IR 05000289/1987010

Insp Rept 50-289/87-10 on 870424-0529.Major Areas Inspected: Event Response,Maint & Surveillance Areas,Reactor Coolant Leak Rate & Restoration of Main Feedwater Isolation Function on High Steam Generator Levels

ML20235H673
Person / Time
Site:	Three Mile Island
Issue date:	07/06/1987
From:	Blough A NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
To:
Shared Package
ML20235H647	List: IR 05000289/1987010 ML20235H644
References
50-289-87-10, NUDOCS 8707150226
Download: ML20235H673 (35)

Text

_-- .__ _ _ _ . - _ _ _ - - _ _ - - - _ - - - - _ _ _ _ - - -

_ - - - _ ,

.+

. .

U. S. NUCLEAR REGULATORY COMMISSION

REGION I

Docket / Report N /87-10 Licensee: ORP-50 Licensee: GPU Nuclear Corporation P. 0. Box 480 Middletown, Pennsylvania 17957 Facility: Three Mile Island Nuclear Station, Unit 1 Lccation: Middi aswn, Pennsylvania Dates: April 24 - May 29, 1987 Inspectors: R. Conte, Senior Resident Inspector (TMI-1)

A. D' Angelo, Senior Resident Inspector (Rancho Seco)

D. Johnsoa, Resident Inspector (TMI-1)

F. Young, Senior Resident Inspector (BV-2)

Reporting i Inspector: F. Young, Resident Inspector-

'

Reviewed By R. Conte, Sertior Resident '.nspector Approved By: d A. Bloug f Chief, Reactor Section No. lA, DRP MN

~Date Inspection Summary:

The NRC resident . staff conducted safety inspections (248 hours0.00287 days <br />0.0689 hours <br />4.100529e-4 weeks <br />9.4364e-5 months <br />) of power opera-tions, focusing on performance in the operations, which included event response; 4 main'enance; and, surveillance area The following events were reviewed:

reactor trips of May 1 and 2, 1987, and the Unusual Event of May 9, 198 !

Items reviewed in the plant operations area were: reactor coolant system leak l rate, restoration of the main feedwater isolation function on high steam gen-erator levels, core physics data at full power, procedural revisions for reac-tor water level indication, and cable / conduit support installations. With re-spect to equipment operability, the following items were reviewed: hydrogen recombiner maintenance / testing, steam reliefs for the steam-driven emergency feedwater pump, main steam safety valve refurbishment, and diesel generator air check valves. Licensee actions on past inspection findings were also reviewe i l

.

8707150226 870707 j PDR ADDCK 05000209 O PDR I

_____--__A

__ - _ _ _ _ . _ _ _ _ - ___ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ - _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

l l

.

. Inspection Summary-(Continued) 2 Inspection Results:

l l

!

Licensee management and the quality assurance department continued their l detailed attention and involvement in the functional areas reviewed. That involvement was particularly notewerthy during the reactor trips of May 1-2, 1987. Corrective actions from past inspections were apparently effective in resolving a concern regarding main steam safety valve performanc Further, for new issues thet arose during this inspection, licensee representatives were also responsive to NRC staff concerns. Licensee commitments made during past inspection periods were verified to be properly implemente During this period, the licensee completed the collection of core physics test data. No unexpected conditions were note Overall, operator response to off-normal events was oriented toward safety and, in general, in accordance with facility procedures. The licensee-identified nonadherence to the-reactor trip emergency procedure appears to be an isolated case and licensee management tcok appropriate corrective actions (paragraph 4.2.5.3). The planned events were adequately preplanned, procedurally con-trolled, and staffed. Post-event reviews were reasonably thorough with cor-rective actions appropriately identified, documented, and evaluated for impact on safe operation Plant response during the reactor trip events was as expecte !

_ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _

.-. . ____-__ _ __ _ ___.

.

.

TABLE OF CONTENTS Page j 1. Introduction and Overview. . . . . . . . . . . . . . . . . . 1 2. Plant Operations . . . . . . . . . . . . . . . . . . .... 2 3. Maintenance / Surveillance - Operability Review. . . . .... 9 4. Event Review . . . . . . . . . . . . . . . . . . . . . . 16 5. Licensee Action on Previous Inspection Findings. . . . . . 27 6. Exit Interview . . . . . . . . . . . . . . . . . . . . . . 30 i

l i

i i

- _ _ - _ - _

.

D DETAILS 1.0 Introduction and Overview 1.1 NRC Staff Activities The overall purpose of this inspection was to assess licensee activ-ities during the power operations mode as they related to reactor safety and radiation protectio Within each area, the inspectors documented the specific purpose of the area under review, acceptance criteria and scope of inspections, along with appropriate findings /

conclusions. The inspector made this assessment by reviewing infor-mation on a sampling basis through actual observation of licensee activities, interviews with licensee personnel, measurement of radia-tion levels, or independent calculation and selective review of listed applicable document The NRC resident office inspectors were additionally supported by a Region V senior resident inspector assigned to Rancho Seco, another Babcock & Wilcox (B&W) designed reacto .2 Licensee Activities During this period, the licensee operated the plant at essentially full power. At the start of the inspection period, reactor power was limited because of steam generator (SG) water level close to or at the high level limit. The high SG water level was due to fouling of the heat transfer surfaces in the SG, The licensee conducted a pre-planned automatic reactor shutdown (trip) on May 1,1987, to allev-iate that fouling problem. The results of that trip were apparently successful; however, on return to full power, an unplanned trip occurred (see Section 4).

Subsequent to the two trips over the weekend of May 2-3, 1987, the plant achieved full power at 11:00 a.m. on May 3,1987. Steam gen-erator "A" and "B" levels were 76 and 84 percent, respectively, of the operating range. The high level operating limit is 92 percen These levels still indicate some fouling of the heat transfer sur-faces and the licensee will review long-term actions to resolve that proble During the remainder of the inspection period, the reactor was oper-ated at full power with some power reductions to 80-85 percent power for minor secondary plant maintenanc An Unusual Event was declared on May 9,1987, due to a substantial loss of emergency communications (see Section 4).

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - _ _ - _ _ _ _ _ _ _ _ _ - _ _ - _ - _ _ _ _ _ -

_ _ _ _ _ - _ _ _ _ _ ,

.

.

On May 14, 1987, a radiation monitor in the intermediate closed (cycle) cooling (ICC) system (ICCS) indicated possible leakage of reactor coolant into the ICCS through the "A" letdown cooler. The cooler was isolated and the "B" cooler was placed in operation. Both of these coolers had been replaced during the last refueling outag At the close of the inspection period on May 28, 1987, there was a release of noble gas (less than 10 curies) through the plant stack during a changeout of a letdown prefilter. During the next inspec-tion period, the inspectors will review the licensee's critique of this even .0 Plant Operations 2.1 Criteria / Scope of Review The resident inspectors periodically inspected the facility to deter-mine the licensee's compliance with the general operating require-ments of Section 6 of the Technical' Specifications (TS) in the fol-lowing areas:

- - -

review of selected plant parameters for abnormal trends;

--

plant status from a maintenance / modification viewpoint, includ-ing plant housekeeping and fire protection measures;

--

control of ongoing and special evolutions, including control room personnel awareness of these evolutions;

--

control of documents, including logkeeping practices;

--

implementation of radiological controls; and,

--

implementation of the security plan including access control, boundary integrity, and badging practice Because of additional resident inspector coverage at this facility, the inspectors focused on the additional attributes listed below:

--

operators are attentive and responsive to plant parameters and conditions;

--

plant evolutions and testing are planned and properly authorized;

--

procedures are used and followed as required by plant policy;

--

equipment status changes are appropriately documented and com-municated to appropriate shift personnel;

_

I

-

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

.

. 3

--

the operating . conditions of plant equipment are effectively monitored and appropriate corrective action is initiated when requi red;

--

backup instrumentation, measurements, and readings are used as appropriate when normal instrumentation is found to be defective or out of tolerance;

--

logkeeping is timely, accurate, and adequately reflects plant activities and status;

--

operators follow good operating practices in conducting plant operations; and,

--

operator actions are consistent with performance-oriented trainin The inspectors focused on the following areas:

--

control room operations during regular and backshift hours, including frequent observation of activities in progress and periodic reviews of selected sections of the shift foreman's log and control room operator's log and selected sections of other control room daily logs;

--

areas outside the control room;

--

selected licensee planning meetings;

--

5/t/87 planned reactor trip to reduce fouling of OTSG, STP N ;

--

5/2/87 - unplanned reactor trip, Emergency Procedure 1210-1;

--

5/2/87 - selected startup activities; i

--

5/6/87 - reclaimed boric acid tank piping flush, Operating Pro-cedure 1104-29E;

--

5/7/87 - waste gas tank release;

--

5/13/87 - letdown cooler isolation; and,

--

5/11-13/87 - "A" train of building spray isolation for main-tenanc !

-_ - - -

- _ . _ _ _ _ _ _ _

,

. 4 During this inspection period, the inspectors conducted direct inspection time during the following backshift hours (3:00 to 7:00 a.m., Monday to Friday and Saturday / Sunday): May 1, 1987, 3:30 p.m. - 10:30 p.m. ; May 2,1987, 6:30 a.m. - 12:30 p.m. ; week of May 4, 1987 (except May 6), 6:30 :00 a.m. ; May 11,1987, 6:30 a.m. - 7:00.a.m.; and, May 12, 1987, 3:30 p.m. - 4:30 As a result of this review, the inspectors reviewed specific events in more detail as noted belo .2 Findings / Conclusions 2. Reactor Coolant System (RCS) Leak Rate The inspector selectively reviewed RCS leak rate data for the- past inspection perio The inspector independently calculated certain RCS leak rate data reviewed using licensee input data and a generic NRC " BASIC" computer program "RCSLK9" as specified in NUREG 1107. Licensee (L)

and NRC (N) data are tabulated below, i

i l

l

_________________________________j

- _ _ - _ _ _ _ _ _ _ _ . -__

• l 5 {

. TABLE 1 j l

RCS LEAK RATE DATA

1 (All Values GPM)- 1 DXTE/ TIME (NUREG 1107) CORRECTED DURATION Lg Ng N N Lg U U l

5/7/87 0.0700 0.07 -0.04 .06 0.0600' )

07:07:40

'

2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> i

5/11/87 -0.0560 -0.07 -0.16 .06 -0.0456 16:44:48 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />

• 5/17/87 0.3546 0.16 0.17 .27 0.2715 01:24:41 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />

• 5/17/87 0.0264 0.02 0.13 .23 0.2386 ,

03:39:34 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />

• 5/17/87 0.1003 0.10 -0.37 .27 -0.2663 07:04:38 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> 5/17/87 0.1262 0.13 0.03 .13 0.1361 10:58:59 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />

• 5/17/87 0.1964 0.20 0.08 .18 0.1882 15:00:54 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> 5/17/87 0.0828 0.08 -0.02 .08 0.0870 17:44:55 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> 5/23/87 0.1019 0.10 -0.01 .09 0.0953 1:38:21 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> '

5/23/87 0.1353 0.14 0.05 .15 0.1475 15:25:57 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> G = Identified gross leakage U = Unidentified leakage L = Licensee calculated N = NRC calculated

- _ _ _ _ _ _ _ _ . _

.- _-- -

.

. 6 Columns 2 and 3; 5 and 6 correlate + 0.2 gpm in accordance with NUREG 1107, (N u is corrected by adding 0.1044 gpm to the NUREG 1101 N due to total purge u

flow through the No. 3 seal from RCP' As noted above, the licensee detected in:reased RCS leakage and abnormal performance of the reactor coolant drain tank (RCDT) level indication during the weekend of May 17-18, 1987. Upon further review, the licensee identified that a sensing line for the RCDT level transmitter had a small j lea It was repaired and RCS leak rate values returned to norma The above-noted tests marked by an asterisk were declared invalid because stable - test conditions were not met or because of the abnormal behavior of the RCDT level indi- I cato Records were on file to document the results of the invalid tests and valid tests were taken at least on a daily basis in accordance with technical specification The leakare plus loss term (Lg or N G) being less than the unidentified va?ue -(N u " b u) is an open issue addressed in NRC Inspection Report No. 50-289/87-09 and NRC staff will continue to monitor i In accordance with NUREG 1107, good agreement existed be-tween licensee and NRC calculation result . Main Feedwater Isolation Function The TMI-1 Restart Hearing established a need for a safety-grade isolation function of the main feedwater (MFW) system from continued feeding of the steam generators (SG's) on high level in the SG's. The restart condition required the isolation function to be installed by Cycle 6 startup (March 23, 1987). However, the licensee identified a tech-nical problen with the SG level instrumentation; possible indicated level oscillations in response to SG pressure oscillations following a reactor trip. The licensee wanted to avoid complicating the post-trip response associated with an inadvertent MFW isolation due to SG level oscilla-tion Accordingly, they wrote a safety evaluation (SE),

which justified keeping the high level isolation function j in defeat until electronic filters were installed in the SG level instrument loop and until they completed a review of the effectiveness of those filters. As a result of the planned trip addressed in Section 4 and the licensee's evaluation of SG level instrumentation performance, the licensee enabled the MFW isolation function on March 14, 198 The NRC staff is reviewing related licensee technical /

safety evaluation J

. _ _ _ _ _ _ _ - _ _ _ _ _ _ _ .

,

. 7 2. Review of Core Physics Data for Full Power As part of the startup sequence, the licensee conducted specific reactor physics tests during power escalation to ensure that the present loaded core accurately reflects the predicted core configu"ation parameters. Operating char-acteristics were collected at zero power, several intermed-iate power levels, anc maximum achievable power of 84 per-cent (see NRC Inspection Report No. 50-289/87-09).

During this inspection period, the licensee was able to achieve full power (100 percent) and, on May 8,1987, they completed data acquisition for that power leve These full power parameters were compared against an acceptance criteria previously generated by the licensee with the assistance of the Nuclear Steam Safety Supplier (NSSS)

vendo On a sampling basis, the inspector reviewed the licensee's data that were collected by:

!

--

Refueling Procedure (RP) 1550-01, " Controlling Pro-cedure for Physic Testing," and Temporary Change Notice (TCN) 87-0070 to this procedure;

--

RP 1550-02, "Zero Power Physic Testing;"

--

RP 1550-04, "Fower Imbalance Detector Correlation Test;" and,

--

RP 1550-08, " Core Power Distribution Verification."

In general, the predicted values generated by the different vendor computer programs were very similar to those as noted in NRC Inspection Report No. 50-289/87-0 The necessary data and applicable portions of the governing procedures were completed as required. Where necessary, the governing procedure was temporarily revised in accord-ance with station procedures to allow the procedure to be performed as writte The inspector concurred with the licensee's conclusion .that the core configuration was acceptable as determined by collected dat _ _ - _ _ - . .__ -

. - - _ _ _ _ _ _

.

l

< *

'2. Reactor Water Level Indication During the design review of the Reactor Coolant. Inventory Tracking System (RCITS), the Office of Nuclear Reactor Regulation (NRR) identified an apparent single failure susceptibility because all channel indications are pro-cessed through a single computer interface " multiplexer" for control room indicatio Accordingly, the NRR staff obtained a commitment from the licensee to proceduralize their ability to take voltage readings using safety grade instrumentation at the (local) signal conditioning cabinets (upstream from the multiplexer in the instrumentation loop). By letter dated March 19, 1987, the licensee com-mitted to procedura11ze obtaining these local voltage readings for reactor water level by October 1, 198 During NRC Inspection 50-289/87-09 ( April 1987), the inspector questioned why it would take so long to change a facility procedure to provide the correlation that should !

already be known, assuming operability of the RCIT Licensee representatives indicated that a Procedure Change Request (PCR) was being processed and that it should be issued by the end of May 198 During this inspection, the licensee issued Revision 3, !

dated May 7, 1987, to Operating procedure (0P) 1103-1,

" Reactor Cooling Inventory Tracking System." Among other ,

changes, the revision incorporated Enclosure III to the l procedure, which provides the formula to calculate reactor vessel and hot leg water level with voltage readings from the signal conditioning cabinets. The inspector was sat-isfied with the timeliness of the licensee's action .2.5 Cable and Conduit Installation During the inspection of safety-related building spaces, the inspector questioned the adequacy of the installation of an instrument grounding cable for the Heat Sink Protec-tion System (HSPS). The cable was loosely tie-wrapped to the outside of safety-related conduit which has HSPS cable Further, the ground cable traversed a path to an area just above a cable tray and both penetrated an adjacent wall which had fire barrier material. The cable tray had a cable tray cover pushed asid Other open bottom cable trays were near the araa. The inspector questioned the seismic installation of ne tie wrapped cable and the adequacy of separation with no installed cable tray cover, along with adjacent open bottom cable tray __

_ - _ - __ _ ._ _ _ _ _ _ _ _ .

-<T .

- 3'

..

Licensee representatives could cot immediately identify the safety function of. the cable and whether or not the . cable ,

.. m

"'

and cable tray were properly installed. They agreed, how -

ever, that the matter- ought to be reviewed and plant elec-

'

trical engineering initiated documentation actions -to have l Technical Functions Division address the inspector'rques-

.

tion .This is unresol.ved pending completion of licensee review

~

and subsequent Region I follow-up revi ew - (289/87-10-01),

4 2.3. plant Operations Summary Licensee management and the quality assurance department continued their detailed attention to and involvement in plant operation The main feedwater isolation function on high SG. level was appropri-

.

' ately . enable The NRC staff is reviewing licensee safety evalua-tions.on this~ function. Appropriate end timely procedure revisions were.made to reflect commitments made to the NRC staff for the reac-tor water level indication.' syste Core ' physics dat'a were founc to be essentially as predicted at full power, thereby completing the post-refueling test progra An unresolved item was identified in the area of cable / conduit installatio . Maintenance / Surveillance - Operability Review 3.1 Criteria / Scope of Review The inspector reviewed the below-listed activities to verify proper implementation of the applicable portions of the maintenance and surveillance programs. This was a spontaneous review to capture on-going activities in the plant spaces as they occurred. The inspector used the general criteria listed under the plant operations section of this report. A more detailed review of equipment operability was also addressed belo Maintenance Reviewed

--

4/30/87 .- Job Ticket (JT) CM 578, Heat Sink Protection System 1 (HSPS)' level filter replacement

--

5/5/87 - JT CL 562, Post-maintenance and re-installation of

, hydrogen recombiner

)

(

'

__ _ _ _ ___

_ - - _ _ _

'

.

. 10

--

5/5/87 - Emergency feedwater (EFW) steam turbine relief valve test and oil change

--

5/14/87 - Resetting of the MFW isolation feature on high steam generator level from 94 percent to 97.5 percent

--

5/12/87 - JT CI 618, CL 243 through 248 and CL-113 related to the repair of main steam safety valves (MSSV's) during March 1987 Surve111ances Reviewed

--

5/6/87 - Post-surveillance review and return to normal of steam-driven emergency feedwater pump (SP 1300)

--

Surveillance Procedure (SP) 1303-11.3, Revision 15, " Main Steam Safety Valves," completed April 19, 1985; October 26, 1986; and, March 21, 1987 The results of this review are addressed in paragraph .2 Selected Equipment Operability Review The inspector reviewed licensee maintenance (preventive and correc-tive) and surveillance activities to assure main steam safety valve (MSSV) operabilit Specifically, the inspector was to verify:

--

procedures required by Technical Specification (TS) 6.8.1 pro-perly implement TS 3.4.1.2.1 and 3.4.1.2.2 related to MSSV operability;

--

procecures required by TS 4.1.2, Table 4.1-2, Item 4, properly implement MSSV surveillance testing requirements;

--

applicable procedures have the proper format and technical con-tent in accordance with applicable section of ANSI 18.7-1976;

--

surveillance / calibration / preventive maintenance were conducted at the proper frequency; and,

--

machinery history records and related surveillance / calibration /

preventive maintenance records were retrievabl .3 Findings / Conclusions 3. Hydrogen Recombiner During the review of the maintenance work on the "1A" hydrogen recombiner, the inspector observed operator align-ment of valves (including containment isolation valves)

_ _ _ _ .

. _ _ ____

.

. 11 associated with recombiner operatio For the scope of maintenance activity conducted, the blower was removed from its housing with pipe flanges taken apar The blower motor had to be energized for bearing temperature measure-ment Flange isolation valves were " blue tagged" shut, but the manual containment isolation valves (CIV's) HR-V-2A and 4A were opened. It appeared to the inspector that the CIV's were unnecessarily opened. The auxiliary operator explained that he followed the applicable operating pro-cedures given to him by shift managemen The problem appeared to be that the operating procedure did not address the off-normal maintenance activity which also required energization of the blower motor outside of its housing. When this was discussed with the Plant Operations Director, he acknowledged that a review of this condition was warranted to avoid unnecessarily manipulating manual CIV' Later in the inspection period, operations personnel sub-mitted PCR No. 1-MT-87-1009 to revise the applicable pre-ventive maintenance procedure (M-148). The proposed change avoids manipulation of the CIV's and, in the interest of efficiency, suggests that the PM and quarterly leak tight-ness tests be scheduled together. The maintenance depart-ment was processing that change. The licensee's planned action alleviates the inspector's concer The inspector concluded that containment integrity was maintained although the operating procedure used did not address the off-normal conditions for the maintenance activit The inspector also noted that the record for the recombiner leak tightness test (SP 1303-11.46) for the May 5, 1987, maintenance activity was missing. This is unresolved pend-ing licensee recovery of the subject records and subsequent NRC Region I review (289/87-10-02).

3. Relief Valves for Steam Turbine The licensee resolved the problem of inadvertent challenge i to the relief valves for the steam-driven emergency feed-i water (EF) pun.p s as noted in NRC Inspection Report No.

i 50-289/87-09. The licensee replaced the steam safety relief l valves MS-V-22A and B during the eddy current outage of spring 1986. Licensee actions are embodied in Technical Data Report (TDR) No. 828, "TMI-1 Turbine-Driven Emergency Feedwater Pump. Testing." The new safety salves were an L __ _ _ - - . - . . - -- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - . - - - - - - - -

.

_

-

-

integral part of the resolution to assure the steam safety reliefs (set at 300 psig) would not lift on subsequent start of the turbine-driven EF pump under normal condi-tions. Also, the instrument air controller for the steam admission pressure regulator (MS-V-6) was replaced with an integral and proportional controlle Testing, during plant shutdown for refueling in October, confirmed positive results with no challenges to the safety relief valve The inspector reviewed the above-noted TDR and test results but also inquired on the periodic testing of these valves with the Preventive Maintenance Manager. The inspector concluded that the licensee had properly installed the i steam safety relief In addition, the inspector noted that these valves had been placed in the preventive main-tenance program and would be tested / repaired on a five year cycle. The inspector had no additional comments in this are .3.3 Main Steam Safety Valves I The inspector reviewed recent repairs and testing that were completed on the Main Steam Safety Valves (MSSV's). Also a review was made of the test and maintenance history since 1983. The main steam systerr uses eighteen MSSV's on the main steam lines between the Once-Through Steam Generators ,

(OTSG's) and the main steam isolation valves (nine per l OTSG). The valves function to relieve pressure in the ,

OTSG's at varying setpoints between 1040 and 1092 psi l The capacity of all eighteen MSSV's at full flow is greater l than the 100 percent output of both OTSG's. Normally, dur-ing reactor trips from full power, the MSSV's will lift to assist in reducing plant temperature prior to the main steam turbine bypass valves modulating open, which occurs l at approximately 1010 psig. This corresponds to a plant temperature of approximately 548 Additionally, two power-operated relief valves, MS-V-4A/B, can function to assist in pressure relief at approximately 1020 psig, although their combined flow capacity is only approximately 6 percent of total OTSG outpu During the reactor trips in Cycle 5 operation, it was ob- '

served that the MSSV's lifted as expected to relieve OTSG pressure, but then some valves reseated and then lifted one i or two more times (see unresolved item 85-25-05, paragraph 5.2). It was determined at that time by the licensee that operator action to control pressure via the turbine bypass valves was not adequate or timely enough to permit a pos-itive reseat of the MSSV' _ _ _________.--------------- _ J

_ __ _ __

l

.

v

-

The present design of the primary and secondary relief valve scheme places added emphasis on the functioning of the MSSV' The' lift point of the primary PORV, RC-RV-2, has been increased to the point (2450 psig) where it is no longer an integral part of the plant response to a reactor tri The setpoint is too high to allow it to operate to ;

reduce primary pressure and temperature following the '

reactor trip. Additionally, the high pressure reactor trip is set below the PORV lift setpoint. This places increased importance on the functioning of the MSSV' Maintenance machinery history records were reviewed for MSSV maintenance conducted during 1983 and 1985. Job Tickets for the maintenance conducted 6R (March 1987) were-also reviewed. During 6R, fif teen of the eighteen MSSV's were overhauled ano one was replace During previous maintenance (1983/1985), varying degrees of work was re-quired to restore the valves to functional status. Many of the valve seats were pitted and internal parts were exces-sively corrode In one instance, in 1987, MS-V-20B was found with frozen internal parts and one valve MS-V-17B required replacemen During testing of the valves in 1985 during restart and ir, October of 1986, just after shytdown for 6R, many of the valves were found to lift signWicantly below the required set pressure Also, during the pre-outage testing in October 1986, many of the valves had evidence of seat leakage (valve weeping after reseat).

It appears that over a period of time between maintenance, !

the valves slowly degrade due to either seat steam cutting I or general corrosio This results in valve heating and subsequent lif t pressures drifting lower than required. It appears that this generally lower set pressure for over half of the valves discovered during October 1986 testing was a significant contributor to the cycling of the valves during the Cycle 5 reactor trip This condition would i have required the operators to manually control main tur- I bine bypass valves at lower than the 1010 setpoint in order to ensure positive reseat of the MSSV' The inspector's review of the recently completed mainten-ance (6R) and testing (March 1987) revealed that generally all valves (16 of 18 were repaired and tested) required i some work due to frozen parts, excessive corrosion and seat steam cutting and pittin Final test results per SP ,

1303-11.3 accomplished on March 23, 1987, were satisfactory I with all valve final setpoints being in specificatio j i

_ _ _

'

,

l i

l

-

i During the reactor trips on May 1-2, 1987, the MSSV's that did lift, reseated properly and did not cycle. Lift times were generally less than one minute. One valve did lift for a much longer period of time prior to reseatin This was not determined to be a problem with the valve, but that the blowdown setpoint may have been set lower than the i pressure being maintained via the turbine bypass valves at approximately 980-1000 psig. This problem was noted by the licensee and will be corrected at the next available oppor-tunit i It appears that the maintenance that was accomplished on the MSSV's during the 6R outage has successfully corrected the seat-weeping problems. The inspector discussed the recent maintenance and testing of the MSSV's with licensee engineering personnel. The conclusions about the setpoint drift and valve cycling generally were that valve heating l due to excessive seat leakage was the cause of. the proble ,

The recent maintenance appears to have temporarily correc-

'

ted the proble The MSSV's are not presently included in a preventive main-tenance progra The valves -are only overhauled when either excessive leakage is present or the valves cannot be successfully adjusted. The licensee is presently reviewing past data on valve lif t and repairs accomplishe Engi-neering personnel will make recommendations on further valve planned maintenance schedules at a later dat !

,

The inspector concluded that the the present time the past problem with the MSSV's does not present a concern, al-though a planned preventive maintenance program for these j valves seems to be warranted based on previous data. The ;

inspector had no other concerns in this area. It should be 5 noted that the B&W Owners Group is also studying this prob- ]

lem (see section 5.2). Future licensee and B&W Owners i Group action on the MSSV's will be reviewed in future )

inspection .3.4 Diesel Generator Air Start Check Valves i During the inspection of safety-related spaces, the inspec- ]

tor observed the automatic charging of the diesel generator j

'

air start receiver The charging lines each had check valves (EG-V-10A/B (A/B)) which apparently seated when the compressor unloaded after the recharging cycle. Based on observations of the receiver pressure gauges, it appeared that check valve seating was ef fective in holding receiver

_ _ _ . ___ -_-_-_-_ _ -

L

-

l l pressure assuining that the piping between the check valves I

and the compressor was unloaded along with the compresso The inspector then concluded that the check valves had an apparent safety function, namely to seat and hold air re-ceiver. pressure above minimum pressure (approximately 225 psig) for satisfactory engineering safety features actua- l tion of the emergency diesel generator. He then querried plant engineering representatives on the design'of upstream piping and whether or not the check valves were in the NRC-staf f-approved inservice test (IST) program established by the license i Plant engineering personnel reviewed the matter and pro-vided additional information on the design and operation of the subject system They confirmed that the upstream piping was depressurized when the compressor unloade That piping was designed to seismic Category I, but the compressor qualifications were not clearly establishe The check valves were not in the IST program; but licensee personnel noted that, based on operations shift tours, the proper seating of the check valves was indirectly verifie That is, operators check air receiver pressure and note unusual and frequent operation of the compressor indicating a leak in the air receiver system or possibly the subject check valve !

The inspector independently confirmed the additional infor-mation provided by plant engineering. However, it was not clear on why the check valves (EG-V-10A/B (A/B)) were not addressed in the NRC-approved IST program, along with the licensee's methodology of verifying the check vsive oper-ability to perform their apparent safety functio No commitments to change the IST program were offered by I licensee representatives. Accordingly, this matter is unresolved pending further NRC staff review in accordance with 10 CFR 50.109 (289/87-10-03).

3.d Operability Summary Licensee maintenance and the quality assurance department were also involved in this area. Maintenance managers were noted to be in plant spaces overseeing vork activities .

The lit.ensee was responsive to NRC staff concerns for those identi-fied during this inspection period and for the MSSV's abnormal per-formance on post-trips. Licensee effort on the MSSV's were note-worthy, i

_ _ _ - - - _ - _ _

__---_-- _----_ _ _---_-_ --_ _----_---_ _ _ _ ,

, *

I

.

l l

4. Event Review Introduction and General Scope of NRC Staff Review During this inspection period, there were several events that the NRC staff reviewed in detail. They were: the reactor trips of May I and 2,1987, and the Unusual Event of May 9,1987. In general, the fol-lowing aspects were considered for each of these events:

--

details regarding the cause of the event and event chronology;

--

functioning of safety systems as required by plant conditions;

--

consistency of licensee actions with licensee requirements, approved procedures, and the nature of the event;

--

radiological consequences (on site or off site) and personnel exposure, if any;

--

proposed licensee actions to correct the causes of the event;

--

verification that plant and system performance are within the

' limits of analyses described in the Final Safety Analysis Report (FSAR); and, p

--

proper notification of the NRC was made in accordance with 10 CFR 50.7 For each of these events, the inspector provided a chronological /

factual summary, specific scope of NRC staff review, licensee find-ings and NRC staff findings. An overall conclusion on licensee per-formance is also provide .2 Reactor Tr Qs 4. Background Information '

Since the Cycle 6 startup (March ,23, 1987), the reactor power output was limited because of high water level limits in the steam generator due to heat transfer fouling. The licensee previously reported that the fouling material was primarily magnatite (an iron corrosion product), which was flaky or loose, clogging the broach hole openings between the tubes and tube support plates in the SG's. The mate-rial could be redeposited (to the bottom of the SG instead of at the broach holes) by a substantial pressure transient in the SG as that which occurred on previous (Cycle 5)

trip !

l

.

.

The licensee's previous experience was these trips had the ,

side benefit of redepositing the fouling material. ' Conse- i quently, the licensee approached NRC staff (NRR and Region l I) during a conference call of April 24, 1987 to discuss l any concerns regarding a planned turbine-to reactor trip.

l The trip was proposed primarily to let the SG's experience the pressure transient which is normal for such trips. The licensee felt this pressure transient would redeposit foul-ing material. The licensee presented their plans for con-trolling the evolution Essentially, the emergency pro-cedure for a reactor trip would be used, but other evolu-tions were planned such as initiating emergency feedwater in a controlled manner and securing reactor coolant pumps

--

all with the intention of inducing transients on the SG's to loosen the fouling material. A special tempora ry procedure would be written as an overall controlling docu-ment and a 10 CFR 50.59 evaluation would be writte The licensee also hoped to test the installation of the elec-tronic filter circuit installed in Heat Sink Protection System (HSPS) SG level indication cabinet The NRC staff noted that the planned trip was unprecedente From a probabilistic viewpoint, these evolutions were in-tentional challenges to safety systems. But, they had the side benefit of proving their operability, if successfu The NRC staff had no immediate safety concern with the licensee's plans, but they remained skeptical both on the future use of intentional trips to solve an economic prob-lem and on its generic implications. Accordingly, the NRC ;

staff saw no cause to stop the licensee from proceeding, l but they also obtained a commitment from the licensee to address in a letter to the NRC staff their long-term plans to effectively resolve the fouling problem without planned i reactor trips. The staff also obtained . a commitment that if the trip was successful, another trip would not be at- I tempted unless they discussed the problem with the NRC staff agai .2.2 Event Chronology At approximately 5:00 on May 1, 1987, the licensee conducted a planned trip of the reactor by initiating a turbine trip for reasons addressed cbove. The licensee's preplanned sequence / procedure also included manually initiating emergency feedwater, manipulating SG level to the natural circulation control setting (50*s on the opera-ting range), and securing reactor coolant pump l

_ - _ _ _ - - _ _ _ . l

t-l-

.

,

'

No significant problems were identified during these evo-lutions and during the licensee's post-trip review. Minor equipment problems occurred and corrective actions were scheduled for either prior to or after startup as addressed belo i The licensee made the reactor critical at 9:40 p.m., May 1, ;

1987, and they had the plant at a steady-state power of 90 l percent by 6:00 a.m. , May 2,1987. The trip was apparently effective in alleviating the power / level limit temporari"v; but, plant data still indicated fouling, more so with tne

"B" SG similar to just prior to the tri !

With steady-state conditions established, the licensee started a nuclear instrumentation (NI) calibration using a plant heat balance calculatio At 7:51 a.m. on May 2, 1987, the plant tripped on high reactor coolant system (RCS)

pressure due to operator error during the conduct of the above-noted calibration. The NI channel in test was inad-vertently left feeding the integrated control system (ICS)

and a technician dropped the NI channel in test to zer The ICS responded by increasing reactor power and, by the cross limits function, significantly reduced- feedwater demand flo The RCS responded to the loss of feedwater and the high RCS pressure situation resulted. There was no initiation of emergency core cooling systems and the emerg-ency-feedwater system was not challenged, i The post-trip plant response was normal with no significant problems identified during the post-trip revie Operator performance problems were addressed by the licensee as noted belo .2.3 Specific Scope of NRC Staff Review for the Reactor Trigs Specific to the reactor trip events noted above, the in- 1 spector verified the below-listed items:

--

initial proper response of the plant to the post-trip window on the pressure-temperature (P-T) plot;

--

personnel properly implemented ATOG procedures and prudently acted on unusual conditions; j

--

identification of the sequential proximate causes for the trip along with a reasonable determination of the root cause; j i

._-. _ _ _ _ _ - _ _ _ _ - _ _ _ _ ______________-__ - - _-__ - __

.,

-

--

post-trip review was conducted in accordance with Administrative Procedure (AP) 1063, " Reactor Review Process;" and,

--

no unreviewed safety issues identified in post-trip review dat In addition to discussions with cognizant licensee person-nel, the inspector:

--

made an independent assessment of post-trip parameter response basea on visible strip charts and indicators in the control room shortly after the events;

--

attended the licensee's post-trip review;

--

reviewed the complete post-trip review packages (No !

87-01 and 87-02); and,

--

reviewed AP 1063, " Reactor Trip Review Process" for adequac . Licensee Findings For the reactor trip and evolutions of May 1,1987, listed below is a summary of the licensee-identified problems /

findings along with licensee resolutions:

(1) One channel of nuclear instrumentation (NI-1) for the  ;

source acted erratically. An additional two channels .1 indicated expected power level for the reacto The licensee traced the problem to a loose 15-volt power supply cabl The licensee cleaned the connec-tors and re-adjusted the cable for startup later that evening and later replaced the cable (NI-1 responded properly during the trip of May 2,1987). The licen-see intends to replace the cable at the next plant shutdow (2) /pparently, the two controllers associated with emerg-ency feedwater flow control valve EF-V-30B to the "B" OTSG responded erratically to the controller demand signal from the control room and remote shutdown panel Prior to startup, the licensee later confirmed l

!

j' I l. : __ __________________________________9

_ _ - - _ _ _ _ _ _

l

.

l

-

proper operation and confirmed opera-tor unfamiliarity with the use of that particular con-trolle Due to subsequent licensee review, it was determined that the controllers at each station are of different desig The control room controllers for EF-V-30A/B/C/D were relatively new and put in during the cycle 6 outage along with the new controllers at ,

the Remote Shutdown Panel (RSP). for EF-30C/ The

'

older models were still installed for EF-30A/B at the RSP. Between the two designs, controller response is different, and this may have led to operator confusion l and the apparent erratic behavior Before the subsequent plant startup the licensee cau-tion tagged the two older model controllers to enhance operator awareness of these unique characteristic An internal licensee memorandum, dated April 7,1987, i (Serial No. 5520-87-0271) from project engineering authorizes the procurement and installation of new controllers for EF-V-30A/B at RSP similar to the other six controllers noted above. This outstanding work is to be tracked by incomplete worklist items for licen- l see Budget Activity No. 41224 l (3) The low SG level alarm (set at 23 inches on the start-up range) was received in the control room during operator difficulties with the EF-V-30B valv The low level limit controller is set at 30 inches. Act-ual level was 27 inches. No emergency feedwater (EFW)

actuation occurred, as expecte The licensee confirmed that one channel could have been that low with an alarm accuracy of i 4 inche EFW actuation does not occur until 10 inches on the startup rang (4) The Heat Sink Protection System (HSPS) OTSG level instrumentation responded properly with no evidence of level oscillation due to secondary plant pressure oscillations. (The licensee evaluated the recorded data and later placed the main feedwater (MFW) isola-tion function in enable on March 14,1987 (see para-graph 2.2.2)).

(5) The OTSG safety valves did not cycle as noted in pre-vious plant trips from high powe i

_ _ _ _ _ _ _ _ _ _ _ _ _ _

.

-

(6) The saturation monitor did not oscillate to below the low margin alarm of 25 F as noted on previous plant trip (7) No problems were encountered with the securing of the reactor coolarit pumps (RCP's) and manual actuation of the EFW syste (8) Overall, response of the plant was as expected without indication of interrupting decay heat removal (DHR).

For the reactor trip of May 2, 1987, listed below is a sum-mary of licensee-identified problems / findings along with licensee resolution (1) Operator error (inattentiveness) caused the wrong switch to be positioned, thereby sending the reactor protection systen (RPS) channel in test to controllers that were controlling the plan The licensee counseled the particular operator and his crew, in particular, and other operators / crews were later notiff ed by a written internal review of the

event.

l l The NI calibration procedure was reviewed to ensure measures were adequate to prevent recurrence. How-ever, enhancements were planned for to the NI cali-bration and RPS calibration procedures to positively confirm proper operator functioning of the switche Labeling for these switches were specifically reviewed by the Plant Operations Director (P00) and he conclud-ed that the labeling was adequate and the planned procedure revisions should be sufficient to preclude recurrenc (2) The operators did not perform the first step of the AT0G procedure - manually trip the reactor. However, they did confirm control rods had fully inserted in response to the automatic trip. The individual oper-ator and shif t foreman were specifically counseled by the POD and these events were also addressed in the event review documentation noted abov The licensee's POD gave a personnel counseling session  ;

to the crew involved prior to startu (3) Plant response was normal, as expected.

l l

l

!.

I 1

__

_ - _ _ _ -

i

.

-

For both trips, total radioactivity (mostly Xe-133) re-leased as a result of the steam generator safety valve actuation was 16.8 microcuries, which was well' below tech-nical specification (TS) limit .2.5 NRC Findings 4.2. General - Both Trips The inspector independently confirmed the licensee find-ings/ conclusions as noted above. Plant response was es-sentially as expected with minor problems note The licensee adequately identified these problems and took/

planned appropriate and reasonable action for immediate correction and to prevent recurrenc I

'

Operator responses to the planned evolutions and/or off-normal conditions were essentially consistent with facility operating and emergency procedures. It appeared that they were conscious of and oriented their actions toward con-firming reactor shutdown conditions and adequate decay heat removal. Management involvement in these events and post- t trip action reviews were noteworth For the trip of May 1, 1987, no formal NRC notification was required; however, as an initiative / courtesy, the licensee incorporated a step in their teinporary procedure to notify an on-site NRC resident inspector before the trip, which i they did. A 10 CFR 50.72 report was appropriately made for

'

the May 2, 1987, . trip and a licensee event report was expected to be issue The AP 1063 was adequate to identify / confirm the root causes of the reactor trips and the post-trip reviews ~ were reasonably thorough to identify appropriate corrective actions before startu A substantial upper management ,

involvement in both post-trip reviews occurred and they included participation by the: Director of TMI-1; the {

existing and new Operations and Maintenance Directors; the Plant Engineering Director; and, the Plant Operations j Director, along with senior site / corporate engineers, a Further, Quality Assurance Department (QAD) representatives .

were actively reviewing events of the May 1-3, 1987, weekend, l l

I

,

i

_ _ _ _ - _ _ __ A

.

-

The planned trips of May 1-2, 1987, had the side benefit to confirm proper functioning of safety system .2. Planned Trip The special temporary procedure (STP) was developed for the May 1,1987, trip and its accompanying 10 CFR 50.59 evalua-tion was reasonably thorough to assure safety and effect coordinated actions during the planned sequence of event Essentially the emergency procedures were used on trip response. Also, the securing of RCP's and EFW actuation were also addressed. Additional crews were brought in to assis The procedures were strictly followed. As a re-sult, good control of the events occurred by licensee per-sonnel and managemen The crews were well prepared and briefed on expected ac-tions and their responsibilities. Site and corporate tech-nical support personnel were also present. Videotaping of the steam generator tailpipes confirmed licensee corrective action to prevent cycling / challenges noted on previous trips during Cycle 5 (see also Section 3).

The licensee reported that in plant chemistry sampling of SG water did not identify loose or suspended material in the samples taken. However, still pictures of the SG re-lief valve tailpipes revealed brownish material existing momentarily from selected SG safety valve tailpipe Cor-rective actions for this trip were appropriate and/or ade-quately tracked for completio .2. Unplanned Trip In responding to the event, while outside in the yard area, the inspector noted steam generator safety valve perform-

, ance and he confirmed, at least during the brief time per-

iod, no chattering of the safety valve Operator response, in general, was good as noted above.

l However, they failed to follow the first step in the AT0G Procedure 1210-1, but they also met the intent of the ATOG procedure by confirming the reactor trip effectively shut down the reacto The inspector considers the licensee's methodology for fol-lowing the ATOG procedures to be noteworthy. It is stand-ard practice that on a reactor trip immediate actions are taken by the operator while the shift foreman reads the

_

__- _ _ _ _ - _ _ _ __ - _ _ _ _

r

-

'24 l

l procedure and confirms actions taken. During Cycle 5 oper-ations, the inspector noted this to be effective in ensur-ing proper implementation of the AT0G procedures. The failure to perform the manual trip action had minimal safety consequences and it was uncharacteristic of licensee past performance in this specific area of procMure adher-enc Accordingly, the inspector considered this an iso-I lated case and a licensee-identified violation of TS 6. (289/87-10-04). The inspector considered this item closed based on the licensee corrective actions / measures to pre-l vent recurrence and the review noted belo The isolated, but negative operator performance aspects, were: (1) Cognitive error by a licensed senior reactor operator (SRO) to confirm his actions in selected remote panel shutdown (RPS) inputs to the integrated control sys-l tem (ICS) and (2) the failure of a licensed reactor oper-l ator (RO) to follow the reactor trip emergency procedur The POD appropriately addressed these issues in an internal memorandum dated May 19, 1987, to the shift crews.

l The inspector also reviewed the procedure enhancement for the operator to specifically confirm which channel of RPS is selected to the ICS during test conditions. The inspec-tor was satisfied with this action based on a review of the l below-listed documents. The confirmation b accomplished l by observing auxiliary relay energized / de-energized states in an RPS cabine Surveillance Procedure (SP) 1302-1,1, Revision 24, dated May 22, 1987, " Power Range Calibration;

--

SP 1303-4.1, Revision 53, dated May 22, 1987, " Reactor Protection System (Monthly Surveillance);" and,

--

Operating Procedure (0P) 1105-2, Revision 24, dated May 22, 1987, " Reactor Protection System."

Also, the licensee reviewed the human factors labeling of the subject switches for inputs to the ICS. They concluded that the labeling was adequate and that operator inatten-tiveness was the root cause of the trip. Measures to pre-vent recurrence included procedure enhancements to the RPS surveillance procedure and the nuclear instrumentation  ;

calibration procedur l The inspector similarly concluded on the root cause of the event and he was satisfied with the licensee's proposed corrective actions or measures to prevent, recurrence for this trip were adequately addresse i I

_ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _

_ _ _ _ - _ _ _ _ - _ _ .

.

-

4.3 Unusual Event of May 9, 1987 4. Chronology At 11:30 a.m. on Saturday, May 9, 1987, the licensee de-clared an unusual Event based on the loss of the NRC emerg-ency notification (ENS) line and loss of power to a sub-stantial portion of the 717 area code (Harrisburg, Pennsylvania) telephone system. The commercial line fail-ure degraded their emergency communication capabilities to implement their emergency plan. The problem was attributed to the power failure of the Bell of Pennsylvania system at about 9:00 a.m. (with full service restored at about 2:00 p.m. that same day) and that affected the ability to tie into the "AT&T" system for long distance call The licensee was able to establish communications with the NRC Operations Center by using their Reading dispatcher in area code 215 for the required 10 CFR 50.72 reports. With com-mercial line service essentially restored, the licensee secured from the unusual event at 2:20 . Scope of NRC Staff Review No NRC staff response was needed for this event. A post-event follow-up inspection on the licensee's review of the event was conducted to determine what lessons learned were identified and/or what corrective actions would be appro-priat The inspector reviewed control room logs on the events and l discussed the matter with the TMI Emergency Preparedness Manager. The inspector also reviewed the licensee internal review of the event, Memorandum No. 6440-87-116, dated May 27, 1987, from TMI Emergency Preparedness Manager to the Director, Nuclear Assuranc . Licensee Review The root cause of the problem was with equipment not di-rectly under the control of the licensee. Licensee repre-sentatives entered into discussion with the appropriate telephone company who gave the licensee assurance that equipment modifications would be made to prevent reoccurr-ence. The licensee learned from the telephone conpany that the problem was due to loss of d.c. power because of a sequential loss of rectifiers causing d.c. batteries to discharge, compounded by failures to place standby recti-fiers on the line. Although the licensee's emergency com-munications system uses dedicated lines (like the NRC's ENS line), these lines were still powered from the centrally located power supplie :,

.

'

.The ~1icensee 'revi;wers four.o that operator response was overall good despite lack of. notification by.the telephone Ho. wever, some lessons were ' learned with, respect compan to the adequacy of contingencies ' for such . communication problems, :The licensee internally established the below-listed action item Investigate feasibility of establishing a radio com-munication- link with PEMA (Pennsylvania Emergency ManagementAgency).

Enhance the listing of all known alternate means of

~

--

communications in the emergency plan implementing procedure .;

--

Ensure lessons learned during this event are passed on to appropriate licensee staff (i.e., emergency / support directors).

4. NRC Staff Findings / Conclusions The licensee review of this event was reasonably thoroug It appears that appropriate l e s'.,on s learned / corrective actions were documented for later' follow-up. The inspector was satisfied with the licensee review and Region I may review the results of the licensee's actions in a subse-quent inspection report.

'

4.4 Event Summary Overall, operator response to off-normal events were oriented toward-safety and, in general, in accordance with facility procedures. The licensee-identified nonadherence to the~ reactor . trip emergency pro-cedure ' appears to be an isolated case and the licensee management took appropriate corrective actio The planned reactor trip was preplanned, procedurally control l led, and additional licensee staff was utilize Licensee management and quality assurance department attention and involvement in the reactor trips of May 1-2, 1987, were noteworth post-event reviews were reasonably thorough with corrective _ action

_

appropriate identified, documented, and evaluated for impact on plant operation I For the two reactor trips, plant ' response was as expected. When required, safety systems appropriately functioned. There were no challenges to the emergency core cooling system {

...

i i

l a

_ __

.

<

2 . Licensee Actions on Previous Inspection Findings 5.1 (Closed) Inspector Follow Item (289/85-SC-04): Revise Electrical Interlock for Fuel Handling Building Crane to Limit Path Travel An issue remaining to be resolved regarding Generic Task A-36, " Con-trol of Heavy Loads Near Spent Fuel," consisted of licensee action to modify the Fuel Handling Building (FHB) crane to limit its travel over certain areas of the spent fuel pool. The licensee completed the required modifications for the new FHB crane travel interlocks system and tested the new interlocks per Test Procedure (TP) 400/ The inspector reviewed the documentation associated with this modif-ication, including Installation Specification TI-IS-412481.001. Also, the completed data from TP 400/0.1 was reviewed. These modifications will prevent the FHB overhead bridge crane from traveling over areas of the spent fuel pool that could possibly contain spent fuel in the future. The only movement that is permitted is movement of an area of the pool that could be used to load spent fuel into shipment can-ister No evolutions of this type have been accomplished or are planned at TMI-1 at the present tim The inspector concluded that movement of the FHB overhead crane has been sufficiently restricted to as to prevent lifting of the type of loads that, if dropped, could damage spent fue The inspector had no other concerns and this item is close .2 (Open) Unresolved Item (289/85-25-05): Ma:n Steam Safety Valve (MSSV) Abnormal Performance This item was opened to document problems with MSSV operation after reactor trips and was updated in NRC Inspection Report Nos. 50-285/

85-30 and 86-13. The performance of the MSSV's a,d the turbine by-pass valve during the reactor trips of May 1-2, 1987, was discussed in paragraph 3. It appears that although MSSV operation was more normal during these trips than during the June 2, 1986, trip, the fundamental problem with MSSV turbine bypass valve interaction has not been corrected. The recent maintenance and testing of the MSSV's has contributed to more acceptable MSSV's lift and reseat action, but the turbine bypass valves (TBV's) still do not fully control steam pressure. after the trip, although this time (May 1-2, 1987), the TBV's did get full open demand signals after the trip. Accordingly, this item will remain open pending further licensee and B&W Owners Group study of this proble NRC Region I will continue to follow i any recommendations that come from the Owners Group reassessment progra !

- - - _ - . . . _ _ - - _ . . _ _ _ _ - _ _ _ _ - - _ - - - _ _ _ _ _ _ - - }

!

.

i

!

5.3 (0 pen) Unresolved Item (289/85-26-06): Nuclear Service Valve Penetration Design Adequacy Th6 nuclear services closed-cycle cooling system (NS) piping to reactor building fan motor coolers is seismic class I piping and it penetrates the containment in the intermediate building. The pre-vious inspector questioned the design location of NS relief valves NS-V-36A/B/C, since they were located between the outside containment wall and the NS containment isolation valves. The relief valves are nominally set at 175 psig and the lines are needed for motor cooling since this piping system is seismically designed and needed for con-tainment post-accident mitigation. The design is in accordance with Final Safety Analysis Report (FSAR) commitments on which the opera-ting license was issued in 197 However, the design does not meet current general design criteria (GDC) No. 57 (10 CFR 50 Appendix A) in that containment isolation valves for closed-cycle systems in the containment either be auto-matic or locked closed or capable of remote isolation. A relief valve is none of the abov The licensee committed to review this concern and document their findings in an internal memorandum (dated November 7, 1986, from Reactor Plant Manager to a GPUN licensing engineer).

The inspector, during this inspection and with the aid of a region-based specialist during the last reactor building - (RB) integrated leak rate test (ILRT), independently reviewed the piping arrangement and location of the NS-V-36A/B/C for possible adverse effects on plant safet In their internal memorandum, the licensee reviewed the purpose of the relief valves and a possible accident scenario for which the relief valves might adversely functio The purpose of the relief valves is to relieve the thermal expansion in the cooler / piping with inlet and outlet valves shu If, during normal or accident situations, with no NS pipe break, the relief valve opened and stuck open, no radioactive fluid would be released since the NS is a closed-cycle water system inside containment. The licensee's documentation also proposed an accident scenario in which NS piping integrity was degraded with design pressure in containment at 53 psi No release would occur since the setting is 175 psi A postulated NS pipe break with containment pressure at the relief setting is beyond the design basis event. Accordingly, this aspect ;

of the unresolved item is resolved, i

- _ _ _ _- - - - _ -

_ _ _ _ _ .

a

'

'However, during this review, the inspector again noted the inaccurate depiction of the particular NS arrangement. The updated SAR Table 5.3-2 (Sheet 5 of. 8, Update-4 7/85) specifies the ,NS penetration as Valve Arrangement No. 23 in SAR Figure 5.3-1 (Update-3 7/84). Valve Arrangement No. 23 does not show the subject relief valve between the outside isolation valve and the outside containment wall. The in-spector noted .that this error is the only inaccuracy of this section of the SAR and that it was originally identified when the inspection finding was identified. To date, the licensee has not corrected the erro A 1icensee representative indicated that the system drawing in other sections of the SAR was accurate and that Figure 5.3-1 drawings are only " general" arrangements. Further, the individual stated that the licensee's position on the matter was that a change was not needed since none of the requirements of 10 CFR 50.71(e) were meant to re-quire a change to the SAR; namely, the design is part'of the original design and no change occurred nor was a safety evaluation conducted in support of a licensee amendment or as directed by the Commissio This item continues to be unresolved penaing the NRC staff (Region I)

review of the licensee's position on changing the subject section of the SA .4 (Closed) Unresolved Item (289/86-17-04): NRC Review of High Pressure Injection (HPI) Flow Indication Response During HPI Flow Surveillance This item was opened in response to Information Notice 85-100 t' t informed licensees of a possible shift in setpoint for cer. a n Rosemont delta-P transmitters. The HPI flow transmitters are of this type. The inspector observed the response of the flow transmitters during tne conduct of SP 1303-11.8, Revision 17, "High Pressure In-jection," as documented in NRC Inspection Report No. 50-289/86-2 The flow transmitter readout in the control room appeared to be nor-mal for the flows that were being maintained during the tes It appeared that the delta-P transmitters responded properly during throttling of flow through each line. Since the test completion and subsequent outage, four HPI flow transmitters have been reading 0, 30, 40, and 50 gpm in the control room. This is with no flow in the lines. It appears that although the control room readout is not i totally accurate with no flow; it is acceptable when the instrument  !

is used with flow presen Accordingly, this item is close The inspectors still have concerns over the general performance of the Rosemont delta-P transmitter and the resultant "zero shift" that occurs when system pressure is increased greater than the calibration pressur This concern is being tracked by unresolved item 289/

87-09-0 __________________________________o

_ _ _ _ _ . __ _ _ _ _ _ _ . __ __ -. _ - . _ _ _ _ _ _ - _ _ _ _

_ _ _ _ _ _ _ _ _ - - _ _ - _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _

a

-

Unresolved Items are matters about which more information is required in order to ascertain whether they are acceptable, violations, or deviation Unresolved items discussed during the- exit meeting are addressed in para-graphs 2.2.5, 3.3.1, 3.3.4, and Inspector Follow Items are significant open -issues warranting follow-up review by the inspector to properly disposition the matter. An inspector

' follow-up item was addressed in paragraph _ - _ _ _ _ _ .

_ _ _ _ _ _ _ _ _ _ _

.

'

I 5.5 '(0 pen) Unresolved Item (289/87-09-10): Remote Shutdown Panel Relay Defect The inspector reviewed the 10 CR 21 report in detail in NRC Inspec-tion Report No. 50-289/87-09. New relays were to be installed during ]

the next cold shutdown. The licensee also committed to several other j inspection and testing action for the new relay. As an interim meas-  !

ure, facility procedures were to be revised to address necessary-  !

manual actions in case a relay f ailed during remote shutdown panel l us By Revision 31, dated April 15, 1987, the licensee revised OP 1202-37, "Cooldown from Outside the Control Room." This revision included adding Attachment 5 to the procedure. The attachment provides block-ing instructions and a correlation of relay numbers to essential com- 1 ponent designators. The licensee reported training on these instruc- {

tions was completed. The inspector was satisfied with the licensee's  !

interim procedural measure This item remains open pending completion of licensee final actions and subsequent NRC: Region I revie .6 Past Inspection Findings Summary Overall, the licensee was responsive to address previous inspection

'

issues / concerns 6. Exit Interview The inspectors discussed the inspection scope and findings with licensee ]

management at a final exit interview conducted May 29, 198 Senior licensee personnel attending the final exit meeting included the following:

J Colitz, Plant Engineering Director, TMI-1 K. Harkless, 10SRG Engineer H. Hukill, Director, TMI-1 i M. Knight, Licensing Engineer, TMI-1  !

S. Otto, TMI-1 Licensing  !

T. Seaver, QA Auditor D. Shovlin, Manager, Plant Maintenance ]1 P. Sinegar, Administrative Plant Maintenance I R. Toole, Operations and Maintenance Director, TMI-1 The inspection results as discussed at the meeting are summarized in the cover page of the inspection repor Licensee representatives indicated that none of the subjects discussed contained proprietary or safeguards informatio l

_ ___.__ __ _ ___ _ J