Part 21 Report - 1995-065

.
REGION IV  MORNING REPORT     PAGE  2                    MARCH 13, 1995

Licensee/Facility:                     Notification:

Wolf Creek Nuclear Oper. Corp.         MR Number: 4-95-0030
Wolf Creek 1                           Date: 03/13/95
Burlington,Kansas                      RESIDENT INSPECTOR                 
Dockets: 50-482
PWR/W-4-LP                            

Subject: REACTOR TRIP RECOVERY AND ESTIMATED CRITICAL POSITION
ANOMALY   
                                                                         
Reportable Event Number: 28504                         

Discussion:

Following a reactor trip during the performance of reactor trip breaker  
testing on March 8, the licensee completed several corrective actions in 
preparation for reactor startup.  The reactor trip was caused when a     
reactor trip breaker was being closed from the control room using the    
reactor trip switch.  The switch failed to maintain circuit continuity   
and resulted in a reactor trip.  Non-vital bus PA-01 failed to fast      
transfer and two reactor coolant trips were deenergized as a result.     
Corrective actions completed prior to restart included procedure changes 
to prohibit use of the reactor trip switch in the control room to close  
the reactor trip breakers if all rods were not fully inserted and        
preventive maintenance on the PA-01 breakers to improve breaker actuation
time.                                                                    
                                                                         
During the reactor startup on March 12, the reactor was made critical at 
10:36 a.m. (CST).  Criticality was achieved at 56 steps on Control Bank C
and the estimated critical position had predicted criticality at 72 steps
on Bank D, a difference of approximately 1000 pcm.  The licensee had been
plotting a 1/m graph and observed that the reactor would achieve         
criticality earlier than predicted by the ECP.  Following criticality and
stabilization, the licensee decided to insert all control bank control   
rods while evaluation of the anomaly was in progress.  Licensee          
procedures call for an evaluation if criticality is achieved more than   
500 pcm different than the ECP.  The licensee determined that the error  
was due to a reactivity bias in the Babcock and Wilcox software program  
that calculates ECPs and various other core performance related          
functions. Insertion of the same precritical data into a previously used 
program from a different vendor predicted criticality at a position that 
was only 21 pcm different from the actual critical position.  Following  
the evaluation, and based on information from previous physics testing   
and core response, the licensee commenced a reactor startup utilizing the
previously used ECP program, and was critical at 6:02 a.m. (CST) on      
March 13.                                                                

Regional Action:

The Resident Inspectors and Region IV personnel will be following the    
licensee's actions and will review the corrective actions during routine 
inspection followup activities.                                          

Contact:  D. N. Graves               (817)860-8141
          J. Dixon-Herrity           (316)364-8653