice do not constitute NRC 
requirements; therefore, no specific action or written response is required. 

Description of Circumstances: 

During a 4 1/2-year period, three UV output circuit failures have occurred 
at North Anna Unit 2. As described later, Virginia Electric Power Company 
(VEPCO) attributes these failures to inadequate surveillance, maintenance, 
or modification procedures for the reactor trip breaker (RTB) switchgear 
cubicle associated with the failed UV output circuit. The first failure 
occurred in April of 1980, the second in May of 1983 and the third in 
October of 1984. The first failure was detected and repaired before the 
plant received an operating license. The second and third failures were 
detected and repaired during refueling outages w and prior to returning to 
power operation. Post-maintenance testing led to the timely, detection of 
these last two failures. 

In April of 1980, a short-circuit failure was detected in the B train UV 
output circuit board at North Anna Unit 2. The Unit 2 maintenance history 
for 1980 is no longer available, thus VEPCO cannot state with certainty that 
this failure was maintenance related. However, because the UV coil for RTB-A 
was replaced 

8503050481 
.

                                                           IN 85-18 
                                                           March 7, 1985 
                                                           Page 2 of 3 

April 3, 1980 just before the B train UV output circuit failure, VEPCO 
indicated that it is likely that some surveillance was performed on the B 
train after the A train UV coil was replaced. 

On May 13, 1983, a short-circuit failure was detected on the B train UV 
output circuit board at North Anna Unit 2. In this instance, the failure was
preceded by maintenance activities on the bypass breaker. for RTB-A. Because
the UV trip coil for the bypass breaker for RTB-A is powered from the B 
train UV output circuit, VEPCO has attributed this failure to a maintenance 
action that inadvertently short-circuited the B train UV output circuit. 

On October 29, 1984, another short-circuit failure was detected on a UV 
output circuit board at North Anna Unit 2. In this instance, the failure was
preceded by the installation of the shunt trip relay, the relay that 
actuates the automatic shunt trip feature of the RTBs; therefore, VEPCO 
attributes this failure to inadequate plant procedures for implementing the 
automatic shunt trip feature of the RTBs. 

On January 12, 1985, while the plant was operating at approximately 96 
percent power, RTB-A failed to open automatically on a valid reactor trip 
signal at Sequoyah Unit 2; however, RTB-B opened automatically, causing the 
control rods to insert. The failure was recognized immediately by the oper-
ator who then followed up the automatic scram demand with a manual scram 
which opened RTB-B. This failure to trip was the direct result of a pre-
existing short-circuit failure on the A train UV output circuit. Sequoyah's 
staff initially believed that maintenance had not been performed on external
components associated with the A train UV output circuit prior to this 
event. The failure, therefore, was originally considered to have been a 
random failure rather than a maintenance related failure. However, a 
subsequent investigation revealed that surveillance activities had been 
performed on the UV trip attachment of RTB-A prior to the event. As a 
result, Sequoyah management now believes that a multimeter was being used to 
check the voltage across the UV coil of RTB-A, and that the meter was set on 
a current setting rather than on a voltage setting, thereby shorting out the 
A train UV output circuit. In this regard, the Sequoyah failure appears to 
be similar to those at North Anna. 

Discussion: 

The UV output circuit (see Attachment 1, UV Output Circuit) is designed so 
that in the absence of a reactor trip signal from the universal logic boards
or switched inputs, the Darlington pair transistors Q3 and Q4 in the UV out-
put circuit will conduct current and energize the UV trip coil of the assoc-
iated RTB. A reactor trip input will result in the turning off of the Dar-
lington pair transistors Q3 and Q4, thereby interrupting current flow to the
UV trip coil of the associated RTB and causing the RTB to open. Similarly, 
if the automatic shunt trip feature has been implemented, the turning off of
transistors Q3 and Q4 would also deenergize the associated shunt trip relay,
thereby tripping the same RTB by a diverse mechanism. 

Failures of the type experienced at North Anna and Sequoyah have resulted in
the shorting of transistors Q3 or Q4. Such short-circuit failures are not 
only 

.

                                                           IN 85-18  
                                                           March 7, 1985  
                                                           Page 3 of 3 

not fail-safe (i.e., they do not cause a reactor trip) but they also would 
prevent the functioning of any subsequent automatic reactor trip signal 
because the failure would continuously maintain power to the UV trip coil of
the associated RTB. Since the Darlington pair transistors Q3 and Q4 are also
being used to control the automatic shunt trip feature of the RTBs, failures
of the type experienced at North Anna and Sequoyah would also incapacitate 
the automatic shunt trip feature of the associated RTB. Thus, the coincident
failures of both the A and B train UV output circuits would disable both 
automatic trip functions (UV trip and shunt trip) of both RTBs, thereby 
exposing the plant to a potential ATWS event. 

Because the UV output circuits are not continuously monitored for failures 
and because each UV output circuit is functionally tested on a 60-day cycle,
one of the two redundant UV output circuits could be inoperable for as long 
as 60 days before the failure was detected. Thus, although the probability 
of an ATWS event occurring due to failed UV output circuit boards is very 
low, the most likely scenario for such an event appears to be one where a 
reactor trip demand occurs while one SSPS train is being tested and the 
other train has a shortcircuited UV output circuit board. For example, if 
the Sequoyah Unit 2 reactor trip of January 12, 1985 had occurred while the 
B train SSPS was being tested, an automatic reactor trip would not have 
occurred. 

The described failures appear to be associated with maintenance activities 
external to the SSPS, including the implementing of the automatic shunt trip
features of the RTBs. To preclude potential failures of the UV output 
circuit boards while implementing the shunt trip feature, Westinghouse 
informed North Anna that it is imperative that the SSPS UV output circuit 
card be removed before performing such work. Prudent actions also would 
include testing the UV output circuit by using the SSPS semiautomatic testor 
after any maintenance work is performed on the associated RTB cubicle. 

No specific action or written response is required by this information 
notice; however, if you have any questions regarding this notice, please 
contact the Regional Administrator of the appropriate NRC regional office or
the technical contacts listed below. 


                                   Edward L. Jordan, Director 
                                   Division of Emergency Preparedness 
                                     and Engineering Response  
                                   Office of Inspection and Enforcement 

Technical Contacts: I. Villalva, IE     N. Merriweather, RII 
                    (301) 492-9007      (404) 221-5577 

Attachments: 
1.   UV Output Circuit 
2.   List of Recently Issued IE Information Notices