Information Notice No. 86-16: Failures to Identify Containment Leakage due to Inadequate Local Testing of BWR Vacuum Relief System Valves

                                                            SSINS No: 6835 
                                                            IN 86-16       

                                UNITED STATES
                            WASHINGTON, DC 20555

                               March 11, 1986

                                   DUE TO INADEQUATE LOCAL TESTING OF BWR 
                                   VACUUM RELIEF SYSTEM VALVES 


All nuclear power reactor facilities holding an operating license (OL) or a 
construction permit (CP) 


This notice is to alert recipients to a potentially significant problem 
involving the failure to conduct adequate local leak rate tests of 
containment isolation valves It is expected that recipients will review 
this information for applicability to their facilities and consider actions, 
if appropriate, to preclude a similar problem occurring at their facilities
However, suggestion contained in this notice do not constitute NRC 
requirements; therefore, no specific action or written response is required

Past Related Correspondence: 

IE Circular 77-11, "Leakage of Containment Isolation Valves with Resilient 
Seals" September 6, 1977 Information Notice 79-26, "Break of Containment 
Integrity", November 5, 1977 Information Notice 85-71, "Containment 
Integrated Leak Rate Tests", August 22, 1985 

Description of Circumstances: 

During containment integrated leak rate testing, three plants had excessive 
leakage associated with the torus-to-reactor-building vacuum breaker valves
In all of these cases, the leakage was not detected by the local leak rate 
test procedure because the valves were not tested with pressure applied in 
the direction assumed for an accident 

Browns Ferry 2 

Browns Ferry Unit 2 conducted a containment integrated leak rate test in 
February 1983 that failed because of an excessive leak rate of about twice 
the allowable limit of 15 percent per day (075La) The leakage path was 
found to be through a flange seal on a valve in the 
torus-to-reactor-building vacuum breaker system This valve (designated FCV 
64-20) is a butterfly valve bolted 


                                                             IN 86-16      
                                                             March 11, 1986 
                                                             Page 2 of 3   

into an 18-inch line connecting directly to the torus The leakage through 
the flange seal was reduced to an acceptable rate by tightening flange 
bolts Local leak rate testing, which is required to be performed every 2 
years, is done by applying pressure between FCV 64-20 and a flapper-type 
check valve that is located on the reactor building side of the butterfly 
valve However, the leaking flange was on the torus side of FCV 64-20 
Consequently, the valve flange was not included in the local testing, but 
was tested only during the integrated testing which is done every 3 to 4 

Peach Bottom 2 

Peach Bottom Unit 2 conducted a containment integrated leak rate test in 
June 1985 that produced an excessive leak rate of about three times the 
allowable limit of 0375 percent per day Most of the leakage was found to 
be going through the stem seal of valve AO-2502B, an air-operated butterfly 
valve located adjacent to the torus in the vacuum breaker line An 
apparently successful local leak rate test performed on this valve prior to 
the integrated test had failed to detect the leakage Local leak rate 
testing is done by applying pressure between valve AO-2502B and the check 
valve located between the reactor building and this valve However, the 
valve stem for AO-2502B is located on the torus side of the valve and, as in 
the Browns Ferry case, this leak path was not subject to the local leak rate 
test pressure 

Duane Arnold 

During a containment integrated leak rate test at Duane Arnold in July 1985,
difficulty was experienced in establishing the test pressure The problem 
was found to be caused by leakage through a hole left by a plug that was 
missing from the body of isolation valve CV4305 This valve was part of the 
torus-to-reactor-building vacuum breaker system and was located on the torus
side of the vacuum breaker line The plug had evidently been removed during 
maintenance conducted on the valve during the same outage as the integrated 
test An apparently successful local leak rate test, conducted on the valve 
after the maintenance, had failed to detect the hole This failure was due 
to the fact that the hole was located on the torus side of the valve disc, 
and the test pressure had been applied to the other side of the valve 


NRC regulations (10 CFR 50, Appendix J, Section III C1) require that local
leak rate test pressure be applied in the same direction as that which would
exist when the valve would be required to perform its safety function, 
unless it can be determined that the results from tests for a pressure 
applied in a different direction will provide equivalent or more 
conservative results Many facilities experience problems in applying this 
rule because of the difficulty of applying a local test pressure for large 
isolation valves connected directly to primary containments After the 
Browns Ferry test failure, TVA identified 14 containment isolation valve 
flanges on each of the Browns Ferry units that were not being tested under 
the local leak rate test procedures then in use After the Peach Bottom 
test, two valves on Unit 2 and five valves on Unit 3 were found to be 
oriented so that the valve stems were not being subjected to local leak rate 
test pressure 

                                                            IN 86-16 
                                                            March 11, 1986 
                                                            Page 3 of 3 

There are modifications and test techniques that can be applied to cause the
local leak rate test to produce "equivalent or more conservative results" 
For example, at Browns Ferry, TVA is committed to solving the valve flange 
problem by installing double seals (gaskets) on the problem flanges Local 
leak rate test pressure can be applied between the seals to produce a local 
test that can be considered equivalent to or more conservative than internal
pressurization This technique may also be used on valve stems that are 
designed to permit double seals In some situations valve stem seals may be 
included in the normally pressurized boundary by turning the valve around 
without reducing the effectiveness of the valve In some cases special test 
devices such as a blank flange may be used to seal the line inboard of the 
inner isolation valve 

No specific action or written response is required by this information 
notice If you have any questions about this matter, please contact the 
Regional Administrator of the appropriate regional office or this office 

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

Technical Contact:  Don Kirkpatrick, IE
                    (301) 492-4510

Attachment: List of Recently Issued IE Information 


Page Last Reviewed/Updated Friday, May 22, 2015