Multi-plant Issue B-56, Control Rods Fail to Fully Insert (Generic Letter No. 81-24)

                               UNITED STATES 
                        NUCLEAR REGULATORY COMMISSION
                           WASHINGTON, D. C. 20555
                                     
                                June 15, 1981

LICENSEES OF ALL GE BWR FACILITIES (EXCEPT HUMBOLDT BAY) 

SUBJECT:  MULTI-PLANT ISSUE B-56, CONTROL RODS FAIL TO FULLY INSERT (Generic
          Letter No. 81-24) 

Gentlemen: 

By letters dated August 29, 1978 and January 10, 1980, you were requested to
provide information on your experience with one or more control rods 
stopping short of the fully inserted position on scram and then settling 
back to notch position "02" or six inches short of full insertion. 

We have evaluate your responses, along with those provided by other BWR 
licensees, and have concluded that the problem of control rods inserting 
only to notch position "02" on scram is attributable to leakage past worn 
stop and drive piston seals. We have determined that this problem is a 
maintenance problem, and is readily correctable by control rod drive 
overhaul at a subsequent outage. We have further concluded that Technical 
Specification requirements regarding control rod operability and shutdown 
margin provide adequate assurance of the capability to place and maintain 
the plant in a safe shutdown condition. Consequently, the failure to fully 
insert events, as described herein, do not represent a significant safety 
issue. 

The enclosed safety evaluation provides the basis for this determination. 

                              Sincerely, 


                              Darrell G. Eisenhut, Director 
                              Division of Licensing 
                              Office of Nuclear Reactor Regulation 

Enclosure:
Safety Evaluation
.

                                UNITED STATES
                        NUCLEAR REGULATORY COMMISSION
                           WASHINGTON, D. C. 20555
                                     
       SAFETY EVALUTATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION
                            DIVISION OF LICENSING
                                June 15, 1981

1.0  Introduction 

     Over the past several years, a number of instances of control rods 
     failing to fully insert on scram have occurred at GE BWRs. These 
     failures to fully insert on scram (FFIS) have involved one or more rods
     stopping short of the fully inserted position (notch position "00") and
     then settling back to six inches short of full insertion (notch 
     position "02"). 

2.0  Causes 

     Both General Electric (GE) and utilities experiencing such problems 
     have identified trhe apparent cause of the problem as leakage past worn 
     stop and drive piston seals which allows scram water to build up in the 
     buffer area of the drive. This results in a hydraulic lock of the 
     drives between notch positions "02" and "00". When a force balance is 
     achieved, the buffer area and the undreside of the drive piston 
     pressures equalize allowing the control rod drive to settle back into 
     notch position "02".

     Orifices near the upper end of the control rod drive (CRD) piston tube 
     are progressively closed by the drive piston as the full rod insertion 
     postion is approached. This slows the CRDs and prevents seal damage as 
     a result of the drive piston slamming into the stop piston. However, in
     the event of excessive leakage past the stop piston seals, the final 
     piston tube buffer orifice cannot pass all of the water during a scram.
     Consequently, the hydraulic lock referred to earlier is created which 
     prevents the final increment of rod insertion. 

     General Electric Company has recommended a revised Control Rod Drive 
     Venting Procedure which required the CRDs to be vented until no air 
     could be detected with a Differential Pressure Cell in the CRD 
     Hydraulic System. By following this procedure, a water hammer condition 
     would not be created by the pressure of excessive air in the buffer 
     area. Such a condition could result in slamming the drive piston into 
     the stop piston resulting in seal degradation. 
.

                                    - 2 -

     An additional cause of seal deterioration has been attributed to high 
     crud levels in reactor coolant during operation. Some BWRs such as 
     Dresden 2 (which has had the highest number of FFIS events) require a 
     pressure breakdown of reactor coolant before purification, with a 
     resultant low volumetric cleanup rate. The high crud levels may inter-
     fere with the tigh tolerances around the stop piston seals causing 
     binding, breakage, and subsequent FFIS events. 

3.0  Evaluation 

     In all FFIS events reported in response to our letters of August 29, 
     1978 and January 10, 1980, control rode which failed to fully insert 
     inserted to notch position "02". In all cases, subsequent manual 
     insertion by the operator resulted in the rods being fully inserted. 

     Even with all rods inserted only to the "02" notch position instead of 
     the "00" fully inserted position, sufficient shutdown margin exists to 
     preclude FFIS from being a safety problem. The addition of reactivity 
     to the core as a result of all rods being withdrawn to six inches short
     of full insertion is less than that of the most reactive rod being 
     stuck in the fully withdrawn position, according to calculations 
     performed by Brookhaven National Laboratories (BNL). 

     Although this problem appears to be a cumulative one (i.e., rods which 
     initially fial to fully insert continue to fail along with additional 
     rods on subsequent scram), it also appears to be a random type failure 
     (as opposed to a failure mechanism which results in "clustered" non-
     insertion of rods such as that which occurred at Browns Ferry 3 on June
     28, 1980). In addition, these failures are time dependent, which allows
     for an increased likelihood of detection and correction of these 
     failures. This deficiency is typically corrected by removing and over-
     hauling the affected CRDs at an outage subsequent to the recognition of
     the FFIS event. 

     Technical Specifications pertaining to the reactivity margin required 
     to be abailable and to the operability of control rods provide adequate
     assurance of the capability to place and maintain the plant in a safe 
     shutdown condition. 

     In summary, the failures of control rods to fully insert as described 
     herein does not present a safety problem since there is negligible 
     effect on reactivity even if all rods should insert to only the "02" 
     position, the capability to manually insert control rods is retained, 
     the overhaul of affected rod drives during subsequent outages rectifies
     the problem, and existing Technical Specification requirements provide 
     adequate assurance of the capability to place and maintain the plant in
     a safe shutdown condition.