Information Notice No. 87-10, Supplement 1: Potential for Water Hammer during Restart of Residual Heat Removal

                                 UNITED STATES
                         NUCLEAR REGULATORY COMMISSION
                     OFFICE OF NUCLEAR REACTOR REGULATION
                          WASHINGTON, D.C. 20555-0001

                                 May 15, 1997


NRC INFORMATION NOTICE 87-10, Supplement 1:  POTENTIAL FOR WATER HAMMER DURING
                                             RESTART OF RESIDUAL HEAT REMOVAL
                                             PUMPS

Addressees

All holders of operating licenses or construction permits for boiling-water
reactors (BWRs).

Purpose

The U.S. Nuclear Regulatory Commission (NRC) is issuing this supplement to
Information Notice (IN) 87-10 to alert addressees to the continuing potential
for water hammer in the residual heat removal (RHR) system of BWRs during a
design-basis loss-of-coolant accident (LOCA) coincident with a loss of offsite
power (LOOP) if the RHR system is aligned in the suppression pool cooling
(SPC) mode of operation.  This supplement also addresses the increased use of
RHR pumps in the SPC mode due to leaking safety relief valves (SRV).  It is
expected that licensees will review this 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.

Background

Most BWRs are designed with a multifunction RHR system that includes a
suppression pool cooling capability for long-term cooling following a design-
basis accident.  To ensure the design basis-heat removal capacity, BWRs also
have technical specification limits on the temperature of the suppression pool
during normal operation.  

Information Notice 87-10, "Potential for Water Hammer During Restart of
Residual Heat Removal Pumps," dated February 11, 1987, informed licensees that
the RHR system at the Susquehanna facility was susceptible to water hammer
loads that could exceed allowable stresses.  Such loads could result from a
design-basis LOCA coincident with a LOOP if the RHR system is in the SPC mode
of operation.  In these circumstances, the LOOP, subsequent valve realignment,
and large elevation differences may allow portions of the system to drain down
to the suppression pool, leaving voids in the RHR piping.  When the diesel
generator reenergizes the buses in response to the LOOP, the RHR pumps will
start and possibly cause water hammer damage in the voided RHR loop.  

In addition to concerns about possible water hammer, IN 87-10 also identified
that the RHR system was being operated in the suppression pool cooling mode
more than was assumed in

9705130215                                         IN 87-10, Supp. 1
                                                            May 15, 1997
                                                            Page 2 of 4


the design basis.  The Susquehanna design basis for LOCA/LOOP assumes that the
RHR system is aligned in the standby configuration (suppression pool cooling
flow path valves closed) approximately 99 percent of the time and in the
suppression pool cooling mode only 1 percent of the time.  Contrary to this
assumption, the RHR system was aligned in the suppression pool cooling mode
nearly 25 percent of the time during some operating cycles because of the
leaking SRVs.  As an interim corrective action, the licensee modified its
operating procedures to allow only one loop of RHR at a time to be operated in
the suppression pool cooling mode of operation.  In addition, the licensee
planned to revise plant procedures to address restart of an RHR pump if it
trips while operating in the suppression pool cooling mode. 

Although not discussed in this supplement, IN 87-10 also noted that
Susquehanna had identified that the core spray (CS) system may be susceptible
to water hammer damage if a LOOP occurs when the CS system is aligned in the
suppression pool mixing mode of operation.  To compensate, the licensee
prohibited operation of the system in the suppression pool mixing mode except
to perform the required surveillance testing. 

Description of Circumstances  

Experience since IN 87-10 was issued in February 1987 indicates that the
suppression pool cooling mode of RHR is now frequently used at many BWRs
during normal operation to remove heat from leaking SRVs and to maintain the
suppression pool temperature within technical specification limits.  Although
the leakage appears to be more frequent in BWRs with two-stage Target Rock
SRVs, other model SRVs have had similar leakage problems.  Half of the 10 BWRs
with Target Rock two-stage SRVs (FitzPatrick, Hatch, Hope Creek, Limerick, and
Millstone 1) have experienced problems.  The increasing amount of time in the
SPC mode increases the probability that response to a LOCA will require
realignment of RHR from the SPC mode to the low-pressure coolant injection
(LPCI) mode which therefore increases the possibility of water hammer damage. 


Analyses by several licensees have revealed potential problems similar to
those described in IN 87-10. 

Millstone 1

In Licensee Event Report (LER) 96-050-00 (Accession No. 9610160456), dated
October 10, 1996, the licensee submitted its evaluation of a LOCA concurrent
with a LOOP and a loss of direct current (dc) power.  Such a scenario prevents
closure of LPCI suppression pool test return valves (which are open for
suppression pool cooling) and allows the LPCI flow to be diverted from the
core to the suppression pool.  Since the licensee's existing LOCA analysis did
not consider the impact of a failure of the dc bus with the suppression pool
test return valves in the open position, the plant was considered to be in an
unanalyzed condition.  Also, since sections of LPCI piping would be voided at
the time LPCI initiates, there is the potential for damage to the system as a
result of water hammer                        IN 87-10, Supp. 1
                                                            May 15, 1997
                                                            Page 3 of 4 


The licensee also reported that SPC was operated more frequently during
periods in which SRVs were leaking.  For example, during the most recent
period when SRVs were leaking, SPC was operated for approximately 320 hours
during a six-month period.  The average duration of operation was 12 hours,
with both LPCI subsystems operating. 

WNP-2

In LER 93-001-02 (Accession No. 9406130009), dated June 3, 1994, the licensee
reported that water hammer could fail the train of RHR in the SPC mode of
operation if a LOOP occurred coincident with a LOCA when both trains of RHR
are being operated in the SPC mode.  Typically, SPC was operated to limit the
suppression pool temperature increase caused by leaking SRVs.  For example,
operators placed two trains of the RHR system in  the SPC mode on September
30, 1991, for almost 3 hours; on July 6, 1992, for more than 6 hours; and on
July 11, 1992, for more than 2 hours.  

Limerick 

IN 95-47, Revision 1, "Unexpected Opening of a Safety/Relief Valve and
Complications Involving Suppression Pool Cooling Strainer Blockage" (Accession
No. 9511270084), reported that shortly after starting up Limerick Unit 1 from
a refueling outage, elevated  tailpipe temperatures indicated that three SRVs
("F," "M," and "S") were leaking.  SRVs "D" and "L" were also observed to be
leaking at some time during the cycle.  Reactor operation continued from March
1994 until September 1995, except for two short mini-outages.  The licensee
frequently operated one or both trains of SPC in order to remove heat from the
leaking SRVs.

Discussion

In addition to the long-term post-accident running of RHR pumps in the SPC
mode as described in the final safety analysis report (FSAR), running RHR
pumps occasionally for short durations in the SPC mode may be described in the
system design basis for BWRs.  However, experience indicates that some
licensees may be running the RHR pumps in the SPC mode more often than was
assumed in their safety analysis.  Extended use (increased frequency and long
duration) of the RHR system in the SPC mode during normal operation may be
outside the original design-basis analysis (LOCA) assumptions.  
 
For many BWRs, the original design closing speeds of the valves in the
system's cooling/test lines were specified as the standard speed
(12 inches/minute) and not the fast-closing valves such as the LPCI injection
valves.  Because the cooling/test return valves take longer to close than the
LPCI injection valves take to open, there is a potential for the core
injection flow to be diverted to the suppression pool in the event of a LOCA. 
The emergency core cooling system performance analysis does not always
consider the longer closing time of the test line valves since they are
assumed to be normally closed.  As the amount of time that the test valves are
kept open increases, the likelihood that the valves will be open at the time
of an accident increases, thereby increasing the possibility that the LPCI
flow could be diverted to the suppression pool.  This may be an unanalyzed
condition for some BWRs                       IN 87-10, Supp. 1
                                                            May 15, 1997
                                                            Page 4 of 4 


When operating in the SPC mode, the RHR system is more likely to undergo a
water hammer event if there is a loss of station power.  Since the probability
of a water hammer event increases as the amount of time the system is operated
in the SPC mode increases, and the likelihood of damage to the system
increases with the frequency of water hammer events, operating in the SPC mode
more often than assumed in the FSAR may be an unreviewed safety question as
defined in 10 CFR 50.59(a)(2)(i).  In addition, a significant increase in the
amount of time the RHR system is operated may affect the amount and types of
preventive maintenance and monitoring activities that are required to ensure
that it is capable of performing its intended function.  

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.


                                          signed by S.H. Weiss for

                                       Marylee M Slosson, Acting Director
                                       Division of Reactor Program Management
                                       Office of Nuclear Reactor Regulation

Technical contacts:      George Thomas, NRR
                         (301) 415-1814
                         E-mail:  gxt@nrc.gov

                         Charles D. Petrone, NRR
                         (301) 415-1027
                         E-mail:  cdp@nrc.gov