United States Nuclear Regulatory Commission - Protecting People and the Environment

Information Notice No. 93-90 Unisolatable Reactor Coolant System Leak Following Repeated Applications of Leak Sealant

                                 UNITED STATES
                         NUCLEAR REGULATORY COMMISSION
                     OFFICE OF NUCLEAR REACTOR REGULATION
                            WASHINGTON, D.C.  20555

                               December 1, 1993


NRC INFORMATION NOTICE 93-90:  UNISOLATABLE REACTOR COOLANT SYSTEM LEAK
                               FOLLOWING REPEATED APPLICATIONS OF LEAK SEALANT


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 an application of an on-line leak sealing
process which substantially degraded the integrity of the reactor coolant
pressure boundary.  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 May 24, 1993, at the Millstone Nuclear Power Station, Unit 2, Northeast
Utilities (the licensee) identified a leak in a body-to-bonnet gasket on
valve 2-CH-442.  This valve is a Velan 2-inch gate valve which is used to
manually isolate the letdown portion of the chemical and volume control system
from the reactor coolant system for maintenance and local leak rate testing of
containment isolation valves.  The valve is upstream of the automatic letdown
isolation valves, and itself cannot be isolated from the reactor coolant
system.  Between June 4, and August 5, 1993, leak sealant was injected
approximately 30 times in an attempt to stop the leakage from the body-to-
bonnet joint.  Five sealant injections were performed by Leak Repairs, Inc.,
and the remainder were performed by Furmanite, Inc.; these activities were
performed under the direction of the licensee.  The injections had marginal
results; after sealing, the valve would remain leak free for various periods
ranging up to 18 days in length and would then start to leak again.

On June 12, 1993, following the 7th leak-seal injection, technicians attempted
to install a body-to-bonnet peripheral clamp to provide a boundary for the
leak-sealant compound.  The clamp could not be installed because of fit-up
problems with the irregularly shaped valve bonnet.  The bonnet had been
installed 90 degrees away from its normal orientation, further complicating



9311190455.

                                                            IN 93-90
                                                            December 1, 1993
                                                            Page 2 of 3


the attempted clamp installation.  Subsequent on-line attempts to seal the
leak included mechanical peening to prevent injection material from extruding
out of the gasket area.  This peening closed the body-to-bonnet joint.

On August 5, 1993, leak sealant was injected to obtain the dry valve surfaces
required for welding a second (hybrid) clamp across the body-to-bonnet joint. 
During the injection, technicians observed that the leakage increased
dramatically, from slight seepage to a 3.1-meter [10-foot] steam jet.  The
technicians also reported that the valve bonnet appeared to lift and that one
body-to-bonnet stud moved.  In response to the increased leakage, the Director
of Millstone, Unit 2, ordered a controlled normal shutdown.  During the
shutdown, the maximum reactor coolant system leakage was 16.3 liters 
[4.3 gallons] per minute.

Discussion

When valve 2-CH-442 was disassembled, it was discovered that one stud was
broken.  Because the body of valve 2-CH-442 is an unisolatable reactor coolant
system pressure boundary, the broken stud changed the character of the
incident from a routine gasket-related failure to a significant structural
failure.  An ABB-CE metallurgical study indicates that the stud broke in
response to loads applied as part of the on-line leak sealing process.  The
most likely potential sources of loads were evaluated.  These likely load
sources are:  (1) drilling, tapping, and injection port installation, 
(2) peening, and (3) injecting.  Subsequent testing performed for the licensee
at the ABB-CE facility indicated that stud loads produced by drilling,
tapping, and injection port installation were moderate.  However, when body-
to-bonnet joints are peened to the point that the edges contact, the adjacent
bolt can be loaded to failure due to the wedging action of the deformed metal.

Inspection revealed that two of the four studs had drill holes in them from
the injection port and clamp installation drilling processes.  The broken stud
and one other stud each had 0.49 centimeter [0.19 inch] diameter holes that
penetrated 0.25 centimeter [0.10 inch] and 0.28 centimeter [0.11 inch] deep,
respectively.  The holes did not appear to contribute to the stud failure. 
The licensee had intended to limit the injection port locations to low stress
zones.  However, the injection ports had actually been drilled in restricted
high-stress zones located near the studs.  Later analysis determined that the
drilling caused very little stress on the studs.

The leak-sealant injection procedure at Millstone permitted some peening
between the body and bonnet to prevent sealant extrusion (it did not allow
peening along the entire perimeter of the valve).  Inspection of valve 
2-CH-442 revealed a significant amount of metal moved by the peening process. 
Peening essentially obscured the body-to-bonnet interface, leaving a groove-
like indentation along the split line, and metal was peened so that it was in
contact with all four bonnet studs.  In addition, chisel-like marks were .

                                                            IN 93-90
                                                            December 1, 1993
                                                            Page 3 of 3


evident around the leak-sealant injection ports where hand peening was
performed.  The licensee determined that the extensive peening was responsible
for the stud failure.

The licensee had made no provision to limit the amount of leak sealant
injected into the valve.  As a result, a total of approximately 2.16 liters
[0.57 gallon] of leak sealant was injected into the body-to-bonnet joint.  

The repeated attempts to seal the valve at Millstone Unit 2 indicated that an
adequate engineering evaluation was not performed.  The licensee evaluation
did not adequately consider the effects of the sealing process and the borated
water on the fasteners.  Also, the evaluation did not adequately consider the
amount or effect of sealant entering the system after repeated injections. 
Further, the evaluation did not adequately consider the operational and safety
consequences of structural failure of the component or the fasteners during
and after the leak-seal attempts.  In addition, management and quality
assurance oversight did not identify the failures to follow procedures, the
failures to adhere to engineering documents and the lack of weight given to
personnel safety considerations.  

Events such as the one discussed above have the potential to cause a loss-of-
coolant accident and to result in personnel injury or death.  This event
illustrates the importance of properly performed engineering and safety
evaluations and the importance of considering occupational safety hazards in
support of on-line leak sealant use.  When ASME Code Class 1 pressure boundary
components are involved, these considerations are especially important to
public safety.  

This information notice requires no specific action or written response.  If
you have any questions about the information in this notice, please contact
one of the technical contacts listed below or the appropriate Office of
Nuclear Reactor Regulation (NRR) project manager.

                                    /s/'d by BKGrimes


                                    Brian K. Grimes, Director
                                    Division of Operating Reactor Support
                                    Office of Nuclear Reactor Regulation

Technical contacts:  Eric J. Benner, NRR
                     (301) 504-1171

                     Geoffrey P. Hornseth, NRR
                     (301) 504-2756

                      Charles G. Hammer, NRR
                      (301) 504-2791

Attachment:  
List of Recently Issued NRC Information Notices
.
Page Last Reviewed/Updated Tuesday, November 12, 2013