IR 05000245/1990001
| ML20057B455 | |
| Person / Time | |
|---|---|
| Site: | Millstone  |
| Issue date: | 08/30/1993 |
| From: | Durr J, James Trapp NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| Shared Package | |
| ML20057B451 | List: |
| References | |
| 50-245-90-01, 50-245-90-1, NUDOCS 9309220058 | |
| Download: ML20057B455 (11) | |
Text
U. S. NUCLEAR REGULATORY COMMISSION
REGION I
)
REPORT NO.:
50-245/90-01 DOCKET NO.:
50-245 LICENSE NO.:
DPR-21 LICENSEE:
Northeast Nuclear Energy Company P. O. Box 270 Hartford, Connecticut 06141 FACILITY NAME:
Millstone Unit 1
,
INSPECTION AT:
Waterford, Connecticut INSPECTION CONDUCTED:
January 8 - 12, 1990 Inspector:
@
9-24-9 3 JameMi. Trapp, Team Leader Date Engineering Branch, DRS
-
Approved by:
d J
Jacq8P. Durr, Chief
' D Ite Engineering Branch, DRS INSPECTION SUMMARY: Routine Unannounced Inspection on January 8 - 12, 1990 (Report No. 50-245/90-01)
AREAS INSPECTED: Review of the Licensee Event Report (LER) No. 89-22, regarding feedwder coolant injection (FWCI) system inoperability declared on November 17, 1989.
(NRC IM92703)
RESULTS: One apparent violation was identified for the failure to perform the required engineering analysis to determine FWCI operability in a timely manner. In addition, two unresolved items were identified for the seismic classification of the air system and the lack of available test result documentation for establishing FWCI system operability.
.
9309220058 930910 PDR ADOCK 0500
G
DETAILS 1.0 PERSONS CONTACTED
1.1 Northeast Utilities Company
- E. Abolafia, Engineer
- M. Bigiarelli, Asst. Eng. Supr.
'G. Cornelius, Gen. Mech. Eng.
D. Gerber, Engineer
- J. Ferguson, Gen. Mech. Eng.
- N. Jain, Reactor Engineer
- W. Noll, Engineering
- E. Olszewski, I&C
- R. Palmieri, Operations
- P. Przekop, I&C
- J. Quinn, Maintenance
- M. Ross, Operations
- S. Scace, Station Superintendent
- T. Starr, Engineering Mechanics
- J. Stetz, Unit 1 Superintendent R. Tobin, GE Site Rep.
- W. Vogel, Engineering Supr.
- C. Wargo, Engineering 1.2 U. S. Nuclear Regulatory Commission
- D. Dempsey, Resident Inspector
- W. Raymond, Sr. Resident Inspector
- Denotes presence at exit meeting held on January 12, 1990.
2.0 REVIEW OF LER NO. 89-22, "FEEDWATER COOLANT INJECTION DECLARED INOPERABLE" As detailed in LER No. 89-22, on November 17, 1989, the licensee declared the feedwater coolant injection (FWCI) system inoperable and entered a seven-day limiting condition for operation (LCO). The purpose of this inspection was to review the events related to the above FWCI system inoperability. The scope of this inspection was to:
Assess the timeliness of problem identification and corrective action;
Review operability analysis which indicated FWCI inoperability;
Review ongoing analysis described in the LER;
Verify technical adequacy of the modification;
Review reportability issues; and
Review FWCI for other loss-of-air vulnerabilities.
- - -
.
-
=
-
-
2.1 Introduction The FWCI system is a high pressure emergency core cooling system which uses existing feedwater and condensate system components to supply cooling water to the core. During a loss of coolant accident sensed by low-low vessel level or high drywell pressure, the "A" or
"B" condensate /feedwater string automatically starts and provides core cooling. Each of three reactor feedwater pumps (RFPs) has a minimum flow valve which provides a recirculation path from the discharge of the RFP back to the main condenser. The minimum flow valves open on low suction flow to the RFP, providing pump protection at low How conditions. The minimum flow valves have air operators which fail open on a loss of instrument air.
'
The feedwater regulating valves (FRVs) are used to regulate flow to the vessel and provide RFP runout protection. The FRVs are normally controlled using the reactor vessel level in
>
the single element control mode. When a high flow is sensed by flow elements beated at the outlet of the high pressure feedwater heaters (downstream of the RFP), the FRVs switch to
+
flow control. Flow control limits the maximum RFP flow to 105% of rated flow to provide pump runout protection. The inspector noted that the minimum Cow valves are located upstream of the Dow elements usul for pump protection. Therefore, the total pump Dow will be the sum of the minimum bypass flow valve and the flow sensed by the elements that provide the pump runout protection.
The RFP and the condensate booster pumps are protected by low suction pressure pump trips. The RFP trip setpoint is 85 psig and has a 10-second time delay. The condensate booster pump low suction pressure trip setpoint is 20 psig and also has a 10-second time delay. Both pumps motors are provided with instantaneous and time delay overcurrent protection.
On November 17, 1989, the FWCI system was declared inoperable when an engineering analysis indicated that the FWCI may not be capable of delivering design flow to the vessel during a LOCA with a concurrent loss ofinstrument air. The loss of instrument air causes the RFP minimum flow valves to fail open and to divert flow from the vessel. The engineering analysis also indicated that the RFP and condensate booster pump would be at
" runout" and be inoperable.
2.2 Discussion of the Events that Led to FWCI Inoperability The inspector reviewed the FWCI concerns discussed in the licensee's own safety system functional inspection (SSFI) conducted in late 1988, and the details of the FWCI inoperability _
on November 17,1989, as detailed below.
!
i I
..
J
-.
-
-
-
._
--
The licensee conducted a SSFI for the FWCl system in 1988. The concerns raised by the SSFI team, and the corrective action taken to resolve these findings were reviewed by the inspector. SSFIinspection observation number 1-6-88, dated November 2,1988, stated that documentation did not exist to verify that the FWCl system would adequately perform its design function.
On November 18, 1988, in response to the above SSFI observation, the licensee's engineering staff stated that preoperational tests were performed demonstrating FWCI operability. However, the documentation was not recoverable. The FWCI availability test No. T84-1-12, performed on June 27,1984, was also referenced as a document to demonstrate FWCI operability. Test No. T84-1-12 was a 20-second duration test which initiated FWCl injection water into the depressurized reactor vessel. The SSFI team member reviewed the results of test No. T84-1-12, and requested additional information. The SSFI team member was particularly concerned about noticeable oscillations in RFP/ booster / condensate pump flows and suction pressures during this short duration test. It was identified that the low suction pressure pump trips may have been challenged. In response to these concerns, the engineering staff provided additionalinformation on December 6,1988. This additional information stated that the licensing basis would be changed to more accurately define the FWCl system in the Uodated Final Safety Analysis Report (UFSAR) and Technical Specifications. The revised system function would allow the booster pump or condensate pump to supply FWCl flow when vessel pressure was sufficiently low. Therefore, the loss of a RFP would no longer be a concern. A memorandum, dated December 12, 1989, stated that this open item would be tracked as part of the Technical Specification revision. The SSFI observation and Test No. T84-1-12 are further discussed in section 2.3.D of this inspection report.
Generic letter GL-88-14, " Instrument Air Supply System Problems Affecting Safety-Related Equipment," requested the licensee, in part, to " verify that the design of the entire instrument air system, including air or other pneumatic accumulators is in accordance with its intended function, including verification by test that air-operated safety-related components will perform as expected in accordance with all design basis events, including a loss of the normal instrument air system. This design verification should include an analysis of current air-operated component failure positions to verify that they are correct for assuring required safety functions." The licensee provided a response to GL-88-14 on February 17, 1989. In their response, the licensee committed to testing the RFP minimum flow valves during the 1989 refueling outage. The response to the generic letter also stated that analyses and design verifications would be completed for Millstone 1, consistent with startup from their 1989 refueling outage, commencing on April 1,1989. The licensee stated that documentation of this analysis and design verification would be provided to the NRC when completed for all four nuclear units operated by the utility.
In June 1989, testing and verification analysis for air-operated systems requested by the generic letter were performed at Millstone 1. During review of this information, the plant
-
engineering informed corporate engineering of a potential for the RFP minimum flow valves
,
,,
yn
--
-
_
to fail open on a loss of instrument air, and to divert flow from the reactor vessel. A meeting was held between the corporate and plant engineering to assess the significance of
this issue, and to propose a solution. At this meeting, the staff discussed a number of factors that could be used to demonstrate FWCl system operability. A change to provide a backup air supply to the RFP minimum flow valve was proposed by the corporate engineering staff to the plant on July 28,1989. However, the FWCI system operability analysis was not initiated to determine system operability with the RFP minimum flow valves open.
The plant staff again questioned the operability of the FWCI system with the RFP minimum flow valves open, and, in accordance with plant procedures, originated a Northeast Utilities Reportability Evaluation form, on October 10, 1989. The plant staff performed the technical evaluations to determine FWCI operability. The evaluation concluded that the condensate booster pumps would experience run-out condition and the FWCI system would not be capable of performing its intended safety function. Corporate engineering independently verified the conclusions of this analysis. In a memorandum dated November 17, 1989, corporate engineering stated they did not have a sufficient basis to disagree with the conclusion which indicated that the FWCI system was inoperable. At 7:00 p.m., on November 17, 1989, the FWCI system was declared inoperable by the licensee.
2.3 Inspection Findings / Assessment A.
Timeliness of FWCI Operability Evaluation In June 1989, during a design review of the instrument air system initiated by NRC GL-88-14, the licensee identified that the RFP minimum flow valves would fail open on a loss of instrument air. The open RFP minimum flow valves were recognized as having the potential to divert flow from the reactor vessel to the condenser, thus degrading the FWCI system.
The licensee's staff performed an undocumented preliminary evaluation of FWCI operability and concluded that the FWCI system was capable of performing its design function. During this preliminary review, the licensee did not consider the fact that the FWCI system instrument air supply for valve control was not designed to satisfy Class I seismic requirements, as required by the original General Electric Company (GE) design i
specification. Also, during the preliminary operability assessment, calculations were not j
performed to determine operability. Approximately four months following the identification of this issue, a plant staff engineer initiated a reportability evaluation form. The technical evaluation, which was used to declare the system inoperable, was completed in approximately 3 days, as stated by the evaluating engineer. Five weeks after initiating the reportability evaluation, the FWCI system was declared inoperable, and a seven-day LCO was entered. The failure to perform timely analysis to a known operability concern is contrary to Technical Specification 3.5.C.3, and is an apparent violation (Violation No. 50-245/90-01-01).
i
,.
-
-
-
.
B.
FWCI System Operability Analysis The inspector reviewed the reportability evaluation and discussed the preliminary results of the analysis referenced in the corrective action section of LER-89-22 with licensee
,
representatives. The analysis referenced in the LER was being completed by GE at the time of this inspection. The purpose of the analysis was to determine if the FWCI system was capable of delivering design flow to the vessel with the three minimum flow valves open.
The analysis used to determine FWCI inoperability on November 17, 1989, showed that the system flow rate would cause the FWCI pumps to be in runout condition. The licensee stated that assumptions made in this analysis for series pump operation were incorrect and would be corrected in the more detailed GE analysis being performed.
The inspector reviewed portions of the preliminary FWCI analysis being performed by GE.
'
This GE analysis uses a static flow model of the FWCI system. This static model does not include the dynamic oscillations in pressure and flow, which occur when the FWCI system is
,
started. The preliminary results of the static model showed that 8000 GPM design FWCI flow would be delivered to the reactor vessel with all three RFP minimum flow valves open.
The inspector had the following concerns regarding the assumptions used in the GE analysis.
The preliminary GE analysis states that, based on the results of test No. T84-1-12 performed in 1984,12,400 GPM FWCI flow is feasible. However, these are the same test results which caused the SSFI to question FWCI operability one year earlier. A review of the test data indicated that each time a flow of approximately 12,400 GPM was approached during
,
this test, the RFP suction pressure was less than the 85 psig trip setpoint. The RFP did not
'
trip due to the 10-second time delay and the short duration of the test (approximately 20 seconds). This test was never performed at a steady state condition with a flow rate near 12,400 GPM.
The inspector also noted that the preliminary analysis did not consider the flows to other systems. A review of the condensate system revealed that the "attemporator spray" valve would also fail open on a loss of instrument air, and cause additional flow to be diverted to
,
the condenser. This will further change the static pressure / flow profile of this system.
i The inspector discussed these findings with the licensee staff at the exit meeting. The licensee stated during the exit meeting that the final results of the analysis, including all assumptions, would be transmitted to the Region I NRC office prior to February 28,1990.
j Subsequent to the inspection, a copy of the report was provided to the inspector. A
preliminary review by the inspector confirmed the analysis was for equilibrium flow j
conditions. The inspector had no further questions concerning this matter at this time.
_
-
-
_
_. -.
-
.
_
_
..
'
C.
Plant Design Change to Provide Backup Air to Feed Pump Minimum Flow Valves A review of plant design change record (PDCR) 1-33-89, " Backup Air to Feed Pump Minimum Flow Valves," was performed. The modification installed a redundant backup high pressure air supply to the RFP minimum flow valves. The design consists of the installation of two high pressure air cylinders which feed a common air header that will normally be supplied by the plant instrument air system. The modification package, including seismic qualification, was found to be detailed and of good technical quality.
,
Modi 6 cation PDCR #1 installed a backup air supply to the FRVs in 1980. The FRVs are designed to fail as-is during a loss of instrument air event. The original GE design specification for the FWCI system states that the air system supply for the control valve be designed to Seismic Class I requirements. The modification package to install the backup air supply for the FRVs indicates that this change was not installed as Seismic Class I.
Discussion with corporate seismic engineers indicated that portions of the backup air supply modification were seismic. However, there was no documentation to demonstrate this. The licensee stated at the exit meeting that a walkdown of the system had been conducted that
morning, and documentation to demonstrate the Seismic Class I qualification of this system would be made available for NRC review. This item remains unresolved pending the licensee providing the documents to show that the FWCI system meets Seismic Class I requirements (Unresolved Item No. 50-245/904)1-02).
Additionally, the inspector noted that the UFSAR, Section 3.2, Table 3.2-1, stated that the
<
compressed air system, piping fittings, and valves are classified as Seismic Class I.
Discussions with the plant staff indicated that this was not the case, and that the UFSAR required revision.
,
D.
Tests to Demonstrate FWCI Operability The licensee conducted a test of the FWCI system in 1984, with the reactor vessel
,
depressurized. The condensate pump was running as an initial condition, and the candensate booster pump and the RFP were started on time delay relays, rather than the norma start signal when adequate pump suction pressure is achieved. ' The test was initiated with a reactor vessel level of approximately 15 inches and the test was terminated when vessel level reached approximately 40 inches. The test duration was approximately 20 seconds.
Test results showed large oscillations in measured flows and pressures throughout the FWCI i
system over the duration of the test. Damping of the pressure and flow oscillations was not readily evident during a review of the test results by the inspector. A similar concern was
!
identified by a licensee staff member when reviewing the test results during the 1988 SSFI.
i s
.-
-
-
_
-..
_
-.
.
~.
--
- - - -
.-
-
.
-
.- -.
.
,
This licensee staff member recalled test data, which indicated oscillation damping was
I available, and satisfied his concern at that time. Further information regarding this issue was requested by the inspector at the exit meeting. This item remains unresolved pending the
,
licensee's documentation and evaluation of the test results to demonstrate the operability of l
the FWCI system from the above test (Unresolved _ Item 50-245/91-01-03).
4.0 EXIT MEETING i
At the conclusion of the site inspection on January 12,1990, an exit interview was conducted
'
with the licensee's senior site representatives (denoted in section 1) to discuss the results and conclusions of this inspection.
At no time during this inspection was written material provided to the licensee by the inspector. The licensee representatives did not indicate that this inspection involved information subject to 10 CFR 2.790 restrictions.
?
.
.
,
f
,
t b
- - -
.
_
,
-
_
_A
~
.s a
.+-
2a.
.-
a.na
-
F
$
?
I
f ENCLOSURE 2 i
,
'
REPORT OF INVESTIGATION
r SYNOPSIS
'
.
CASE NO. 1-90-008 a
!
i
i
.
-
!
-f i
a
C I
e l
!
i p
!
, -
-
-
. ~.
_., _,
.
CASE No. 1-90-008
p.
c.,,,
'
Nj ).,Mc ;j
!'
United States ig'-4#
<
'rJuclear Regulatory Commission
\\.....
-
'
Report of Investigation
>
Millstone Unit 1:
Earassment and intimidation /Reportability.
Violations
'
i
.
.e.
Office of Investigations Reported by 01:
RI
-
.
\\
i
_ -
-
f
-
- - - -.
- - - - - - - - -
SYfv0PSIS On March 13, 1990, the Regional Administrator, fiRC Region 1, requested that an investigation be initiated concerning (1) the alleged harassn.ent and intimidation (H&l) of a fiortheast Utilities (t<U) employee, with the knowledge of senior line management, in order to influence the results of a reportability detemination by this employee regarding the operability of the Feedwater Coolant Injection System (FWCI); (2) the alleged H&l by senior line management of that employee in retaliation for a reportability detemination regarding the operability of the FWCI system; and (3) the alleged willful failure of licensee management to address legitimate safety concerns regarding the operability of the FWCl system by improperly attempting to influence the results of the operability and reportability determination while delaying corrective action, notification and reporting of a significant safety issue until the required system modification was ready to be implemented.
The 01 investigation substantiated that the engineer was harassed by the Millstone Unit 1 Engineering Manager in an effort to influence the Engineer's evaluation of the safety system.
It was also substantiated that the Millstone Unit I Superintendent and the Engineering Manager discriminated against the Engineer by not selecting him to fill a personnel vacancy within Unit 1 Engineering, in retaliation for.is technical evaluation and conclusion regarding the safety system. The investigation detemined that there was a deliberate delay in declaring the safety system inoperable by the licensee's organization, utilizing administrative means and attempts by the Engineering Manager to avoid reportability entirely.
By delaying the operability determination for the FWCI system, the licensee exceeded the seven day Limiting Condition for Operation on the FWCI system which constitutes a potential Technical Specification violation.
The investigation detertr'ned that the system modification was not initiated until after the licensee declared the system inoperable and made the required fiRC notification.
x Case tio. 1-90-308 1