Part 21 Report - 1995-074

ACCESSION #:  9412120191
                       LICENSEE EVENT REPORT (LER)

FACILITY NAME:  HOPE CREEK GENERATING STATION             PAGE: 1 OF 5

DOCKET NUMBER:  05000354

TITLE:  Single condition which could have affected the operation
        of multiple systems - failure of cooling water isolation
        valves for diesel generators to stroke open

EVENT DATE:  11/10/94   LER #:  94-017-00   REPORT DATE:  12/08/94

OTHER FACILITIES INVOLVED:                          DOCKET NO:  05000

OPERATING MODE:  1   POWER LEVEL:  100

THIS REPORT IS SUBMITTED PURSUANT TO THE REQUIREMENTS OF 10 CFR
SECTION:
50.73(a)(2)(v)

LICENSEE CONTACT FOR THIS LER:
NAME:  Lou Aversa, Senior Staff             TELEPHONE:  (609) 339-3386
       Engineer - Technical

COMPONENT FAILURE DESCRIPTION:
CAUSE:      SYSTEM:       COMPONENT:       MANUFACTURER:
REPORTABLE NPRDS:

SUPPLEMENTAL REPORT EXPECTED:  NO

ABSTRACT:

On Saturday, October 22, 1994, during performance of a quarterly In
Service Test of the Safety Auxiliary Cooling System (SACS) two room
cooler isolation valves failed to stroke open.  The valves were
associated with the diesel room coolers for the "B" and "D" diesel
generators.  The valves are air operated and designed to fail open
on loss of air to the actuator.  The valves were mechanically agitated
and subsequently stroked open.  The air supply to these two isolation
valves was tagged closed and the valves were left in the open position
to ensure the safety function to provide cooling in the diesel generator
room was satisfied.  System Engineering was contacted and informed of the
failure as this was a repeat of a similar condition identified in
September of 1993.  System Engineering immediately initiated a root cause
investigation to determine the cause of the stroke failures.  On November
10, 1994, the investigation determined that a common mode failure existed
and that the condition was reportable.  The valve stroke frequency was
increased to weekly from quarterly because at this frequency no failures
have been noted.  The root cause of this event is design deficiency.  All
valves will be stroked weekly until the packing configuration is
corrected.

END OF ABSTRACT

TEXT                                                          PAGE 2 OF 5

PLANT AND SYSTEM IDENTIFICATION

General Electric - Boiling Water Reactor (BWR/4)
Safety Auxiliary Cooling System (EG) EIIS IDENTIFIER (CC)

IDENTIFICATION OF OCCURRENCE

TITLE (4): Single condition which could have affected the operation of
multiple safety systems - failure of cooling water isolation valves for
diesel generators to stroke open.

Discovery Date: 11/10/94
Event Dates: 10/22/94
Event Time: 1427
This LER was initiated by Incident Report No. 94-185

CONDITIONS PRIOR TO OCCURRENCE

Plant in OPERATIONAL CONDITION 1 (Power Operation)
Reactor Power 100% of rated, 1109 MWe

DESCRIPTION OF OCCURRENCE

On Saturday, October 22, 1994, during performance of a quarterly In
Service Test of the Safety Auxiliary Cooling System (SACS) two room
cooler isolation valves failed to stroke open.  The valves were
associated with the diesel room coolers for the "B" and "D" diesel
generators.  The valves are air operated and designed to fail open on
loss of air to the actuator.  The valves were mechanically agitated and
subsequently stroked open.  The air supply to these two isolation valves
was tagged closed and the valves were left in the open position to ensure
the safety function to provide cooling in the diesel generator room was
satisfied.  System Engineering was contacted and informed of the failure
as this was a repeat of a similar condition identified in September of
1993.  System Engineering immediately initiated a root cause
investigation to determine the cause of the stroke failures.  On November
10, 1994, the investigation determined that a common mode failure existed
and that the condition was reportable.  The valve stroke frequency was
increased to weekly from quarterly because at this frequency no valve
stroke failures have been noted.

ANALYSIS OF OCCURRENCE

The Safety Auxiliaries Cooling System (SACS) provides all station cooling
requirements for Emergency Core Cooling Systems (ECCS) and Diesel
Generator Systems.  The SACS system consists of two independent redundant
loops each of which are normally lined up to provide cooling to
associated safety related equipment.  Selected safety related loads can
be manually transferred to an operable loop if one loop is

TEXT                                                          PAGE 3 OF 5

ANALYSIS OF OCCURRENCE

inoperable.  Redundant coolers and fans are provided in each ECCS pump
room and diesel generator room.  The cooling water isolation valves are
pneumatically operated, fail open type gate valves.  A solenoid valve
operates an internal three way pilot which repositions to vent air from
the actuator upon receipt of a valid open signal.  All valves involving
stroke failures are Anchor-Darling flex wedge gate valves utilizing
Hiller pneumatic actuators.

The failure experienced during this event is similar to that identified
in a previous License Event Report (LER 93-006-00).  The analysis
performed for the previous event attributed the failure to excessive air
pressure being supplied to the actuator causing the valve disk to bind
in the seat.  This condition resulted when the packing style was changed
which reduced the packing drag on the valve stem.  The design change
package did not account for the reduced packing drag and did not lower
the air supply pressure to the actuator.  The additional seating force
and gate travel reduced the ability of the spring in the actuator to
drive the valve open.  This condition was alleviated by stroking the
valves on a weekly basis rather than quarterly stroking frequency
specified under the In Service Test program.  During the period of the
initial investigation all valves were stroked weekly with no failures
noted.  During the follow-up testing after the air pressure was reduced,
half of the valves were stroked weekly while the remaining valves were
returned to quarterly stroke frequency.  Following two consecutive
successful quarterly valve strokings all valves were returned to the
quarterly test frequency.  During all the initial and follow-up testing
no stroke failures were noted.

Following the failure identified on Saturday, October 22, 1994, System
Engineering initiated a root cause investigation starting with an
inspection of the air supply regulators.  The air supply regulators were
found to be set correctly and were eliminated as the root cause of the
failures.  The valves were then scheduled for testing and disassembly to
inspect for stem galling or mechanical failure.  As a precaution all
valves were returned to the weekly stroke frequency which has resulted in
100% success rate.

The disassembly of the valves did not indicate any abnormalities in the
disk and seating area other than some minor scratches that would be
expected for a flex wedge designed valve.  The stuffing boxes were
unpacked and inspected for packing order, number of packing rings,
presence of lubricant and foreign material, and location of the lantern
ring.  No abnormal conditions were noted.  The stems were inspected and
deemed to be in good condition.  The valve stem diameter, stuffing box
dimensions, and lantern ring were measured with no abnormalities noted.
The valves stroke length and actuator stroke length were measured and
determined to be correct to assure proper actuator spring compression.

TEXT                                                          PAGE 4 OF 5

ANALYSIS OF OCCURRENCE

In conjunction with the above activities the manufacturers for the
actuator (Hiller) and the valve (Anchor Darling) were contacted to
determine if the actuator/valve combination were adequately designed.  A
conference call was conducted with representatives of Anchor Darling and
Hiller to review their findings and it appears that the current design is
adequate for all the valves ( 3", 4" and 6"), although the 3" design
margin is smaller than the 4" and 6" valves.  Anchor-Darling believes
that the current packing configuration is the most probable cause for the
valve stroke failures.

The original valves were supplied with Crane 187I packing.  The packing
was replaced with Chesterton Graphfoil packing.  This replacement was
done on a one for one bases (nine rings of Crane were replaced with nine
rings of Chesterton).  The current industry standard for Graphite type
packing is to use four or five ring configuration, regardless of the
original number of packing rings, with a carbon bushing to make up any
spacing difference.  The standard packing gland torque calculations are
based on a standard five ring configuration, therefore using these
calculations for a large number of packing rings may give non--
conservative results.  Calculations indicate that reducing the number of
packing rings to four or five, can reduce the dynamic packing load by up
to 50% thereby reducing the total force needed by the actuator.

There has been industry experience with graphite type packing that
indicates if a valve is left stationary for long periods of time (this
time is not well defined), that the packing may tend to stick causing
static packing loads to increase substantially.  This increase in static
loads may cause the total load to exceed the capability of the actuator.
Additionally, industry experience has shown that a higher degree of stem
finish is necessarily when using graphite packing.

A contributing factor to the valve stroke failures may be attributed to
the disk friction coefficient used in the original actuator sizing
calculations.  Testing of motor operated valves (MOV) under GL 89-10 has
indicated that the valve friction coefficient used by many valve
manufacturers is generally non-conservative.  For an air operated valve
that relies solely on spring pressure to open, a non-conservative valve
factor could have a significant impact on the actuators ability to open
the valve.

SAFETY SIGNIFICANCE

This event posed minimal safety significance.  Redundant components were
operable to maintain sufficient ECCS equipment operable.

PREVIOUS OCCURRENCES

There has been one previous event reported for this particular condition.
See LER 93-006-00.

TEXT                                                          PAGE 5 OF 5

APPARENT CAUSE OF OCCURRENCE

Based on the initial analysis the most probable cause of the binding is
packing configuration combined with the long period of time between valve
stroking.  Contributing factors may include a non-conservative valve
friction coefficient used in calculating the actuator size, and the style
of packing versus stem finish.

CORRECTIVE ACTIONS

Based on the available information the following corrective action are
being implemented.  As the root cause investigation continues further
corrective actions will be implemented as appropriate:

All valves in the affected population will be stroked on a weekly basis
until the root cause investigation is completed and all corrective
actions to assure proper valve performance are complete.

The packing manufacturer will be contacted to develop a four or five ring
packing configuration with appropriate gland torque values.

Engineering will evaluate the need to improve the finish of the valve
stems to minimize the packing drag.

The actuator manufacturer will be contacted to determine whether a larger
spring is required to increase capability margin.

Diagnostic testing will be utilized to determine actual seating thrusts
and adjust the air supply as necessary to ensure the desired seating
thrust is obtained.

                                        Sincerely,

                                        R.J. Hovey
                                        General Manager -
                                        Hope Creek Operations

SORC Mtg. 94-082
Recommended approval: Yes
C Distribution

ATTACHMENT TO 9412120191                                      PAGE 1 OF 1

     PSE&G

Public Service Electric and Gas Company P.O. Box 236 Hancocks Bridge,
New Jersey 08038

Hope Creek Generating Station

                                                  December 8, 1994

U. S. Nuclear Regulatory Commission
Document Control Desk
Washington, DC 20555

Dear Sir:

HOPE CREEK GENERATING STATION
DOCKET NO. 50-354
UNIT NO. 1
LICENSEE EVENT REPORT 94-017-00

This Licensee Event Report is being submitted pursuant to the
requirements of 10CFR 50.73(a)(2)(v).

                                   Sincerely,

                                   R.J. Hovey
                                   General Manager -
                                   Hope Creek Operations

LAA/

Attachment
SORC Mtg. 94-082
C Distribution

The Energy People

*** END OF DOCUMENT ***