ML20216C588
| ML20216C588 | |
| Person / Time | |
|---|---|
| Site: | Beaver Valley |
| Issue date: | 06/19/1987 |
| From: | Lester Tripp NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| Shared Package | |
| ML20216C500 | List: |
| References | |
| 50-334-87-10, NUDOCS 8706300316 | |
| Download: ML20216C588 (6) | |
See also: IR 05000334/1987010
Text
~
1
r
.
!
.
U. S. NUCLEAR REGULATORY COMMISSION
REGION I
Report No.
50-334/87-10
i
,
Docket No.
50-334
Licensee:
Duquesne Light Company
One Oxford Center
301 Grant Street
Pittsburgh, PA 15279
Facility Name: Beaver Valley Power Station, Unit 1
Location:
Shippingport, Pennsylvania
i
Dates:
June 5 - June 17, 1987
4
Inspectors:
F. I. Young, Senior Resident Inspector, BVPS-1
{
S
M. Pi dale, Resident Inspector, BVPS-1
i
nh/
///87
Approved by:
1
.
1. E. Trf pp, Chief, Reactor Projects Section 3A
ate
Inspection Summary:
Inspection No. 50-334/87-10 on June 5 - 17, 1987.
Areas Inspected: Licensee's action to address the concerns associated with the
inoperability of the Chlorine Detection System, required to be operable per Tecn-
nical Specification 3.3.3.7.
This inspection involved 103 inspection hours by two
resident inspectors.
Results: One apparent violation was identified:
failure to comply with Technical Specification 3.3.3.7, which requires three independent channels (trains) of the
chlorine detection system with their alarm and trip setpoints, adjusted to activate
at chlorine concentration of less than or equal to 5 ppm, to be operable in modes
1 through 4.
All three independent channels (trains) of the chlorine detection
system were found to be inoperable since initial installation on September 29, 1980.
The root cause of this apparent violation was the failure to properly implemer;t
the licensee's Quality Assurance Plan which requires, in part, that design control
measures be provided for verifying or checking the adequacy of design changes to
the plant.
l
l
l
!
8706300316 B70622
DR
ADOCK 0500
4
-
-
-
-
-
-
. -
.
-
.
.
DETAILS
1.
Introduction and Overview
On May 4,1987, the licensee determined that the control room chlorine detec-
tion system was incapable of properly detecting the chlorine concentration
j
in the control room.
This system is further divided into three subsystems,
i..e. three channels.
By Technical Specification 3.3.3.7 (Chlorine Detection
Systems), the licensee is required to maintain at least two channels of this
system operable in modes 1 through 4.
The Limiting Condition for Operation
j
(LCO) associated with this Technical Specification further states that if no
{
trains are operable, the ventilation system for the control room is to be
j
placed in the recirculation mode.
On May 25, 1987, the licensee placed the
control room ventilation system in that configuration prior to entry into
i
mode 4.
)
A review and inspection of the chlorine detection system was conducted by the
resident inspectors to determine the safety significance and the status of
1
the chlorine detection system that was installed, and reasons for the failure.
I
of the licensee to properly design and test the chlorine detection system.
This special inspection was conducted between June 5-17 to determine the de-
tails and the root causes of the above.
2.
Background
In response to the Atomic Energy Commission's letter dated November 13, 1974,
the licensee committed to upgrade Beaver Valley Nuclear Power Station, Unit
1, so that the control room would be protected against toxic gases such as
i
At that time, the licensee commenced an engineering design of a
l
system that would protect the control room against certain toxic gases.
One
of the prevalent toxic gases that could possibly be drawn into the control
J
room was chlorine.
This is due to the fact that non-seismic chlorine tanks
are located near the ventilation air intake structure. These chlorine tanks
l
are used in the chlorination of secondary plant systems for biological growth
control.
An engineering study and proposed engineering modification package
.
was initiated and worked under Design Change Package (DCP) No. 156.
This
I
package modified the control room ventilation system to provide an automatic
'
isolation of the control room upon indication of chlorine intake.
To accomp-
lish an automatic isolation, quick response chlorine detectors were located
in the fresh air inlets to the control room.
These detectors were designed
L
to be able to detect a chlorine concentration of a minimum of 5 ppm. Within
i
10 seconds after this concentration is encountered, the detectors were to
i
detect and initiate a complete closure of the control room isolation dampers,
placing the control room ventilation system in the recirculation mode.
In
addition, the DCP added automatic startup features for the control room emer-
gency bottled air system (CREBAPS) to ensure pressurization of the control
,
room with uncontaminated air.
This insured that if there were any leakage
paths in the control room, the leakage flow would be from the control room
to the surrounding environment.
\\
l
]
..
3
.
The licensee installed three chlorine detectors and three blower units, whose
design would take a sample from the ventilation intake airstream of the con-
1
trol room, pass a sample of this air across the detector, and return it back
into the intake airstream.
Each of.the chlorine detectors with its associated
i
blower unit is considered an independent channel and constitutes one of three
'
subsystems of the chlorine detection system.
In 1980, the licensee completed
!
the design, installation and testing of the system.
The Technical Specifica-
!
tions for Beaver Valley Unit I were modified to incorporate the chlorine de-
l
tection system requirements in modes 1 through 4.
The effective date of this
Technical Specification change was September 29, 1980.
3
In May 1987, in support of Unit 1/2 control room modifications, the licensee
)
determined that the Unit 1 chlorine detection system was not functioning pro-
i
perly.
At that time, the licensee commenced an investigation into the root
cause of why the chlorine detection system at Unit 1 would not adequately
respond to known sources introduced into the airstream,
l
3.
Licensee's Findings
i
On May 4, 1987, the licensee discovered that the system was incapable of pro-
perly drawing a sample from the control room intake air duct.
A walkdown and
review of the system by the licensee determined that the sample blowers that
-j
are used to draw a portion of the airstream off to be passed across the
chlorine detectors had inleakage from adjacent spaces.
The design of the
l
blower was such that cooling ventilation air for the motor casing was drawn
i
from the space that the blower was located in.
This cooling air flow that
was drawn into the blower constituted the total flow that was seen by the
individual chlorine detectors.
In addition, further testing by the licensee
!
found that reverse flow existed upstream of the blower and detector and into
)
the main control room ventilation system (also from surrounding area in-
2
leakage via motor cooling holes).
This blower design had existed since final
system testing in 1980.
The licensee concluded, however, that the chlorine
detector eventually would respond to an increase in chlorine that was beir.g
brought into the control room via the air intake vent.
This is because the
adjacent space where the blower is located is also part of the control room
ventilation envelope.
4.
NRC Findings
The inspector reviewed DCP 156 documentation regarding the design and testing
phases of the modification.
System proof test No. T-44A-156-3 also was re-
viewed by the inspector.
The test was used to check proper installation and
to verify that the operability requirements for the chlorine detection system
had been satisfied.
The test used a rotometer (for air flow measurement),
which was installed downstream of each of the individual subsystem blowers.
The flow test results indicated that required acceptance criteria were met.
However, the actual flow verified was not control. room ventilation system
sample flow.
Instead, it was the flow that was drawn into the chlorine de-
tection system, primarily through the blower motor cooling holes from the
ventilation equipment room.
Additionally, no mechanism was provided during
. . .
.
i
I
.
4
..
i
l
.
the flow test to detect flow direction; the air flow upstream of the blower
was reversed, as identified in the licensee's recent system testing.
The
{
source of this reversed air flow was also from the ventilation equipment. room,
drawn into the blower motor and eventually into the control room ventilation
3
system ducting.
The post modification flow testing of the chlorine detection
. system appeared to be inadequate in that the system design deficiencies were
l
not identified.
Technical Specification 3.3.3.7 requires that the chlorine detection system
shall be operable when the reactor is in modes 1 through 4.
The design in-
j
adequacy was discovered on May 4, 1987 with the plant in mode 5.
The licensee
{
immediately began the DCP process to replace the existing control room chlor-
!
ine detection system with a new system, similar to that which Unit 2 uses.
!
DCP No. 814, Control Room Detectors, was developed and approved on May 12,
1987.
The new detectors have been located in the existing ventilation and
equipment room, and all equipment for the existing chlorine detection system
.
has been removed.
The new in-line detector probes have been mounted on the
i
air intake duct and the inputs to the Solid State Protection System and
benchboard controls have remained the same.
Although installation, testing
and turnover of the new chlorine detection system was not completed prior to
entry into mode 4, the provisions of TS 3.0.4 (Entry Into An Operational Mode)
)
did not apply. Therefore, the licensee placed the control room ventilation
'
system in the recirculation mode of operation as required by the Action
Statement of TS 3,3.3.7 on May 25, 1987.
During control room walkthroughs,
the inspectors routinely verified that the control room ventilation system
was operating in the recirculation mode.
On June 16, 1987, the turnover pro-
cess for DCP 814 was completed, and the new chlorine detection system was
j
operationally acceptd by the Operations Group.
The control room ventilation
j
system was subsequently placed in its normal operating made on June 17, 1987.
5.
Summary and Conclusion
/
This inspection determined that the licensee was in apparent violation of
Technical Specification 3.3.3.7 (Chlorine Detection Systems) for periods when
the plant was in modes 1 - 4 from September 29, 1980 through May 4, 1987.
From discussions with licensee personnel and review of Licensee Event Report
87-10, the licensee confirmed that the plant was in apparent violation of the
associated Technical Specification.
TS 3.3.3.7 requires that two out of three
channels (trains) in the chlorine detection system must be operable when the
plant is in modes 1 through 4.
As previously stated, all three channels
(trains) of the chlorine detection system were considered to be inoperable
for this entire period because the chlorine detection system could not perform
its intended function as described in the licensee's Design Change Package
No. 156 or the Technical Specifications.
Review of the licensee's design change process in effect for the time period
1974 to 1980 indicated that weaknesses existed in the area of engineering
review and post-modification testing.
This design change package did receive
reviews subsequent to the original design; however, it appears that an in-
4
depth review was not performed on the key components such as the blower unit
q
)
.,.
5
.
or the chlorine detectors'to ensure that they met the original design objec-
'
tives.
During the post-installation testing phase of the system, the licensee
performed discrete-functional testing of key components independently.
The-
post-installation testing did not require that an integrated functional test.
of the entire system be performed.
Due to this accepted test. practice,.a
design error.that was incorporated into this package.was not discovered.
The
inspector concluded that failure to properly install and test.a modification
to the plant was an apparent violation of the licensee's Quality Assurance.
-(QA) Plan.
Section 3.1.4 of the QA Plan requires in part that_ design control
measures shall provide.for verifying or checking the adequacy ofl design'such
as.by the. performance of design reviews or by the performance of a' suitable
testing program.
The licensee's design change program did not properly ensure
these criteria were adequately met.
'
'
~
The significance of the chlorine detection system's inability'to respond as
required. is reduced by the fact that the system inleakage. flow being drawn
into the blower from the adjacent space is part of the control room ventila-
tion air envelope; thus, if chlorine gas would be drawn into the intake duct,
the air in the space where the blower is located would also experience an
increase in chlorine gas. The chlorine detection system would have responded.
eventually; however, the response time would have been delayed and it would
not have met the criteria as set forth in the design change package or the
!
Technical Specifications.
In' addition, the. licensee had a portable chlor.ine
detector available since 1983 to the' control. room operators-for indepen' dent
verification of the concentration of the chlorine gas in the control room.
Discussion'with several operators indicated that on-shift personnel-were aware
]
of how to operate this instrument and would have used.it if chlorine was
'
sensed by. control room personnel.
The licensee took timely corrective ~ action
once the problem had been identified-on May 4, 1987.
The_ inspector considered the potential for this. type of failure to be incor-
)
porated into the plant to be a significant programmatic weakness .in the lic-
-]
ensee's previous design change process.
6.
Exit Interview
The senior resident inspector discussed the inspection scope and findings with
i
licensee management in an exit interview conducted on June 18, 1987.
This
meeting included a discussion of the safety significance as understood by the
NRC of this event.
The following licensee personnel attended the final exit
meeting:
J. D. Sieber, Vice-President, Nuclear
J. O. Crockett, Sr. Manager, Nuclear Operations
K. D. Grada, Manager, Nuclear Safety
N. R. Tonet, General Manager, Engineering and Records
L. R. Freeland, Nuclear Operations Supervisor
D. C. Hunkele, Director, Operations Quality Assurance
4
.
.
6
.
G. S. Sovick, Sr. Licensing Supervisor
F. J. Lipchick, Sr. Licensing Supervisor
S. A. Nass, Supervising Engineer
,
The inspection results as discussed at the meeting, are summarized in the
cover page of this inspection.
The licensee representatives indicated that~
none of the subjects discussed contained proprietary information.
l
J
!
l
\\
I
i
i
!
l
,