ML20198L923
| ML20198L923 | |
| Person / Time | |
|---|---|
| Site: | Pilgrim |
| Issue date: | 10/21/1997 |
| From: | NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| Shared Package | |
| ML20198L921 | List: |
| References | |
| 50-293-97-05, 50-293-97-5, NUDOCS 9710270233 | |
| Download: ML20198L923 (39) | |
See also: IR 05000293/1997005
Text
. . . . - - .. . _ , .- - - . . . . . ,
,
a
e ,_
U.S. NUCLEAR' REGULATORY COMMISSION
REGION l .- -
Docket No. 50 293
Licensee No. DPR 35
Report No.- 97 05
Licensee: Boston Edison Company _
800 Boylston Street
Boston, Massachusetts 02199
i
Facility: Pilgrim Nuclear Power Station :
Location: Plymouth, Massachusetts
Dates: May 14 August 28,1997
Inspectors: L. Prividy, Systems Engineering Branch
B. Higgins, Systems Engineering Branch
'S. K! din, Systems Engineering Branch
G. Morris, Electrical Engineering Branch
Approved by: Eugene M. Kelly, Chief, Systems Engineering Branch
Division of Reactor Projects
.
4
'
J
9710270233 971021
PDR ADOCK 05u00293
, _ e. PDR
4
, _- . .,. # . -, -. _ v -
. _ _ __ ._- __ ___ __. __. . . _ _ _ _ . _ _ _ _
.
.
EXECUhVE SUMMARY
Pilgrim Nuclear Power Station
NRC SpecialInspection Report No. 50 293/97 05
May 14 August 28,1997
Several findings and observations from a BECo service water self assessment (SWSOPI)
conducted in January 1995 were not adequately evaluated or corrected. These issues
represented a pattern of inconsistencies and engineering weaknesses associated with
design control, corrective action and reportability.
Inadeouate Safety Evaluations (eel 97005 01)
- Containment overpressure was credited in 1984 to demonstrate adequate net
positive suction head (NPSH) for the core spray and residual heat removal pumps
when considering the pressure drop due to insulation debris expected on newly-
Installed suction strainers. However, crediting overpressure was not part of the
Pilgrim licensing basis prior to July 3,1997, and therefore, represented an
unreviewed safety question. (Section E1.1)
- Revisions approved on July 18,1997, to residual heat removal (RHR) Operating
Procedure 2.2.19.5, "RHR Modes of Operation for Transients," did not receive
written safety evaluations and yet resulted in changes to reactor building closed
loop cooling water (RBCCW) and RHR operation which differed from that described
in the Pilgrim FSAR and as approved in License Amendment No.173, and represent
potential unreviewed safety questions. (Sections E1.4 and E1.5)
Inconsistencies Between Plant Analyses and Operatina Procedures
Several examples involving the reactor building closed cooling water (RBCCW), residual
heat removal (RHR), and standby power systems were identified where plant analyses or
calculations of record were not consistent with nor properly translated into applicable
operating procedures. (eel 97005-02)
- On several occasions in 1994 and 1995, the Pilgrim Station was operated with
RBCCW inlet temperature greater than the original licensing design basis value of
65 F. The NRC had not approved operation at temperatures above 65*F, which
was the value assumed for containment response in accident analyses, prior to
July 3,1997. (Section E1.2)
- Design basis information regarding isolation of RBCCW system flow to nc,1 essential
loads and the establishment of an appropriate RHR flow rate was not correctly
translated into RHR Operating Procedure 2.2.19.5. (Sectiont E1.4 and 5)
- Design basis information from the emergency diesel generator loading calculation
PS-79 was not correctly translated into Standby AC Power System Operating
Procedure 2.2.8. (Section E1.6)
.
,
i
l
Ineffective Corrective Action
'
l
SeveralInstances from the January 1995 SWSOPl self assessment represent potentially
significant conditions adverse to quality which were not adequately identified, evaluated
and corrected in a timely manner. (eel 97005 03)
- A single failure vulnerability was identified by the NRC which involved the
malfunction of a selector switch for the " swing" (or fifth) salt service water (SSW)
pump. The failure causes both headers to remain cross-tied fed by a single pump in
an unanalyzed potential runout condition, placing the plant outside the design basis.
(Section E1.3)
- The lack of isolation of RBCCW flow to nonessential loads and discrepancies with
the diesel generator loading calculation had been initially identified (but not formally
entered in the PR process) during the January 1995 SWSOPl self assessment, but
had not yet been corrected or substantially addressed. (Sections E1.4, E1.6)
- A long-standing question remained unresolved regarding the impact of ambient air
on diesel generator operation when exceeding the maximum design value of 88 F.
(Section E1.7)
Reportability
Several examples were identified where unanalyzed conditions outside the Pilgrim Station
design bases were not appropriately reported in accordance with the requirements of
10 CFR 50.72 and 73. (eel 97005-04)
- No report was made regsirding exceedir.u the 65'F RBCCW inlet design temperature
on several occasions in 1994 and i995 (Section E1.2).
- An untimely report was made regarding a single failure vulnerability which placed
the SSW system in a configuration with one pump supplying both headers, but in a
runout condition. (Section E1.3)
- No report was made concerning the potential for operation outside the design basis
regarding isolation of RBCCW flow te nonessential loads. (Section E1.4)
- An untimely report was made (and later retracted) regarding the potential for
operation significantly below and above the 5100 gallon per minute design basis
flow rate for RHR in the long-term containment heat removal mode. (Section E1.5).
- For at least a seven-year period of December 1988 through March 1996, Pilgrim
Station was operated in a condition wherein, if a main steam line break were to
occur, drywell temperature profiles could have exceeded established environmental
qualification (EO) limits. Although identified by BECo in early 1996, this condition
(introduced by a modification to reduce drywell spray capability in 1987) was not
reported and apparently would have resulted in containment temperatures
approximately 35 degrees higher than the analyses of record, in apparent violation
of 10 CFR 50.49. (Section E1.8)
lii
TABLE OF CONTENTS
PAGE
EX EC UT IVE SU M M ARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II
El Conduct of Engineering ........................................ 1
E 1.1 Not Positive Suction Head (NPSH) Requirements (UNR 97 0103) ..... 1
E1.2 SSW System Design Inlet Temperature (UNR 9 5 21 -0 2) . . . . . . . . . . . . 5
E1.3 3 alt Service Water System Single Failure . . . . . . . . . . . . . . . . . . . . . . . 9
E1.4 Reactor Dullding Closed Cooling Water Flow Margir. . . . . . . . . . . . . . . 13
E1.5 Design Basis RHR Flow Range ............................. 15
E1.8 Diesel Generator Load Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . 19
E1.7 Diesel Generator Ambient Temperature . . . . . . . . . . . . . . . . . . . . . . . 21
E1.8 Drywell Spray Flow Modifications . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
E1.9 SSW System inservice Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
E1.10 Automatic Starting of SSW and RBCCW Pumps during ECCS Testing . . 28
E1.11 Sea Water Levels Assumed in SSW System Ana'yses . . . . . . . . . . . . . 28
E1.12 Heat Transfer Capability of Safety Related Heat Exchangers . . . . . . . . 29
E1.13 SSW System Piping Through-Wall Leakage . . . . . . . . . . . . . . . . . . . . 31
E 1.14 Pla nt Walk downs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
E7 Ouality Assurance in Engineering Activities . . . . . . . . . . . . . . . . . . . . . . . . . 33
E7.1 Independent Oversight of SWSOPl Corrective Actions ............ 33
E8 Miscellaneous Engineering Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
E8.1 Review of Updated Final Safety Analysis Report (UFS AR)
C o m m i tm e n t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
V. Management Meeting . . . . . . . . . . . . . . . . . . . . . . . . . . ................ 33
X1 Exit Me eting Sum m ary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
iv
- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ - _ _ _ _ _ - _ _ _ _ _ -
.
.
, iknort Details
111. Enaineerina
E1 Conduct of Engineering
Backnroyed
in January 1995 the Boston Edison Company (BECo) conducted a self assessment
of the Pilgrim safety related cooling water systems with the intent to meet the
objectives of NRC Temporary Instruction (TI) 2515/110 " Service Water System
Operational Performance Inspection (SWSOPI)." The NRC monitored these activities
as documented in NRC inspection Report 50 293/95 01. Thermal hydraulic margins
for the cooling water systems woro verified by BECo and subsequently
demonstrated by testing. Over 115 action items were created to followup on the
observations and recommendations from the SWSOPl report that required corrective
action.
The objective of this NRC inspection was to evaluate those corrective actions and
followup activities, including salt service water (SSW) system issues of significance
such as the design inlet temperature question (i.e.,65'F vs. 75'F). The inspection
also addressed existing NRC open items (namely 97-01 03,95 21 02).
E1.1 Not Positive Suction Head (NPSH) Reauirements (UNR 97-01 031
a. lDipaction Sqqpe (92903)
The inspector reviewed DECO nuclear safety evaluations, GE studies, the Pilgrim
UFSAR, and correspondence supporting BECo's use of containment overpressure to
satisfy emergency core cooling system (ECCS) pump NPSH requirements. BECo's
past reliance on crediting (,ontainment overpressure was determined by the NRC to
be an unreviewed safety question (USO), as documented in an attachment to
Inspection Report 97 01 issued on March 18,1997,
b. Observations and FindiDal
1@4 Desian Basu
BECo Nuclear Safety Evaluation (SE) No.1638, Revision 1, dated August 31,1984,
was performed to support changing all piping insulation in the drywell to a flexible
blanket type and the installation of enlarged ECCS suction strainers. The safety
evaluation cover sheet stated that ECCS pump NPSH margin was evaluated "...with
maximum debris loading on the strainer and no overpressure." However, the
licensee later concluded that SE-1638 was incorrect based on discussions with
Gerioral Electric (GE) representatives and in subsequent evaluations and analyses
performed by BECo. The licensee reasoned that results of analysis "...at a peak
supprossion pool temperature of 185*F indicated that the availablo NPSH at this
high temperature with torus overpressuto (calculated with containment sprays)
would be bounded by the NPSH available for a pool temperature of 130*F and a
torus pressure of 14.7 psia." It appears that this was also the rationale used to
_ - _ - - _ - - _ _ _ .-
1
.
2
support the following statements (on ECCS NPSH)in Section 3.0 (page 17) of an
August 1984 GE report (AC-070-0884), " Effects of Fiberglass Insulation Debris on
Pilgrim ECCS Pump Performance":
"The lowest margins calculated, (i.e., the worst " clean screen" conditions)
for NPSH were 15 feet for the RHR pumps and 12 feet for the core spray
pumps, at 130*F pool temperature and atmospheric pressure. These
represent conservative marci ns since no credit is taken for torus
pressurization caused by heating of the air "
In discussiors with cognizant BECo staff and by review of GE written responses to
licensee questions, the inspector ascertained that several iterations of NPSH
calculations occurred in 1984. The GE San Jose design record file (DRFA00-
01713) which supports their study of debris loading effects on ECCS pump
performance was not available for review during this inspection; however, a copy
was being obtained for subsequent verification by the inspector. Nonetheless, GE
correspondence indicates that the final design of the newly installed strainers was
based on a worse caso debris formation (fiberglass) of 23 cubic feet, which was
then predicted to be distributed thrcughout the suppression pool and transported
circumferentially, eventually migrating onto the strainer screens in proportion to the
respective pump flow rates. Corresponding head (pressure) losses through the
shredded debris layer were estimated to be 14 feet for RHR and 6 feet for core
spray; both within the minimum available clean screen" NPSH margins quoted
above.
A key to understanding BECo's design bases and methods for calculating NPSH is
(former) FSAR Figure 14.510, "NPSH Availability for RHR and Core Spray System".
This figure existed in the Pilgrim FSAR for over 25 years, until recent revision.
Based on discussions with BECo staff, review of subsequent BECo safety
evaluations and calculations, RHR pump design information and the GE response to
specific questions posed in 1995, the inspector found that time-dependent NPSH
analyses have always been a part of the Pilgrim design bases. These analyses
explicitly depend upon suppression pool water temperature and containment leak
rate assumptions both ev donce of a consideration of overpressure in determining
NPSH margin. Most signifiaant (and the source of much confusion), the smallest
NPSH margin predicted in tne original (19711984) Pilgrim design was found to
occur relatively late in the containment accident response. FSAR Figure 14.510
illustrates where the limiting margin occurs, independent calculations by the
inspector show this point is at approximately 28 hours3.240741e-4 days <br />0.00778 hours <br />4.62963e-5 weeks <br />1.0654e-5 months <br /> into the event, when the
containment pressure he reiurned to 14.7 psia (atmospheric) and suppression pool
water temperature has been cooled to approximately 130135 degrees.
This is a!ao consistent with eaily GE design guidance (SCER 109) which would
typically evaluate minimum NPSH available for pumps taking suction off the
suppression pool based on 130*F pool water totnperature and atmospheric pressure
ever the pool, Also, the type of pumps provided by GE (e.g., Pilgrim uses single-
stage cast models) provides insight to the design. The inspector noted that this
could easily be miscont, trued at. an assumption of no containment overpressure; to
l
1
- _ _ _
.
.
3
the contrary, it represents the endpoint in a time and temperature-dependent
analysis of NPSH which explicitly requires consideration of subcooling and hence
overpressure. In fact, at the tirne of peak predicted torus water temperature (166'F
originally) approximately 10 foot of margin was shown, about 80% of which is
attributable to containment pressure above atmospheric.
Therefore, while the minimum " clean screen" NPSH margin used in the 1984 design
is referenced at torus conditions of 130'F and 14.7 psia, the torus airspace
pressurization predicted at times earlier in the design basis event was both credited
and necessarv in overcoming the debris loading head lossos. A separate and
independent calculation by the inspector shows that, for torus water temperature
greater than approximately 130'F, containment overpressure (above 14.7 psia) was
required (from August 1984 until April 1997) for the RHR pump to ovwcome the
predicted debris loading on its strainer. Lastly, even at a 130*F/14.7 psia
condition, the subcooled nature of the torus water and consideration of the
"prossure term" in the NPSH calculation still contribute over 12 psi or approximately
29 foot of hood (although not necessarily overpressure) which is factored into the
1984 margin ovaluation.
In a GE Heport (NEDC 30915) performed in 1985 entitled " Pilgrim Decay Heat
Removal Capability", NPSH margins were quoted that are consistent with the 1984
strainer modification. The assumptions and inputs associated with this GE
evaluation woro not consistent with the Pilgrim licensing basis described in UFSAR
Section 14.5 (which was, itself inaccurate). However, with respect to the
consideration of overpressure in the design, the report (Section 5 and Figure 51)
refers to the minimum margin as ".. occurring at a lo'v pool temperature...but as
pool temperature increases, the available NPSH incraases due to wetwell
pressurization". The inspector considered this statement to further corroborate the
use (and necessity) of containment overpressure in the design.
The inspector concluded that a USO was introduced in 1984 but, because SE 1638
was unclear and poorly documented, the explicit credit from that timo forward for
overpressure was not obvious. And, inconsistencies existed (in 1984) between a
number of GE evaluations and several revisions (in April, Juno and August 1984) to
SE 1638. The inspector found that analyses and evaluations at that time were
unclear, and lacked the clarity and rigor nooded to unambiguously conclude what
the design and licensing bases were with respect to NPSH. This was later
corroborated by the independent opinions of outside reviewers engaged by BECo to
review the issue, and ultimately by the NRC in February 1997. Finally, Pilgrim
UFSAR Section 14.5 was never appropriately updated between 1984 and 1996 to
reflect the design during that period, representing an apparent violation of 10 CFR
50.71(c)(4), (eel 9700510)
! 1996 Safetv Evaluations
A now SE 2971 was approved on March 25,1996 to correct SE 1638. This 1996
, evaluation supported the previous replacement of all piping thermalinsulation in the
drywell with Owens Corning NUKON fiberglass blanket, and once again concluded
,
I
l
- .
1. - . ._. __
. . - - - . . - - - _ - _ _ - - - -- -__ - _ - . .
.
.
4
that the 1984 plant modification did not constitute a USQ. Although SE 2971 did
not specifically address the issue of crediting containment overpressure, it refers (in
Section F.4, page 9) to an NPSH margin discussion contained in the original Pilgrim
FSAR. The SE states (original) FSAR Figure 14.513 shows that the containment
remains at or above atmospheric pressure (even in the case of continuous spray).
Still another SE 2983 was approved on March 25,1996, and was characterized as
a " design verification of the plant safety analysis design basis...at 75'F".
However, SE 2983 incorrectly concluded that no unreviewed safety question
existed. The inspector reasoned that this was incorrect because not only was the
use of overpressure (at 65'F) considered to be a USO, but additional amounts of
containment pressure and different assumptions were necessary to justify operation
at 75'F. In a more recent letter (#2.97 004), dated January 20,1997, DECO
requested NRC review and approval for including containment pressure as a
component of NPSH margin in the Pilgrim licensing basis.
ADdenendent Reviewg
The licensee undertook severalindependent reviews of this issue. The inspector
reviewed portions of the following reports summarizing these studies:
1. Once identified by BECo as part of the 1995 SWSOPl self assessment,
Problem Report (PR) 95.9417 was initiated. A BECo Memorandum, dated
March 7,1996, documented the results of a BECo Quality Assurance (QA)
review related to PR 95.9417. That review identified several observations
considered by the OA Department to be " key contributors in allowing the
NPSH condition (viz. the invalid use of overpressure) to go undetected and
create an operability question." For example, the review of vendor design
information (numerous GE evaluations) was cited as a suggested area in
need of assessment.
2. Fauske & Associates performed a review of BECo documentation related to
the design criteria for evaluating NPSH for the RHR and core spray systerns.
The review was performed to assess the technical basis associated with the
available NPSH. The report, dated March 6,1996, concluded that the BECo
evaluations *were performed in a credible manner using available information
in a consistent way."
3. An independent review was also performed by Yankee Atomic Electric
Company, and the results reported in a letter to BECo, dated June 5 1996.
This assessment concluded that containment overpressure was not credited
in the Pilgrim licensing basis. The conclusion was based on the fact that
BECo smendments to the final safety analysis report related to NPSH did not
clearly state reliance on overpressure, and there was no BECo response or
challenge to the AEC's Safety Evaluation which acknowledged that a
positive NPSH margin would be available without requiring overpressure.
.
_
. . . - . _
. . _ _ _ _-.
.
.
5
c. Conclusions
Crediting containment overpressure prior to July 1997 was not in accordance with
the Pilgrim licensing basis, and therefore, constitutoo a USO. This represents the
first of four examples of an apparent violation of 10 CFR 50.59 (eel 97005 01)
The Pilgrim design for ECCS pump NPSH, particularly its reliance upon containment
overpressure, was not adequately described after 1984 in the UFSAR .
The design initially submitted to the NRC by BECo explicitly referred to the presence
of containment pressure in evaluating the NPSH margin avellable, as depicted in
FSAR Figure 14.510. Positive NPSH margin was originally ohown to be available
for the entire course of the event, including points where the containment is
predicted to return to atmospheric pressure (14.7 psla). The NRC, however,
established an originallicensing basis in the August 1971 Safety Evaluation Report,
which was judged acceptable on the basis of margin without requiring containment
overpressure even, as later evaluated, if containment spray was utilized.
A modification to the ECCS pump strainers in 1984 was supported by NPSH
= calculations that required credit for containment overpressure to accommodate
debris loading, but were ambiguously described in terms of where the limiting
design margin occurred. The condition selected (130*F and atmospheric pressure
in the suppression pool) was neither clearly explained nor adequately justifled; the
1984 design was, therefore, not strictly based upon calculations that used
overpressure. However, by extension of the analysis (beyond what's documented
in 1984 evaluations), overpressure was clearly required for the RHR pump at
suppression pool temperatures above 139'F. The statement in BECo evaluations
and GE studies at that time, that this was a " conservative margin", is misleading
and requires speculation to ascertain whether the condition occurs early (i.e., initial
10 20 minuter) or late (beyond 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />) in the design basis event. In either case,
an overpressure is needed for rollable ECCS pump performance between those
timos, particularly at the point of peak torus water temperature. It is notable that,
with Pilgrim operation at SSW temperatures in excess of 65aF, and correspondingly
higher peak torus water temperaturos, the overpressure required for the 1984 1997
strainer design was about twice that currently approved by Amendment No.173.
The NRC has sinco approved a more current licensing basis for Pilgrim in Uconse
Amendment No.173 (NRC letter to BECo, dated July 3,1997), allowing credit for a
specific amount of containment overpressure (1.9 psig from 20-100 minutes, and
2.5 psig thereafter) in determining available NPSH for the core spray and RHR
pumps.
E1.2 SSW System Dosinn inlet Temperature (UNR 95 21 02)
a. Inspection Scope f92903)
in Inspection Report (IR) 95 21, the NRC identified an unresolved item related to
elevated SSW inlet temperatures. The inspector reviewed related correspondence
and DECO Safety Evaluation No. 2983, dated March 25,1996.
_
- _ .
.
.
.
6
b. Observations and Findinas
In February 1990, the NRC questioned the Pilgrim licensing and design bases for the
ultimate heat sink in two regards:
1. The original licensing design basis temperature established for the Pilgrim
2, T% method used to measure and establish SSW inlet temperature during
ns mial plant operation.
The NRC's poaltion, established in February 1996, was that since 65'F was
assumed in UFSAR Chapter 14 accident analyses, that number was considered to
be part of the Pilgrim licensing design basis Consequently, whenever the RBCCW
heat exchanger inlet temperature exceeded 65'F, the plant should have been
considered outside its design basis, in addressing the method of measurement,
DECO's past practice of using a 30 day " rolling average" to establish a temperature
(for comparison against the 65'F design basis) was considered by the NRC to be
inconsistbnt with good engineering practice and, therefore, unacceptable. For
instance, inspection guidance has existed since 1988 (e.g, Manual Chapter
9900,STS 3.7.5) which cautions against the use of a time average to effectively
"... dampen peak values occurring due to higher daytime temperatures or tidal
effects;" otherwise, temperature could significantly exceed that assumed in
accident analyses. Existing (" instantaneous") temperature was more appropriate in
determining if the plant is operating within its design basis. These issues were
discussed with the licensee in a management meeting held on February 22,1996.
Orlainal DosinD_ Considerations
During the February 22,1996 meeting, BECo presented background information
associated with an October 1968 ocean water temperature study which formed the
basis for UFSAR Figure 2.4 2, " Maximum, Minimum, and Mean Temperatures for
Cape Cod Bay and Boston Tide Stations." The licensee argued that a peak
temperature of 75'F and a monthly mean (during August) of 65*F were " carried
forward into the design," and were considered as conservative since the Pilgrim site
l was shown to be approximately five degrees cooler, But, from 1968 until issuance
of the Pilgrim operating license on September 15,1972, the licensee maintained -
that various temperatures (55'F,65'F,75'F) were used to evaluate the RHR
RBCCW and TBCCW heat exchangers (for shutdown cooling, accident and normal
operation, respectively). However, containment heat removal capability was
l analyzed using 65 F, and this was consistently described in the UFSAR.
Nonetheless, the licenseo concluded that 65'F was only a design input, based on
l historical monthly mean seawater temperature, and that "... excursions to
75*F...were identified in' site characterization studies."
Proposed changes to the Pilgrim operating license had been previously considered
by BECo in 1984 and 1985, but were not completed. Similarly, in a review of NRC
Information Notice 87 65, the issue of operating in excess of 65 F was again
_ . _ _ _ _ _ _ _ ____.._._ _ _ __ ._ -_ __ ._ - _ . - _ _ _. __
. l
!
'
.
[
7
,
evaluated by the licensee. The more recent BECo position indicates a i
misunderstanding of the Pilgrim licensing basis. The NRC views the temperatures ,
presented and discussed in UFSAR Section 2 as data used by BECo to support a 65
degree SSW design inlet temperature. Although the historical data presented did
not preclude the use of a higher ternperature, BECo elected to establish 65'F as the ;
licensing and design basis for Pilgrim. Viewing the 65'F temperature as merely a
design input was inconsistent with UFSAR Chapter _14 accident analyses which
had, since original licensing, been reviewed and approved (by the NRC) for 65'F.
Rollina Averaae.g ,
,
By late 1995, BECo justified the operation of the plant above 65'F using what they i
termed a " risk averaged" approach Citing the original site studies of environmental
conditions in the Cape Cod / Boston area (viz. FSAR Chapter 2), the licensee ,
reasoned that exceeding 65'F seawater temperature "did not violate the Pilgrim !
licensing basis...and that all water temperatures within the range of values
contained in UFSAR Section 2.4.2 are within the licensing basis (i.e., .175'F)."
This rationale formed the basis for use of a " rolling average" of SSW inlet
temperatures which is inconsistent with acceptable engineering methods of
establishing an ultimate heat sink design temperature.
In NRC Inspection Report 50 293/95 22 (Section 4.1.3), a chronology of events
related to elevated SSW temperature are presented that demonstrate a
misunderstanding of the Pilgrim licensing bases and a reluctance by BECo to
recognize 65'F as an operating limit, such that the plant had been operating'outside
its NRC approved licensing design basis on several occasions. These events are
,
briefly summarized below:
Qalft Document Eymt
July 7,1994 PR 94.9297 SSW inlet reaches 07'F; Onsite
Review Committee concludes
temperatures up to 75*F.
Feb. 23,1995 Procedure 2.2.32 Administrative limit included in
Revision 35 of procedure to allow
inlet temperaturo up to 75'F.
Aug.3,1995 Engineering memo Approved SSW excursions above
65'F and use of avmge monthly
temperature (" rolling" 30 day
average).
,
,- ,-. ,, . . . - . , - . , - m,...., - ~ . . , c._,-, , . , . , , -
--
,--w, . . , , ... . . - , --., y-m, my,-.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ - _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ - - - _ _ _ _ - ___________ ___-____ ____ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .
.
.
8
Aug.7,1995 Engineering memo Established rolling eight hour
average of 75'F for SSW inlet
temperature in the short term with
specific instructions included in
Operations Standing Order 95 09.
Sept. 4,1995 30 day average exceeded (67'F).
Sept.14,1995 Engineering memo New 30 day " rolling" average limit
established at 67*F.
November 1995 BECo Licensing "PNPS Seawater Temperature
Document Licensing Basis" descrloes a
risk average approach
Qperatinn Procedures
Historically, Cape Code bay water temperatures were known to exceed 65'F during
the months of July September. The licensoo performed a statistical study in March
1996 of SSW temperatures over the 20 year period of 1976 1995, calculating an
average number of hours above 65'F for each year; the mean time for the 20 year
period was 108 hours0.00125 days <br />0.03 hours <br />1.785714e-4 weeks <br />4.1094e-5 months <br /> (above 65'F). The " hottest" summers (based on a maximum
average for 24 hoor periods) occurred in 1981 (69*F),1990 (71.7'F) and 1995
(70.8'F). The worst case recorded period was during the summer of 1995. In the
11 day period of August 17 29,1995,65'F was continucIly exceeded. The peak
recorded value in that period occurred on August 10 at 73.7*F, lasting about 11
hours above 70"F. Several days in that period occurred wherein SSW temperature
remalnad above 70*F for approximately an 810 hour0.00938 days <br />0.225 hours <br />0.00134 weeks <br />3.08205e-4 months <br /> duration.
Notably, no safety evaluation was completed when SSW temperatures exceeded
65'F. As previously discussed in NRC Inspection Report 50 293/95 22, SSW
Operating Procedure 2.2.32 contained an administrative limit in the Section 5.0
precautions and limitations stating that the RBCCW system remains operable with
SSW inlet temperatures up to 75*F. The inspectors considered the lack of a
written safety evaluation to address operation of the RBCCW system at temper-
atures above 65'F to be inconsistent with the Pilgrim licensing basis at that time
(i.e., FSAR 14.5.3) and the second of four examples of an apparent violation of 10
CFR 50.59. (eel 97005 01)
The limit of 75 F was added on February 23,1995 as part of procedura revision
35. However, at that time, no formalized analysis existed to support plant
,
operation above 65'F. Boston Edison Quality Assurance Manual, " Design Control",
requires that engineering and design activities associated with plant design changes
and modifications of nuclear safety related structures...aro accomplished according
I
to ANSI Standard N45.2.11 1974, and BECo Nuclear Engineering Services Group
l Proceduto No. 3.05, " Design Calculations", applies to all safety-related design
analyses.This represents a failure to assure that design basis information (65'F
1: censing design basis SSW inlet temperature) was correctly translated into
_
_.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ - _ _ _ _ _ _ _ _ .
.
l
.
9
operating proceduros (Pilgrim Proceduro 2.2.32) and is one of four examples of an
apparent violation of 10 CFR 50, Appendix B, Critorion Ill, Design Control.
(eel 97005 02)
Moreover, NRC approval of the use of containment overprossure would have boon
required at that time to permit operation up to 75'F (see Section E1.1 above).
Consequently, during the times when Pilgrim Station was operating at SSW
temperatures abovo OS*F, the plant was (until Mar:h 1990) in an unanalyrod
condition with respect to containment heat removal, ECCS pump performance, EDG
electricalload profilos, and environmental qualification that significantly
compromised plant safety and was (until July 3,1997) outsido its design basis, but,
no report was medo to the NRC. This is one of six examples of an apparent
violation of 10 CFR 50.72 and 50.73. (eel 97005 04)
This issue was eventually resolved when, on May 14,1997, BECo requested NRC
approval for Pilgrim operation up to a UHS temperature of 75'F in conjunction with
a licenso amendment to permit crediting containment overpressure for ECCS pump
NPSH requirements. This request was based, in part, on DECO Safety Evaluation
SE-2983 which DECO charactorized as a " design verification for safety related
cooling systems using a 75'F maximum SSW temperature." On July 3,1997, the
NRC approved Amendment No.173, permitting Pilgrim operation up to a UHS
temperature of 75'F.
c. Concludo.na
DECO used " rolling averages" and other relatively informal means (e.g., engineering
memoranda) to address or reconcile operation with SSW abovo 05'F. On numerous
occasions, particularly the suminer of 1995, the Pilgrim Station was operated
outsido the NRC approved design. In contrast, the licensoo established operating
instructions to permit plant operation at inlet temperatures up to 75'F without a
formally approved safety ovaluation.
E1.3 Salt Servico Water System Sinato Failure
a. IDinoction Sconol029321
The 1995 BECo SWSOPl self assessment team questioned the SSW system design
with respect to single activo failures, with an observation that plant personnel did
not always recognize the safety significance of the SSW system pressure switches.
The pressure switches fulfill a safety related function to start standby pumps in the
event that header pressuro is reduced below a sotpoint of 3.3 psig. Separato
pressure switches exist for each of the SSW headers. If the header pressure is
greater than 3.3 psig and increasing, the switch is satisfied, (i.e., additional
automatic SSW pump starts will not nceur).
l
!
Three specific maintenance and test observations were made by the team:
- A recent examplo regarding the plugging of a pressure switch impulse lino
had not been ovaluated for potential common modo f ailure.
l
l
l
1. __ .. . - _. - -- . --. - .-
______ _____________________-__ _ - _ _ _______ ______ _ ___ ____ _ _ ________ _ _ _ _
.
.
10
- Air accumulation in tne CSW pump discharge pressure indicators did not
prompt an evaluation of the potential for air accumulation in the header
pressure switches.
- All consequences of jumpering the SSW pressure switches were not
recognized during maintenance activltios.
At the timo of this inspection, BEco considered any SWSOPI related pressure
switch issues to be closed. The inspector reviewed the various corrective actions
that had been completed to address the above observations. This review included
the function (s) of the pressure switch within the context of overall SSW system
design after postulating cortain single f ailures,
b. Qbservations and Findinas
Based on reviews of plant records and interviews with instrumentation and control
(l&C) personnel who were responsible for maintaining the pressure switches,
adequate correctivo actions were found to have been taken regarding plurging and
venting of the pressure switch impulse lines. Procedure 8.E.29.1, "SSW
Instrumentation Calibration and Functional Test," performed annually to calibrato
the pressure switches, had been revised after the 1995 SWSOPl to formally include
stops to flush and vont the instrument lines. l&C personnel indicated that the past
plant practice had been to flush and vent these lines upon restoring the instrument
to service, even though the procedure did not specifically require it. The inspector
also discussed the impulso line plugging question with soverallicensee personnel
(l&C technician and l&C engineer) who stated that this has been the only known
occurrence of this kind in a pressure switch sensing line, supporting their conclusion
that it was apparently an isolated occurrence.
Desian Bases
The SSW system consists of five pumps, one of which is considered an additional
' swing" pump. The swing pump is arranged to permit operation on either SSW
header according to the position of two normally open DC powered motor operated
valves (MO 3808 and 3813)in a common section of discharge piping. One of
those valvos is designed to close upon a loss of offsite power (LOOP) based on a
preset position of the Loop Select switch, effectively dividing the SSW system into
two independent loops which is consistent with Updated Final Safety Analysis
Report (UFSAR) Section 10.7.5. In a design basis loss of coolant accident (LOCA)
coincident with a LOOP, the emergency diesel generator (EDG) load sequencer
establishes the automatic start of one SSW pump in each loop. The pressure
switches located in each loop would automatically start additional pumps (af ter
established timo delays) if header pressure is less than 3.3 psig. The switches ,
therefore, control sequenced addition of SSW pumps, for EDG load considerations.
. - _-.
_. __ _ __
.
.
11
1995 SWSOPl Guestions
A separate Problem Report 97.9408 was issued on July 1,1997, to establish
whether failures in the DC power system should have been considered in the SSW
system single failure analysis in response to Generic Letter (GL) 8913. The
licensoo initially performed a fallste modes and offects analysis via a
January 11,1990 contract study to address GL 8913. The 1995 SWSOPl team,
however, found several weaknesses associated with the licensee's original
evaluations and, could not conclude that all single active failures had been
appropriately evaluated for the SSW system. While the team independently
examined what it termed "significant vulnerabilit4s" and reasoned that any
additional unidentified failures "wouldn't Jeopardize SSW function," they
nonetheless concluded that BECo's analyses performed at that point were
incomplete. In fact, this specific failure was questioned (MECH-03, Page 3)in the
final SWSOPl report issued on April 25,1995. BECo opened SWSOPl Action item
SW95.003.01 to address this question. BECo later resolved the single failure item
on April 29,1997, as documented in a specir.1 report S&SA9712. However, the
DC power failure in question had not been included in Report S&SA 97-12.
Los9 of DC Power / Selector Switch (Problem Reoorts)
Based on discussions between the inspector and DECO staff during the first two
weeks of the inspection, a system single failuro vulnerability that represented
operation outside the plant design basis was identified. The condition involves a
loss of DC power during a design basis event at specific tidal conditions and Cape
Cod Bay water temperatures. An assumed DC power failure for the train opposite
to which the swing pump is selected to align, results in the expected loss of an
associated EDG and the related SSW pumps. However, both header isolation valves
(MO 3808 and 3813) remain open, and the headers are not split".
BECo issued Problem Report 97.2040 on June 8,1997, to evaluate the problem.
Hydraulic calculations for both a design high tido of + 13.5 ft. (water levelin intake
I
structure above pump suction) and a design low tide of -7.1 ft. were performed
I
using the existing pressure switch setpoint of 3.3 psig. The results for the high tide
case indicated that the automatic sequenced start of a single SSW pump would
alone satisfy the pressure switch. Therefore, both SSW loops would remain cross-
tied with one pump supplying flow to both headers. The resultant flow was higher
l
than normal and beyond the existing pump curve (approximately 5155 gpm), and in
l an apparent runout condition. However, BECo estimated that pump cavitation
would not be expected since sufficient net positive suction head (NPSH: was
, available although the licensee had to extrapolate to make that conclusion, The
l results of the low tide case indicated that a second SSW pump in the same loop
(predicted to be cavitating briefly) would need to be started by the pressure switch
logic after a 30-second time delay, since tho header pressure would remain less
than 3.3 psig. However, BECo recognized that, at other intermediate tide levels (a)
there could be conditions under which one only pump would remain operating, in a
runout condition; (b) with the SSW headers cross-connected; and (c) insufficient
NPSH. The pump would likely experience cavitation until operator action was
I
, _ _ - . .- , - . _ - _ - ,
_- . - .- -- - . .__ .-_- - -- - - -- .- . - . - - .. .. .
!
,
.
'
12
taken. Such cavitation conditions had to be assumed to persist for ten minutes, !
baced on no operator action to start additional pumps or close a header isolation
valve, and would have to be evaluated for impact on SSW pump and system ,
operability,
based on engineering judgment and discussions with the pump vendor, and on the-
adequacy of existing procedures and training, BECo concluded on June 27,1997,
that the SSW system was operable. Disposition of PR97.2040 also relied on the
assumption that operators woutJ recognize and diagnose the failure condition, such
that the headers would be manually isolated and/or an additional pump started soon
after the 10 minute mark. BECo concluded that this conditicn was not reportable,
but also intended to confirm their engineering judgment in this matter by performing '
a special pump test at the postulated runout condition.
Banortability
The single failure deficiency represented ineffectiva corrective action for a problem
first identified during the original 1995 SWSOPl, and for which several opportunities
to resolve it were missed in the subsequent 30 months. The inspectors considered
that this significant condition adverse to quality was the first of five examples of an
apparert violation of 10 CFR 50, Appendix B, Criterion XVI, Corrective Action. ,
(eel 97005 03)
A 10 CFR 50.72 report was made on July 18,1997, stating that the plant had been
operated beyond its design basis in light of this single failure vulnerability. BECo
implemented a temporary modification to shut one of the header isolation valves to
remove the vulnerability. Since the issue regarding system design and reportability
appeared to be sufficiently understood by early June 1997, the inspector considered
that this report was untimely and the second example of an apparent violation of 10
CFR 50.72. BECo later submitted Licensee Event Report 97 01100, " Service
Water System Single Fanure Vulnerability", to the NRC on August 1,1997.
Still ano'her Problern Report 97.9413 was issued on July 3,1997, to clarify an
inconsistency in the plant design information submitted in support of the original
(NRC) SER for Pilgrim. While reviewing the UFSAR and associated documents in
response to PR 97.9408, BECo noted that SER Section 7.4, Salt Service Water
System, stated "On loss of AC power the loops are automatically isolated by
redundant valvss." This would imply that both header isolation valves (MO 3808
and 3813) closo, which differs from the current design where the selective isolation
. of only one valve occurs upon loss of AC power. This is also inconsistent with
~
UFSAR Section 10.7. The inspectors' questioned whether this apparent lack of
redundancy was merely an administrative oversight or a legitimate design
,
deficiency, and this therefore will remain open pending the disposition of
I
PR 97-9413. (URI 97005 06)
_._,,Lm. .,. . . '- ,y y -.- ,m.. ,,., , . - . , ..
_ _ __ . m _ . _ _ . _ __ ___ __ . . .- __ _ . _ _ ___
,
j
.
13 !
!
c. Conclusions
A new SSW system single failure vulnerability was identified representing operation
.
(ptlor to July 1997)in a enndition outside the Pilgrim Station design basis. This
problem was not reported in a timely manner in accordance with 10 CFR w0.72, and
was also viewed as an example of ineffective corrective action associated with
questions first raised during the 1995 SWSOPl. Once the new single failure i
vulnerability was identified, BECo engineering personnel provided a comprehensive .
i
technical response with a bounding analysis of tidal conditions, and took
- appropriate actions to isolate the SSW headers. An inconsistency in the licensing ;
basis for the SSW header independence, specifically the redundancy of the cross-
connect valves, remains unresolved. ;
E1.4 Reactor Buildina Qlosed Coolina Water Flow Marain ;
a. Inspection Scone (92903)
- The 1995 SWSOPl self assessment identified several weaknesses related to the
reactor building closed cooling water RBCCW flow model, including some
components where the model predicted little or no flow margin. The inspector
reviewed the revised flow model and hydraulic analysis (SUDDS/RF 95 74),
Calculation lAs364, " Containment Heat Removal," RECCW Pump Test Procedure
8.1.11.19, and RHR Operating Procedure 2.2.19.5.
b. Observations and Findinas
RBCCW Flow Model
The revised analysis (model) Vedicted flows supplied to various plant components
for several modes of operation including design basis (accident) conditions with
" degraded" RBCCW pumps. The latter case assumed that the pumps performed
32.3% lower than their design (pump curve) values at all flow rates. This reduction
accounts for the degradation permitted by Pilgrim Technical Specification 4.5,
associated with pump surveillance testing.
The 1995 SWSOPl had noted little or no flow margin existed for some components,
especially low flow " users" such as pump seal water and motor oil coolers. The
inspector was concerned that the high resistance presented to the system by the
low flow users may require significantly higher pump head to supply adequate flow
to those components. The hspector confirmed that flow model results were -
aopropriately used as inputs to Calculation No. M-664,' Revision 0, " Containment
Heat Removal" approved on March 25,1996. Calculation M 664 demonstrated
that the RBCCW system is capable of adequate heat removal, assuming SSW inlet
temperatures up to 75"F and using RBCCW flows predicted by the hydraulic
analysis. The inspector noted that the existing RBCCW pump performance exceeds
I
the minimum Technical Specification 4.5 B requirement of 1700 gpm at 70 feet of
- head. This would result in higher actual flows than those used in the containment
i
heat removal analysis and, therefore, additional flow margin. ,
L
i ..
t
_
- w --v-T-ivwr-re*9-gw-,ry--i'9 pgy,-q,,.--gyv-g gy-_ awn-.mg,*wgy -
y,,-gya-9- y 9y-yyir-i- ,g_.. .py- 9 g n_p-. i yf7.p ,
i
l
l
.
14
l
BEco Design Engineering personnel stated that Test Procedure No. 8.l.11.19 did
not account for instrument uncertainties. Further, although some field testing has
been initiated (in response to a SWSOPl observation), testing has not yet been
completed as of this inspection to validate the flow model. The completion of field
testing, v .ildation of the flow model, accounting for instrument uncertainties, and
any additional analysis to quantify flow margin remains unresolved. (URI 97005 07)
bolation of Flow to Non Safetv Related Loads
The 1995 SWSOPl self assessment team noted that the RBCCW hydraulic analysis
assumed that flow to nonessential loads was isoldied during design basis
conditions, but there was no guidance provided to operators to isolate these
components (e.g., in Pilgrim Operating Procedure No. 2.2.19.5, "RHR Modes of
Operation for Transients"). Without specific guidance or instructions to operators,
flows to nonessential loads may not be isolated; consequently, RBCCW could be !
diverted from safety related components. This issue was tracked as an open item
on an informal SWSOPl database (IADB ID SW95.0002.05), but had not yet been
addressed at the time of this inspection. The inspector considered this as a second
example of untimely and inadequate corrective action for a potentially significant
condition adverse to quality, and an apparent violation of 10 CFR Appendix 8, .
Criterion XVI. (eel 97005 03)
There are no formal analyses to dernonstrate or quantify adequate RBCCW flow if
non-essential loads were not isolated. in response to the inspector's questions on
this issue, the licensee revised Procedure No. 2.2.19.5 in July 1997. The inspector
reviewed the approved Revision 2 and found that the procedure change had not
received a written safety evaluation, and was stilllacking in that it would not
require isolation of nonessential luads if suppression pool temperature remained
below 130*F. It was not clear that suppression pool temperatures would remain
below 130 degrees for all design basis scenarios of interest; and, in those
instances, flow to the nonessential loads may still not be isolated. Flow could
therefore still be diverted from so-called "small um " enfety related components,
such as area coolers, RHR pump seal water coolers, cod the core spray pump
bearing cooler which is calculated to recolve 1012 gpm in order to adequately cool
the upper motor thrust bearings. (URI 97005 08)
The inspector concluded that the unisolated nonessential loads constitute a different
case than those analyzed in the RBCCW flow analysis, and is not bounded by the
design basis calculations of record. Similar to the 75'F SSW temperature issue
(Section E1.2), the Boston Edison Quality Assurance Manual, " Design Cnntrol,
requires that enginccring and design activites associated with plant design changes
and modification of nuclear safety-related structuru...are accomplished according to
ANSI N45.2.11 1974, and DECO Nuclear Engineering Services Group Procedure No.
3.05, " Design Calculations", applies to all safety re'ated design analysis, in spite of
the revision to Procedure 2.2.19.5 in July 1997, no formal analyses exist to confirm
that sufficient RBCCW flow will be supplied to the core spray pump bearing cooler
during all design basis conditions. The failure to have an adequata basis for not
isolating nonessential loads as well as to assure that design basis information is
- - - - - -_...- - - - - . - . . - . - . . - ~ - - - - . ._ - _-
'
.
.
15
I
correctly translated into Operating Procedure 2.2.19.5 is the second example of an
apparent violation of 10 CFR Part 50, Appendix B, Criterion Ill, Design Control j
(eel 97005 02).Further compounding this issue, the inspector noted that Pilgrim ;
FSAR Section 10.5.5.3, " Accident and Transient Operations," describes the design !
'
basis operation of RBCCW At 600 seconds, operators manually initiate cooling for
the RHR heat exchanger by starting a second RBCCW pump, valving in flow, and
"
... isolating the nonessential loads on each loop" (page 10.5 3). Therefore, the !
licensing basis of RBCCW described in FSAR Section 10.5.3 lb also inconsistent !
with Operating Procedure 2.2.19.5, and represents the third example of an apparent
violation of 10 CFR 50.59. (eel 97005 01)
c. Conclusions
The evaluation of flows supplied using reduced RBCCW pump capability (32.3%)
adequately addressed the degradation issue raised during the 1995 SWSOPl self !
assessment. The potential for RBCCW flow diversion because of flow supplied to i
nonessential loads represents operation in an unanalyzed cendition outside of the
Pilgrim Station design bases, and represents the third example of an apparent r
'
violation of 10 CFR 50.72 and 73. (eel 97005 04)
E1.5 Desian Basis RHR Flow Ranae
a. JDinaction Scope (92903)
The inspector reviewed Calculation M 664, Revision 0, dated March 25,1996, ,
" Containment Heat Removal," '"-h recalculated the design basis containment heat
transfer and pressure / temperature n...Jonse. Calculation M 664 utilizes input from
.
GE accident analyses, including the suppression pool temperature profile. The
calculation uses a so-called "K" factor to model heat exchanger performance as well
as overall system heat transfer,
b. Observations and Findinag
i
An RHR design flow rate of 5100 gallons per minute (gpm) is assumed in
Calculadon M 664. How:,ver, revisions prior to July 1997 of Pilgrim Operating
-
'
Procedure 2.2.19.5, directed operators to throttle RHR flow in a range of 4800-
5100 gpm for containment cooling during design basis conditions. In addition to
procedurally allowing flow loss than design, the inspector found that neither the
procedure nor the calculation account for instrument uncertainty at either extreme.
Consequently, the actual flow supplied to the RHR heat exchanger could be higher
than 5100 gpm, or lower than 4B00 gpm. There was no formal documented
analysis to confirrn that adequate heat removalis provided at the lower flow, *
Including the effects of instrument error. Further, flows greater than 5100 gpm
!
L
l
!
'
. - - . . - - - - , , - , , , , . - _ . - . . - ,_ -, _ . _ - , - - - ,,
.. _
- _ _= .- .
-- _ _ --.
-
.
.
10
may exceed design limits for the heat exchanger, affecting tube vibration and
structural loading.
Probiern Report 97.9363
in response to these issues, Problem Report (PR) No. 97.9363 was issued which
judged RHR to be operable based on engineering judgement "... demonstrating
conservative assumptions in the containment heat removal aneiysis." The inspector
observed that the initial reportability screening review for PR 97.9363 did not
require an evaluation in accordance with 10 CFR 50.72. The licensee revised
Operating Procedure 2.2.19.5 to stipulate that RHR pump flow not exceed 5000
gpm when suppression pool temperature exceeds 130'F and one loop of
containment cooling is used during torus cooling. The licensee contemporaneously
developed analyses (SUDDS/RF# 97 84, Revision 0, July 14,1997, " Evaluation of
RHR Heat Exchanger at 5600 gpm") to confirm that the RHR heat exchanger can be ;
operated at shell side flow rates up to 5600 gpm without affecting structural
integrity.
On July 18,1997, the licensee notified the NRC that plant operating procedures
specified RHR system operation in the emergency mode that was contrary to that
assumed in design basis accident analyses. However, BECo retracted this
notification on September 12,1997, based on an evaluation (using engineering
judgement) to affirm acceptable RHR performance.
Dulan Bases
The inspectors expressed concern for the effects of flow rates at both ends of the
range. The inspectors considered the basis for the operating range of 4800 5100
gpm to be principally anecdotal, considering normal operation rather than design
basis. The range was selected with consideration of the fact that the
instrumentation used to throttle RHR (full range of 0 20,000 gpm) would be at the
low end of indicated flow. More rigorous evaluation after this inspection --
quantified instrument inaccuracy et 2.8% (of full scale) or approximately .t.560 gpm
in the range of interest.
Using procedures in place prior to July 1997, RHR flow with one pump operation
could therefore have been as low as 4,240 gpm, which is more than 15% below
that assumed in the design basis calculation of record (M-664). Although not
verified during this inspection, the licensee later indicated that the overall K-f actor
utilized in determining containment response would decrease by about 5%,
affecting the predicted torus water temperatures and heat removal from the
containment. While the containment apparently remains operable (bases for
Technical Specification 3.5.8 unchanged at 64 million BTU /hr), the inspectors found
that the design margins would have been reduced. On the upper end, an indicated
flow of 5100 gpm could be as high as 5660 gpm which exceeds structuralintegrity
limits for the RHR heat exchanger. The failure to assure that design basis
information (i.e., appropriate RHR flow range) was correctly translated into
operating procedures (Pilgrim Procedure 2.2.19.5)is the third example of an
-. _ _ _ _ - - - - . - . . -
.
.
17
apparent violation of 10 CFR Part 50, Appendix B, Criterion Ill, Design Control. (eel
97005 02)
Current Licensina Basla
The inspectors also noted that the licensing basis (prior to Amendment No.173)
described in the P;lgrim UFSAR was inconsistent with respect to RHR operating
procedures. Specifically, FSAR Ssetion 4.8.5.4.1, "LPCl with Heat Hejection,"
states (Revicion 20, February 1997, Pago 4.8 0) that rated heet removal from the
containment by removal of one RHR pump from LPCI service, closure of the RHR
heat exchanger bypass valve, and "... valve adjustments as necessary to establish
5100 gpm through the RHR heat exchanger."
Further, in FSAR Chapter 14.5.3, the design basis containment response end core
standby cooling system NPSH analyses are described. Design margins are
determined for maximum suppression pool water temperature (page 14.611) and
RHR pump NPSH required (page 14.512a) based on establisliing a maximum flow
rate of 5100 gpm. Specifically, FSAR Section 14.5.3.1.2, " Containment Response"
(Revision 20, February 1997) describes the transition at two hours after an accident
to a one pump LPCI Heat Rejection Mode to "... provide lated heat removal from the
containment." Similar to other FSAR descriptions, valve adjustments establish
5100 gpm flow through the RHR heat exchanger, and this modo is assumed to
"...run continuously throughout the remainder of the accident response" voth a
peak suppression pool temperature predicted to occur at 5 to 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> af ter the
accident begins. This flow rate is also summarized in FSAR Tobie 14.5 6.
With approval of Amendment No.173 on July 3,1997, the peak suppression pool
water temperature is higher as a result of 75 F SSW and oth-r considerations
including increased decay heat assumptions imposed as a condition to the
amendmont. RHR pump NPSH is calculated based on 185'F pool water
temperature and 5100 gpm; at an increased flow up to 5000 9pm, frictional losses
and required NPSH (extrapolated at 27 feet) would change, thereby decreasing the
margin availablo and placing more importance on the existence of an ovespressure
to compensate for the deficit in suction head. More significantly, following either
Step 9 (page 15 of 26) or Step 10 (page 22 of 26) in Procedure 2.2.19.5
establishes flow at essentially a runout condition and potentially beyond vendor-
certified test data for both flow and required NPSH.
Oneistina Procedures
The inspectors found that revisions to Operating Procedure 2.2.19.5 in effect prior
to July 1997 contained guidanco to operators which was inconsistent with the
design basis calculation (M 664) of record as well as the licensing design basis
described in the Pilgrim FSAR. While steps woro taken to correct the operating
procedure during this inspection, it remains inconsistent with the FSAR as of
October 3,1997. Operators are now instructed to throttle open either the LPCI
(28A/B) or the RHR torus test return valve (36A/B), "...to achieve loop flow at
maximum RHR pump capacity not to exceed 5600 gpm." The licensee contends
. __.
. .
_ -_ _ _________-_ _ -____-___-_- - - __ _ _ _ _ _ _ _ _ _ _ _ - - _ _ _ - _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _
'
.
18 :
that this step will ensure flows in excess of 5100 but loss than 5600 gpm. While
the Inspectors did not review the revised calculations to ensure desired flows
remained between 5100 5600 gpm, the inspectors did note that this represents a
change to the facility as described in FSAR Sections 4.8.5.4 and 14.5.3. Sinco no
safety evaluation was performed to reflect those design changes (i.e., higher system
maximum flow rates hbovo 5100 and up to 5600 gpm), this was considered as the
fourth example of an apparent violation of 10 CFR 50.b9. (eel 97005 01)
Further, the ins;)octors were concerned that if the margins for suppression pool
water temperature or EC/;S pump NPSH associated with Pilgrim Amendment No.
173 were decreased, then this problem may also represent a potential USO.
Amendment No.173 was based upon RHR flow of 5100 gpm which required 23
feet of NPSH and affects head losses, EDG loadings, and strainer pressure drop.
Calculations of record, which support the approved (and current) licensing basis,
show that while the core spray pumps are moro limiting with respect to NPSH, the
'
RHR pumps also require a small amount of containment overpressure (approximately
0.04 psig). With the approval of 2.5 psig of overpressure beyond 100 minutes in
Amendment No.173, the RHR pumps are predicted (and licenced) to have about six
foot of suction head margin at R100 ppm. However, at the increased flow of 5600
gpm this margin is reduced to about four feet but, more significantly, a near runout
condition is created, thereby affecting the long term reliability of the pump and
increasing the likelihood of pump failure. Lastly, considering both the uncertainty
with measured flow and certified pump performance at flows in excess of 55;')
opm, RHR heat exchanger structural integrity and the potential for component
failure is similarly in question.
ReportoMity
Regarding the adequacy of containment heat removal at RHR flows lower than 5100
gpm prior to July 1997, the inspectors considered the potentit.Ily higher peak
suppression pool temperature to be significant. While the FSAR value reported on
, page 14.511 of 178'F was changed after July 3,1997, by Amendment No.173
(and is in fact also affected by assumptions regarding CSW temperature,
containmont overpressure, and ECCS strainer design), the inspectors considered the
lower flow case to be an unanalyzed condition that significantly compromised plant
safety, as well as a facility change outsido the Pilgrim design basis, not
appropriately evaluated in accordance with 10 CFR 50.59.
In reviewing the NRC staffs' safety evaluation associated with Amendment No.173
(Section 3.1.2, page 5 and 6), the peak calculated pool temperature with 65'F
SSW had been 166'F, prior to July 1997. The inspectors concluded that - if 75'F
SSW and credit for overpressure were invalid prior to July 1997 -- the possibility of
RHR flow in the long term heat removal mode with one pump potentially as low as
4240 opm or as high as 5660 gpm (particularly for long periods of time) similairly
represent operation in an unanalyzed condition outside of the design basis.
Therefore, the retraction on September 12,1997 (two months after an ENS report,
but with no LER issued), is the fourth example of an apparent violation of 10 CFR
50.72 and 50.73. (eel 97005-04)
_ , _
_ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ___ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ . _ _ - _ _ - _ _ _ _ _ _ _ _ _
.
.
10
c. Conclusions
Prior to July 1997, Operating Procedure 2.2.19.5 permitted operation of the RHR
system at flows both above and below design basis, without a rigorous and '
formalized analysis of the effects on not only component performance but overall
containment response. Operation at RHR flows substantially higher or lower than
5100 gpm would be outside the Pilgrim design basis, and presents problems to
operators because of f.560 gpm uncertainty in measured RHR flow. This issue
reflects problems rogarding design control, reportability and safety evaluations that
were (and still are) not accurately reflected in the Pilgrim UFSAR.
E1.6 Diesel Generator load Calculations
a. Scoco of insocction
The 1995 SWSOPl self assessment questioned diesel generator (EDG) loading,
assuming various single failures r,uch as the failure of the SWS header pressure
switches or the failure of the EDG load shed circuitry. The inspector reviewed
BECo's response to this ope, item including related calculations and operating
procedures,
b. Observations and Findinos
The inspector reviewod ceiculation PS 79, " Emergency Diesel Generator Loading,"
Rev. 4, dated January 19,1996, and calculation comment sheets outstanding
against PS 79 up to comment sheet PS 7019. The inspector confirmed that the
calculation addressed all the loading combinations questioned during BECo's 1995
self assessment. The inspector confirmed that the calculation resulted in
demonstrating a worst case or smallest margin of 01 kilowatts (kW) below the
2000 hour0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br /> limit (i.e. 2750 kW). This case was based on operator actions (beyond
10 i.11nutes) to remove certain loads and in the event of the failure of a SSW
pressure switch.
During the review of the calculation, the inspector noted two additional deficiencies.
The 250 Volt de battery charger did not include the power drawn during current
limit operation similar to what had been included for the 125 Volt de battery
chargers, in response to the inspector's question, the licensee estimated that this
could amount to an additional 6.4 kW for each 250 Volt battery charger with a
110% current limit. Also, the calculation did not address the effect on generator
load by the motor driven pqmps frequency variation from a nominal 60 Hertz.
Motor speed, and therefore pump speed, varies with frequency; pump speed, in
turn, affects flow and power drawn by the pump. BECo initiated Problem Report
(PR) 97.9541 to address the frequency variation, but did not expect substantial
changen in calculation results based on a tight governor control of diesel speed to
+ /. 0.2 5 E
l
-
e
l
l
_ . - - ~ _ _ . _ . _ _ .. _____ _ _ _ _ _ _. _ _ . _ _ _ _ _ _
'
.
'
, :
20
,
The inspector also noted the cuver sheet to calculation PS 79 indicated the analytls
did not require revision to affected d6 sign documents. BECo acknowledged that l
they did not have a cross reference list of documents affected by this calculation. !
Calculation comment sheet PS 7912 had already identified the fuel capacity
calculation as an affected document prior to this inspection. The inspector further
noted that the cover sheet did not address other documents such as operation and
surveillance procedures.
The inspector comparret the diesel generator loads documented in design basis
calculation PS 79 to Operating Procedure 2.2.8, " Standby AC Power System", Rev.
42, dated August 8,1997. Attachment 5 to this procedure, Diesel Generator
Loa-fing Table, is a twelve page guide uhsd by the operators when they must
remove Partain loads (in the licensing bases beyond 10 minutes). The inspector
found eiumerous discrepancies betw9en thic table and calculation PS 79; six out of '
ten l>sds an page one of the procedure were wrong. The discrepancies found in
calculation PS 79 would not necessarily result in an automatically applied calculated
dieselload above the 2000 hour0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br /> ratlhg. BECo indicated that they had known about
the inconsistencies since the 1995 SWSOPI, and had issued tracking item SW l
95.0029.04 to revise Procedure 2.2.8.
However both calculation PS 79 Revision i 4) and Procedure 2.2.8 have since been
revised Revision 42), but without ensuring consistency. The inspector found that '
Engineer ng staff had not issued any Interim instructions to Operations cautioning
them ab eut this discrepancy. BECo indicated that following a start of the diesel
gJneratf/, operators Would be dispatded locally to monitor its operation, including
engine temperature; therefore, it was believed that they would not overload the
diesel. The inspector noted that not only was the tracking item still open, but it
was not included or entered as part of the formal P;lgrim corrective action process.
Knowledge of the discrepancy iIso apparently pre dated the 1995 SWOOPl
observation (i.e., PCAO 88 506). Moreover, the inspector noted that proper
removal of loads is critical to not exceeding the 2750 kW design limit.
Section 5.3 of Procedure 2.2.8, Precautions, notes that the connected load is
reduced to less than 2750 kW after 10 minutes. This is the 2000 hour0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br /> rating for
the diesel but does not account for the accuracy of the kW meter. BECo issued i
Problem Report PR 97.9559 on September 5,1997 to address this issue. BECo
indicated that any inaccurar e in the meter would be less than the 10% overload
capablilty of the niachine, and would be correc,tod by the operators dir atchnd to :
- the diesel prior tr. reaching the t.vo hour (in 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />) overload limit,
c. Conclustom
The failure t1 correctly translate the dasign baals results of calculation PS 79 into
the loading table in Procedure 2.i',8 and the lack of consideration of the accuracy of
the kW meter constitute the fourth exemnle of an apparent violation of 10 CFR 50,
Aogr h B, Criterion 'll, Design Control. (eel 97005 02)
,
'
l
, y, _ y.,my...-_,.~,,. . . , . , . . ~ .r,.,,. . ,.,._.-,,_m--- y-.,.,...,._,m.y
_ _ _ _ _ _ _ _ _ _ _ _ . - _ _ _ - _ _ _ _ _ _ _ _ _ _ _ __ _ _ _ ____ _ __ _ _ --__ _ - ___
,
.
21
Similar to the finding related to the SSW pressure switches (refer to Section E1.3),
the question related to EDG loading calculations was an example wherein the single
failure analysis committed to under Generic Letter 8913 had been found to be
lackmg. But more than two years after the SWSOPl team's concern for what had
been characterized as a weakness (one of seven observed) in the procedural
guidance provided to operators regarding EDG loading (particularly SSW and
RBCCW pumps), numerous inconsistencies between calculation PS 79 and
Operating Procedure 2.2.8 persisted. Failure to ensure consistency between the
design basis EDG loading calculation and the operating procedure represented a
potentialiy significant condition adverse to quality, indicative of untimely and
inadequate corrective action and the third example of an apparent violation of 10
CFR 50, Appendix B, Criterion XVI. (eel 97005 03)
E1.7 Diesel Generator Ambient Temocrature
a. Scone of insocction
One of the open items from DECO's 1995 SWSOPl self assessment was a question
of the maximum allowable ambient temperature for the diesel generator and ita
offect on engine operation. The SWSOPl team noted the specified maximum
ambient temperature of 80'F had been exceeded in the past, but the consequences
for engine cooling and combustion air were not thoroughly addressed. The
inspector evaluated DECO's progress on resolution of this problem, and reviewed the
Pilgrim FSAR, Coltec studies and Operating Procedure 2.2.8.
b. Qhpervations and FindiDDI
8ECo initially addressed the problem of operating the diesel beyond its specified
maximum ambient temperature by attempting to take advantage of any built-in
margin in Jacket water or lube oil temperature limits. BECo did not place any interim
limits on dieselloading when operating above an 88'F ambient until Revision 42 of
Procedure 8.9.1 (" Emergency Diesel Generator and Associated Emergency Bus
! Surveillance") was issued on June 20,1997. Jacket water coolant was changed
'
from a 50/50 glycol mix to pure water in June 1997 to take advantage of the
increased heat capacity and thermal conductivity.
Maximum Deslan Basis Ambient Temoerature
i
Historical data for the years 1994,1995,1996 and 1997 indicate that the ambient
temperature exceeded 88'F in each year, prior to the switch to 100% water
coolant. FSAR Table 2.315, " Temperatures - Plymouth, Mass," indicates " extreme
maximum" temperatures above 88'F; the highest temperature recorded in this table
is 102*F for the months of June, July, and August. The 88 cogree value is
apparently a "1% ASHRAE exceedance rate"; nonetheless, the EDG design
specification performance ratings are based on this number.
1
___ _ ________--____ __ - __- _ ______ -_ _ _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ - _ _ _ _ _ _
,
.
22
BECo had not performed an operability evaluation until the time of this inspection,
when problem report PR 97.9303 was initiated. The inspector was concerned that
this evaluation was based upon oreliminarv information obtained from the diesel
supplier, Coltec, on Novembei 8,1996 (Report R 6.00-0260). Coltec's evaluation
did not unilaterally endorse operations above 80'F ambient, nor change established
operating limits. Therefore, at temperatures in excess of 95'F, the diesel generator
and other components ins!de the room could have been subjected to temperatures
beyond their design if the closels had been called upon to operate in a design basis
event. This represented operation in an unanalyzed condition prior to June 1997
significantly compromised plant safety, outside of the Pilgrim design basis of 88'F
ambient fur rated EDG performanco, and is the fifth example of an apparent
violation of 10 CFR 50.72 and 50.73. (eel 97005 04)
June 1997 Desinn Channe
The inspector found that ambient temperature has exceeded 88'F during the
summer months. Worst case recent data from Met Tower readings indicate e
maximum ambient of 97'F reached between 1:00 to 2:00 P.M. on July 14,1995
and May 21,1996. Temperatutes at lower ground elevations -where outside air
enters the EDG Duilding and is mixed with room air--are expected to be higher (e.g.,
heating off of black top and surface effects). The inspector independently used the
predictions of the Coltec Analysis to make the case that the ability of both EDG's to
achieve rated 2750 kW output was in question during past instances where ambient
temperature exceeded 95'F. At temperatures in excess of 95*F, this condition
could result in derated engine performance, even with the "new" predict;ons for a
50/50 mix.
Nonetheless, the results of the preliminary Coltec evaluation were used as part of a
Pilgrim Design Change (PDC 9715) approved on June 17,1997, to permit EDG
operation at outdoor temperatures up to 93'F with a 50% jacket water / glycol
mixture, and up to 102'F with a 100% water mixture. These limits were formally
instituted within Procedure 2.2.8, " Standby AC Power System," Section 5.2,
Administrative Limits. The die.els were operated with the 100% water coolant (for
the first time) during the summer of 1997. BECo attempted to verify the predicted
jacket water temperature in the Coltec evaluation by testing at 2000 kW load and
one hour runs during the months of June, July and August 1997. The expected ten
degree drop in Jacket water temperature did not occur, and the change was
ultimately considered indeterminent because it did not produce the results predicted;
the licensee subsequently initiated atiother problem report PR 97.9536 on August
25,1997 and replaced the 100% water with the 50/50 glycol mix in early October
1997.
HQnertobility Evaluation
The EDG design needed to be modified to accommodate operation above the 88'F
design basis, yet the licenseo maintained in an August 16,1997 evaluation that this
was not reportable since it is...a nominal design parameter... not a station operating
or licensing limit... it is a component limit." The inspector disagreed with the
. __ _ _ _ _ _ _ .
)
.
i
'
23
evaluation, noting not only that the design needed to be thanaed in sune 1997 but :
'
also that FSAR Section 10.9.3.9 and FSAR Table 10.9 2 require revision to address
the 88'F outdoor ambient and related engine coolant (i.e., antifreete). Since both ,
EDG's would be potentially affected in a common mode sense by high temperature, ;
the inspector considered the functionalMy issue as significant. The engineering !
evaluation associated with PR 97.9383 condudec; operability based on 93'F l
ambient air with a 50/50 glycol mix; this remilns inconsistent with past Station
experience at outside air in excess of 93*F, even af ter the June 1997 change was ,
later found to be not well founded.
in fact, the evaluation of reportability was delayed, and dispositioned eight weeks
ofter the operability determinations, suggesting a flaw in the reportability process, .
and an emphasis upon operability in contrast to design basis focus. l
!
Safety Evaluations -!
The common mode failure of both EDG's represents a risk significant implication.
The inspector therefore questioned the scope of the licensee's evaluation of high .
ambient temperature and suggested that the diesel room could possibly experience
an even hotter temperature when the outside ambient exceeds 88'F. The inspector
found that the generator was rated for 2600 kW in a 40*C (104'F) ambient. The ,
diesel generator specification (6498-M 0, Rev.1, dated September ^,1969) l
Indicated that the room design temperature was 105'F. BECo responded by issuing
problem report PR 97.9540 on August 28,1997, which indicated the room
temperature could possibly reach 13 degrees above outside ambient during diesel
operation. Therefore the room temperature could reach 115'F.
During the evaluation stage for this problem report, BECo administratively limited
the operation of the diescis at 88'F outside air temperature. The safety evaluations
for both the June 1997 change to 100% water and the August change back to a
50/50 mir. of glycol did not address highor air temperature effects on key engine
performance characteristics such as fuel consumption rate. Further, the overall
impact on accelerated engine wear and the possible engine power de rating which
90100*F ambient air would conceivably have, were also not addressed in either
SE 3102 (in June) or SE 3110 (in late August). The evaluations were concluded to
be narrowly focused (again upon operability, similar to the design work), and were
not critical of the more safety significant aspects of the issue.
The inspector found that safety evaluations, 3102 and 3114, (upon which Revision
41 to procedure 2.2.8 was based), were not comprehensive and-in retrospect-
'
technically incorrect. _ Further, the evaluations were generally qualitative in nature ,
and based on preliminary input not properly validaied. Even when post modification
tests in June did not achieve the expected results, the engines were considered
operable through the next four months. The testing did not conclusively
demonstrate design basis EDG function above 88'F ambient; therefore, operability
with water / glycol mixturo above that temperature remained questionable,
_ ._ . _ . _ _ ,_ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ __ . _ . , , __
.. . _. . - .
.
.
24
c. Conclu'dQa!
DECO's corrective actions, for the realistic likelihood of operating bo h diesel
generators above their specified design outdoor ambient maximum temperature of !
88'F, were untimely and narrowly focused. This issue, which pre-dated the 1995
SWSOPI, also presents a concern for 'perability prior to summer 1997 with 50/50
Olycol Jacket water coolant when amt nt exceeded 95'F on occasion. This is the
fourth example of a significant condition adverse to quality which has not been
appropriately resolved and an apparent violation of 10 CFR 50 Appendix B, Criterion
XVI, Correctivo Action. (eel 97005 03)
E1.8 Drvwell Sorav Flow Modifications
a. Inspection Engno.122R_Q21
The inspector reviewed portions of several Problem Reports (PR), a past
modification of the RHR spray header caps to reduce drywell spray capability,
related Safety Evaluations 2133 and 3090, and a more recent chango to increar,e
drywell spray flow to 1245 gpm. Desien Change PDC 86 52A was engineered in
1987 to replace the existing upper and lower drywell header spray caps with
assemblies having six of seven nozzlos capped. The resulting lower spray flow rate
reduced the risk of structural damage from an inadvertent spray Initiation in a hot
dry atmosphere, and allowed for a beyond design 5 asis alternate spray configuration
in certain risk significant scenarios. However, the reduced spray capability
impacted the limiting containment temperature profile used for environmental
qualification (EO).
b. Observations and Findinas
1987 Sorav Modification
BECo determined (based on a 1987 GE study) that a spray flow of 300 gpm was
sufficient to maintain the drywell airspace temperature to less than 340 degrees as
well as below the drywell liner structural design limit of 281'F. The calculation
(M6601, Revision 0) predicted a flow rate of 720 opm for the maximum det an
spray capability from December 1988 through May 1997.
DE Model Errors
in January 1996, BECo identified a question related to the development of new
drywell temperature profiles associated with higher (75'F) SSW inlet conditions.
BECo found that the preliminary profiles were not similar to the analysis of record.
The computer modelling error involved certain small break sizes and an incorrect
" simplifying" aswmption that drywell temperature was equal to the reactor vessel
temperature. Consequently, a slightly higher peak temperature (334*F versus
330 F) was predicted and from one hour to ap roximately 220 hours0.00255 days <br />0.0611 hours <br />3.637566e-4 weeks <br />8.371e-5 months <br /> after the
l
event, substantially higher drywell temperature (averaged approximately 35 degrees
higher) than the analysis of record since 1987 (SUDDS/RF 87 917, dated
'
_
.
.
25
September 2,1987, "Drywell Temperature Analysis"), representing an apparent
violation of 10 CFR 50.49(o)(1). (eel 97005 05).
A problem report was generated to assess operability after the model error was
discovered. New (corrected) EO profiles were produced- both for 65'F and 75'F
SSW- and five motor operated valses (located insido the drywell) were identified as
having position indice.tlon not fully qualified. The evaluation at that time considered
the position indicators for the five containment isolation val'" in question to be
able to "... survive the 30-day mission time with the higher drywell temperature,"
but this was based upon engineering judgement and required later re testing to
demonstrate they woro "qualifiable."
The safety function of containment isolation was deemed fully operable, because of
tho short active mission period (to close) and the subsequent time nooded using
position indication for operators to verify the change of state of the valves. As part
of the corrective action associated with PR 96.9028 initleted on January 26,1996,
BECo recommended that a requirernent be added to Nuclear Engineering Procedure
3.01, " Review, Evaluation and Accoptance of Supplier Design Documents". to
review inputs and assumptions used by suppliers in safety related analysos to the
extent practical. The inspector noted that this procedure had not been changed,
representing the fifth example of untimely / inadequate corroJive action and an
appars.nt violation of 10 CFR 50, Appendix B, Criterion XVI. (eel 97005 03)
BECo representativos stated that a reportability evaluation had been conducted in
1996, in which the condition was fourid not to be reportable, in response to NRC
questions, BECo completed another reportability evaluation on September 5,1997,
which (onco again) concluded that the condition was not reportable. The inspector
disagrood with this determination on the basis that the drywell temperature profilos,
which constitute the analyses of record for EQ purposes, establish a design basis
condition for the plant. And, between 1987 and 1996, the drywell temperature
profiles of record could have been exceeded during postulated main steamlino break
(MSLB) accidents. Hence, this condition represented an unanalyned condit on that8
significantly compromised plant safety, and which was outside of the plant design
basis and the sixth examplo of an apperent violation of 10 CFR 50.72 and 73.
(eel 97005 04).
Further Nonconservatism
Problem Report 97.9101 originated on February 12,1997, identified two concerns
related to the drywell spray flow: (1) the now (proposed) drywell temperature profile
based on a 75*F ultimato heat sink temperature assumed design RHR pump
performance, as opposed to the minimum pump performance permitted by Technical
Specifications; and (2) the analysis was based on a hydraulic calculation that
underestimated the difference in elevation between the minimum torus water level
and the uppJr drywell spray header. The engineering evaluation associated with PR
97.9101 determined that the affected equipment was operable if existing RHR
pump performance was considered. The inspector noted that these errors resulted
in reduced spray flow capability, below the 720 gpm value upon which EQ profiles
i
. . . _ , _
- . . _ _ _ _ _ _ _ _ _ . . . _ _ _ _ . _ _ . _ _ . - . _ . . _ _ _ _ . _ _ . .- ___
,
,
.
V
'
!
26 I
and containment response were based. Also notable was that the EO calculations
of record throughout this entire period in question (19881997) were based on
65'F SSW, which had been exceeded on many occasions.
Increased Sorav Caoability
l
In conjunction with Pilgrim Amendment No.173, the NRC required further analysis
using a higher decay heat input, which results in higher predicted drywell
temperatures. To offset the resulting temperature increase, a modification was
instituted during the May 1997 outage (PDC 97 06) increasing the number of spray '
nozzles on the upper drywell spray header from 104 to 208, and from 104 to 196
on the lower header. The resulting spray flow :apability was increased to 1245
gpm. BECo concluded in Safety Evaluation No. 3090 that the increase in flow rate
offset the drywell temperature increase that would otherwise result from the higher
dacay heat input.
c. Conclusio.D1
from 1988 until early 1996, Pilgrim was operated in a condition where design basis '
drywell temperature profiles could have exceeded established Environmemal
Oualification design temperature limits for equipment exposed to this environment
during a postulated (MSLB) accident. While the 1987 computer modelling error was
a subtle problem not easily identified, it represented a violation of 10 CFR 50.49 not
reported to the NRC, and for which appropriate corrective action (under PR
96.9028) has not been instituted.
i
E1.9 SSW Svstem Inservice Testina
! ,
'
a. Inspection Ergp1(92903)
The 1995 SWSOPl self assessmont team identified a SSW system testing weakness
regarding flow measurements. The ultrasonic flow instruments installed by plant
design change (PDC) 91 10 to support inservice testing (IST) were not reliable or
formally incorporated into IST program by Relief Request RP 5. The inspector
- reviewed plant records and had discussions with plant personnet regard
- ng the
licensoe's IST program and its irnplementation, including the use of ultrasonic flow
instruments in periodic SSW pump tests. The inspector also reviewed the
licensee's actions regarding LER 9510 which involved an erroneous reportability
l
determination rendered in 1994 regarding the past operability of the SSW pumps,
b. Observations and Findinas ,
l -Relief Reauest RP 5
Procedure 8 l.1,1, inservice Pump and Valve Test Program, Revision 8, dated
March 11,1997, was in effect at Pilgrim for defining IST program requirements.
The basis of Relief Request RP 5 indicated that the flow measurement requirement
during the SSW pump test was not achievable due to lack of instrumentation.
. -_ ~ - . - - . _ , - - - , _ . - - _ . . _ -
.
- . . - - . _ .
. ,
, .
l
27
DECO's basis for relief stated that while shutoff head testing would still bs used,
ultrasonic flow instruments would also be used during a proving period to determine
their reliability and accuracy to support SSW pump fu4 flow testing.
Full flow quarteriv testing with ultrasonic flow instruments in accordance with
Procedure 8.5.3.2.1, SSW System Pump and Valve Operability Tests with Full Flow
Test Conditions, began in 1996. Prior to this time, BECo had taken measures to
properly calibrate the ultrastnic flow measurement equipment to ensure that the
ASME Code accuracy requirements would be met when using the equipment for
official IST tests. The celibration work was done at A's' t. abs with the 22. inch
titanium piping spool representative of the Pilgrim Sh f.pmg. The ultrasonic flow
equipment was installed so as to match the results against the Alden Lab
instrumentation which had been calibrated to the NationalInstitute of Standards and
Technology.
BECo adopted the use of the ultrasonic flow instrumentation for SSW pump
inservice testing per Procedure 8.5.3.2.1 in October 1996. The inspector verified
that BECo had taken appropriate measures to ensure that the flow instrument
accuracy requirements were properly accounted for during testing and that
adequate instructions were included in the Procedure 8.5.3.2.1 for operating the
flow measurement equipment. BECo was preparing Revis;on 9 of the IST program
to retract Relief R squest RP-5 since they were now conforming to the ASME code
requirement for ficw measurement.
LER 9510
During a previous inspection (NRC Inspection Report 95 22), the inspector identified
an erronooJa reportability determination made by the licensee regarding several
degraded SSW pumps. This item was documented as part of Unresolved item 95-
2102. This determination was the result of a reportability evaluation conducted in
conjunction with Problem Report 94.9385. The PD, in turn, was written to
document a discrepancy between Technical Specification (TS) requirements for a
SSW pump head (2700 gpm at 55 feet TDH) versus a SSW system calculatun that
concluded a SSW pump should delivu. 2700 gpm at 87.5 feet TDH. Upon further
review of the reportability evaluation, the licensee determined instances where SSW
pumps were inoperable due to fa: lure to meet the TS requirement and hence
reportable in accordance with 10 CFR 50.72. This development caused the
issuance of LER 9510.
The inspector evaluated the effectiveness of the corrective acticas taken to
determine if similar erroneous reportability determinations may have been made
regarding other issues. The licensee had conductcd a 3 Jerson review of about
25% of the reportability evaluations conducted in the sa'ne 1 year period. The
licensee concluded in LER 9510 that the erroneous tor ortability determination was
an isolated occurrence. However, the inspector noted that this conclusion was (in
retrospect) not well supported based on the findings from this current NRC
inspection, specifically considering the numerous examples regarding licensee's
f ailure to adhere to the reporting requirements of 10 CFR 50.72.
_
.
.
28
c. CRD_qlusloas
BECo has adequately implemented quarterly full flow testing of SSW pumps using
ultrasonic flow instruments in accordance with the ASME Code requirements for
flow measurement. The conclusion in LER 9510, that a past erroneous
reportability determination was an isolated occurrence, is not well sue . 4rted Lased
on other examples of reportability problems in this report,
i
E1.10 Automatic Startina of SS$ and RBCCW Pumos durina ECCS Testina
a. Insoection Scoce (92903)
The 1995 SWSOPl self assessment team observed a testing weakness in regarding
technical specification ECCS testing. These tests are performed every refueling
outage to confirm that the emergency diesel generator adequately accepts
emergency loads, but did not confirm the start of smaller loads such as the SSW,
RBCCW, er TBCCW pumps. The inspector reviewed the applicable procedures and
discussed them with cognizant Engineering and Operations personnel,
b. Observations and Findinas
The applicable plant procedure that describes ECCS testing is Procedure 8 M.3.1,
"Special Test for Automatic ECCS Lo,ad Sequencing of Diesels and Shutdown
Transformer with Simulated Loss of Off Site Power". The inspector confirmed that
this procedure had been revised to include specific steps to record the automatic
ctart of the SSW, RBCCW, and TBCCW pumps which adequately resolved the 1995
SWSOPi comment. The inspector also reviewed the results of the most recent
special ECCS test conducted in April 1997. The inspector verified that one SSW,
one RBCCW, and one TSCCW oump automatically started durir1 this planned ECCS
test.
-
c. Conclusions
BECo adequately vm t +he 1995 SWSOPl self assessment observation regarding
the confirmation of v.tallloads starting during ECCS testing.
E1.11 Sea Water Levels Assumed in SSW System Analyses
-
c. Inspectior. 3cooe (92903)
The inspector reviewe! severallicensee documents to understand the sea water
- levels used to deterrt ie SSW system flows for assessing RBCCW heat exchanger -
performance. The documents included appropriate sections of the UFSAR and
BECo Safety Evaluation (SE) 2982 which reduced the minimum required SSW flow
rate to the RBCCW heat exchanger from 5000 gpm to 4500 gpm during accident
'
conditions,
,
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - ______I____________________ .
.
4
29
b. Observations and Findinas
The inspector noted that BECo SE 2982 and Section 10.7.S. Sa:t Service Water
System Description, included a tide level of 8.3 feet as a parameter int the SSW
pump's required performance of 2700 gpm at 87.5 feet TDH actr.ss 'he pump's
impeller. The inspector further noted that this tide level wu not consisMnt with
7.1 feet minimum tide level used in the analysis to evaluate the new single 'ailure
vulnerability discussed in Section E.1.3 A third tide level vrdue of - 13 feet,9
inches related to SSW pump performance is also included in Section 2.4.4.2 of the
FSAR defined as "the minimum sea water level at which the service water pumps
maintain design conditions." The inspector discussed this information with th6
cognizant design engineer and obtained the following clarification.
1. The tido level used in accident analysis calculations was - 7.1 feet which is
the yearly astronomical minimum low tide. This value is used to determine
minimum SSW system performance required to perform the emergency
containment cooling function.
2. The basis for the tide level of - 8.3 feet previously referenced in SE 2982
and the UFSAR could not be substantiated and will be removed from the
3. The rated performance of a SSW system pump is 2700 gpm while
developing 95 feet TDH across the pump's impeller. The minimum sea water
level for the SSW pump maintaining this design performance is - 13 feet,9
inches.
'To inspector verified that the cognizant engineer had submitted an UFSAR change
raquest in June 1997 (for the next annual UFSAR update) to the BECo Licensing
Department to clarify the sea water level requirements applicable to SSW pump
performance, as discussed above,
c. Conclusions
The inspector obtained and was satisfied with the clarification of the sea water
levels applicable to SSW performance to support accident requirements. BECo was
processing a change to the UFSAR to clarify this sea water levelinformation.
E1.12 Heat Transfer Capability of Safety Related Heat Exchanaers
a. insoection Scoce (92903)
lne inspector reviewed a number of activities concerning the performance of safety
related heat exchangers. This included procedures for periodic heat exchanger
thermal performance tests, periodic SSW system full flow tests for the RBCCW heat
exchangers, and recent replacement of the etBCCW heat exchanger heads.
l
. _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ -
_
!
!
.
30
b. Observations and F;ndinos
During the recent refueling outage, the licensee modify the heads of both RBCCW
heat exchangers to prc, vide a long term resolution of por head cracks. The most
recent head problem, which was reported in LER 96 008, was a through-wall crack
in the channel cylinder of the "B" RBCCW heat exchanger due to fatigue caused by
cyclic loading of the attached lower partition plate. The original heat exchangers
wem under designed such that the channel partition plates and channel heads were
thinner. With the increased cylinder shell and partition plate thickness and some
changes to the tube inlet flow pattem, the inspector was initially concerned that the
hydraulic loss in the new head might significantly increase and hence have a
negative impact on prior SSW system calculations. However, after discussion with
the cognizant engineer responsible for the modifiention, the inspector loamed that
this potential problem area had been reviewed with the heat exchanger vendor who
concluded that the increase in the hydraulic losses in the new head would be
insignificant. The inspector had no further comments.
In reviewing the plant records that documented the replacement of the RBCCW heat
exchanger heads, the inspector noted that BECo had performed a SSW system full
flow test on each heat exchanger in accordance with Procedure 8.5.3.14, SSW
Flow Rate Operability Test, prior to returning it to service. Procedure 8.ti.3.14
provides a method of demonstrating that each SSW system loop can provide 4500
gpm to its respective RBCCW heat exchanger without exceeding the heat
exchanger's design basis pressure drop. Plant records indicated that the "B"
RBCCW heat exchanger modification had been completed in March 1997 with a
satisfactory retest per Procedure 8.5.3.14. However, on April 3 and 4,1997, with
the plant shutdown, the "B" RBCCW heat exchanger f ailed to meet the procedure's
pressure drop acceptance criteria. The heat exchanger was declared inoperable and
immediate corrective actions were taken to backflush it repeatedly.
The inspector reviewed the associated Problem Report 97.9261 wherein BECo
concluded that the direct cause was attributed to macrofouling from a severe
northeaster storm of April 1,1997. This Problem Report had been closed and the
inspector questioned the thoroughness of the licensee's corrective actions as
follows:
1. Were BECo's corrective actions appropriate knowing that the "A" RBCCW
heat exchanger was undergoing modification at the time and not available for
cooling?
2. Were lessons teamed from this problem such that existing storm readiness
procedures could be improved?
During subsequent discussions the inspector determined that BECo had acted
appropriately to correct the flow blockage of the "B" RBCCW heat exchanger.
However, regarding lessons leamed from the problem, BECo agreed to improve
Procedure No. 2.1.37, Coastal Storm - Preparations and Actions. Provisions would
be included to check for potential safety system degradations, such as macrofouling
i
1
. _ __ _ _ _ _ _ _ _ _ _ _ _ _ - - _ _ _ _ _ _ _ - _ _ _ - __
'
1
1
.
31
in portions of the SSW system, after a storm. This procedure change would result
in a proactive rather than a reactive approach to the effects of a storm. The
inspector had no further comments.
BECo had compiled Procedure No. 8.5.3.14.1, RBCCW Heat Exchanger Thermal
Performance Test, and used this new procedure to test the "A" RBCCW heat
exchanger during the 1997 outage. The "B" RBCCW heat exchanger will be tested
during the next outage. The procedure clearly presented the acceptance criteria
which included appropriate allowances for uncertainty of the test results. The test
results showed fouling factors well below allowable values,
c.- Conclusions
BECo has been adequately testing the RBCCW heat exchangers at full flow
conditions, demonstrating design basis capability. The new test exchanger thermal
performance test procedure contained clear acceptance criteria.'
E1.13 SSW System Pioinn Throuch Wall Leakaae
a. Insoection Scone (92903)
The inspector discussed the long term corrective actions being taken regarding a
recent through wallleak in a section of SSW system piping,
b. Observations and Findinas
On June 23,1997, a through wallleak (3 places) developed in the SSW outlet
piping of the "B" RBCCW heat exchanger near the downstream flange of a butterfly
valve, MOV 3806. BECo developed a corrective action p!an to effect short term ,
repairs of the affected piping. The NRC reviewed this matter as discussed in NRC
Inspection Report 50-293/97-07. The inspector was also concerned about BECo's
long term corrective actions, including the root cause analysis, since this problem is
intimately related to recommended action ill of Generic Letter 89-13 regarding the
need for a routine inspection and maintenance program for open-cycle service water
system piping to ensure that erosion and corrosion cannot degrade the system
performance.
BECo issued Problem Report 97.9399 on June 24,1997, to address the problem.
The defective carbon steel piping spool was determined to be part of the original
SSW system piping. Initial evaluations attributed the cause to a localized loss of
rubber lining and subsequent erosion / corrosion of the carbon steel pipe. The
inspector discussed the long term corrective action plans with the cognizan* deign
engineer who also was involved in the pipe patch implemented for the short term
repairs. Several engineers were assigned to perform a root cause analysis which
would not be finalized until the section of piping was removed for analysis (probably
the next refueling outage). An inspector followup item will be opened to review the
results of the final root cause analysis of the through wall piping leak near MOV-
3806 (IFl 97005 03).
_ - _ - - _ _ _ _ _ _ _ - _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - -
- . _ - . - = - - . _ - - . . . _---- - - - -.-- - -.
.
-9
32
c. - Conclusions
BECo took adequate corrective actions to effect short term repairs of the SSW
piping through-wall leaks. The root cause analysis of the piping leaks has not been
finalized, pending the removal and inspection of the pipe section during the next
outage.
E1.14 Plant Walkdowns
b
a. Inspection Scooe (92903)
The inspector observed the mat 6 rial condition of various SSW and RBCCW system
equipment during plant walkdowns. The inspector discussed these observations
with BECo personnel.
'
b. Observations and Findinos
The inspector noted tha modifications (Plant Design Changes 94-29 and 94 31) that
had been implemented to upgrade the screen wash system piping and strainers and
the screenwelllevel instruments. During a walkdown in the reactor building, the
inspector questioned the preventive maintenance measures applicabie to the
equipment hoses that connect the RBCCW piping to the ECCS pump area coolers.
BECo confirmed that these hoses were original plant equipment. While the exterior
of thest. hoses were periodically inspected, the licensee could not confirm the
condition of the interior of the hoses, and decided to issue maintenance work
>
requests for replacement. The inspector had no further comments.
During a plant walkdown in the intake structure, the inspector questioned the SSW
system pressure switch tubing's ability to withstand seismic interaction with
adjacent SSW discharge piping in two pipe floor sleeve locations.1he licensee's
civil / structural engineer confirmed that restraints were in place at both sleeve
locations to protect the tubing from the potential interaction. The inspector found
this resolution to be acceptable,
c. Conclusions
Material condition in the areas toured and for the equipment observed was
acceptable. Several walkdown observations that were noted by the inspector were
satisfactorily resolved.
_ , _
. __ . . _ . . _ . .-. _ _ _ _ . _ . _ _ . _ _ - . _ .. . _
.
.
33
'
E7- Quality Assurance in Engineering Activities
E7.1 independent Overslaht of SWSOPl Corrective Actions
a. _ Inspection Scope
- The inspector met with BECo quality assurance (QA) management representatives
to understand the QA Department's effo.ts taken in the past two years regarding
BECo's corrective actions for the 1995 SWSOPl self assessment.
b. - Observations and Findinas
Although there was a OA department review conducted regarding the use of
containment overpressure in calculating the NPSH available for the ECCS pumps *
(see Section_ E1.1), there did not appear to be any other substantial QA effort
regarding the corrective actions associated with the 1995 SWSOPl self assessment _
findings. A QA finding in December 1995 indicated that the action items in the
corrective action tracking system related to_the SWSOPl findings were not being
maintained, and many of the items were still(at that time) open,
c. Conclusions
Oversight by the QA Department regarding _1995 SWSOPlissues was limited and,
ultimately ineffective with respect to bringing forward the Criteria XVI findings from
this NRC inspection.
E8 Miocellaneous Engineering issues
E8.1 Review of Updated Final Safety Analvsis Report (UFSAR) Commitments
A recent discovery of a hcensee operating their facility in a manner contrary to the
FSAR oescription highlighted the need for a special focused review that compares
plant practices, procedures, and/or parameters to the UFSAR descriptions. While
performing the inspections documented in this report, the inspector reviewed the
applicable portions of the UFSAR that related to the areas inspected and verified
that it was consistent with observed plant practices, procedures, and/or parameters
except as noted in Sections E1.2, E1.3, and E1.11.
4
V. Manaaement Meeting
X1 Exit Meeting Summary
The inspector' discussed the findings with the licensee staff and management at the end of
each weekly visit, such as the substantial exit meeting of July 18,1997 A final exit
meeting was held on August 28,1997. The licensee acknowledged the findings
presented. No proprietary materials were knowingly retained by the inspectors or disclosed
- in this inspection report.
e
a,
, - . - , , . - -
. , , , + + ,-
4
.
34
PARTIAL LIST OF PERSONS CONTACTED
Boston Edison Comoany
E. Boulette Senior Vice President, Nuclear
P. Harizi Mechanical Engineer
J. Alexander QA Manager
H. Oheim General Manager, Technical Section
T. White Mechanical Engineering Manager
T. Trepanier Acting Plant Manager
N. Desmond Regulatory Relations Group Manager
R. Haladyna Regulatory Affairs Engineer
P. Kahler Regulatory Affairs Engineer
M. Jacobs Operations Manager
J. Keene Regulatory Affairs Manager
B. Chenard Electrical Engineering Manager
J. Coughlin Electrical Engineer
Nuclear Reautatory CommissigD
E. Kelly DRS Systems Engineering Branch Chief
R. Laura Senior Resident inspector, Pilgrim
R. Arrighi Resident inspector, Pilgrim
INSPECTION PROCEDURES USED
IP 92903 Followup - Engineering
'
ITEMS OPENED, CLOSED, AND DISCUSSED
Ooened
Safetv Evaluation
97-05 01 eel Unreviewed safety question on NPSH issue for ECCS pumps
97-05-01 eel No safety evaluation to address operation of RBCCW above 65"F
97-05 01 eel Licensing Basis of RBCCW inconsistent with operating procedure
97-05-01 eel No safety evaluating for design change of RHR ilow
Desian Control-
97-05-02 eel Design control issue re operation above SSW design inlet temperature
97-05-02 eel Failure to have adequate basis for not isolating nonessential loads and
assuring design basis information
97-05-02 eel Design control issue f ailure to translate design basis information (EDG
loading calculations) into standby AC power system procedure and
accounting for kW meter accuracy
-- - - - - - -- - -- - - - - - - - - - - - -
. - - - - -.
e
35
97 05-02 eel Design control issue on failure to assure design basis info (appropriate
RHR flow range) was correctly translated into operating procedures
Corrective Action
97 05-03 eel Inadequate corrective action regarding single failure analysis
performed in response to GL 8913
97-05-03 eel inadequate corrective action regarding SWSOPl finding of
discrepancies between Standby AC Power System Procedure and the
diesel generator loading calculation
97-05-03 eel Inadequate corrective action regarding SWSOPl finding of
inconsistency between operating procedures and design basis info
(assumed isolation of nonessential loads)
97 05 03 eel Inadequate corrective action taken to assure that the design basis
maximum ambient temperature of 88'F for the emergency diesel
generators was controlled properly
97-05-03 eel Nuclear Engineer Procedure not revised as part of corrective action
from PR 96.9028
Reportability
97-05-04 eel Failure to report operation beyond design basis SSW design inlet
temperature
97-05-04 eel Untimely report filed re new single failure vulnerability
97-05-04 eel Failure to report possibility of being outside design basis since
procedures did not require isolation of RBCCW flow to nonessential
loads
97 05-04 eel Failure to report possibility of operation outside design basis at RHR
flows at or below 5100 gpm
97-05-04 eel Failure to report a condition outside the plant design basis regarding
an adverse drywell temperature profile
, 97-05-04 eel Diesel generator ambient temperatures resulting in components
!
potentially subjected to temperatures beyond their design
97 05-05 eel Drywell temperature averaged approximately 35*F higher than
analysis of record since 198710 CFR 50.49(e)(1)
97-05-06 URI lsolation redundancy for salt service water valves
97 05 07 IFl Complete benchmark testing of RBCCW system to validate flow
I
model
97-05-08 URI Confirm that adequate RBCCW flow provided to core spray pump
when flow is not isolated to nonessential loads
l 97-05-09 IFl Understand the final root cause analysis of the SSW system through
wall piping leaks
l 97-05 10 eel Failure to Update FSAR
l
l Closed
50-293/97-01-03 URI Unreviewed safety question on NPSH issue for ECCS pumps
50-293/95-21-02 URI Elevated SSW inlet temperature at Pilgrim