05000335/LER-2003-003

LER-2003-003,

Event date:
Report date:
Reporting criterion:	10 CFR 50.73(a)(2)(ii)(B), Unanalyzed Condition
3352003003R02 - NRC Website

FACILITY NAME (1) DOCKET NUMBER (2) LER NUMBER (6) PAGE (3)

Description of the Event

On September 25, 2003, St. Lucie Unit 1 was in Mode 1 at 100 percent reactor power.

Quality Assurance (QA) personnel were conducting a scheduled triennial fire protection audit. The audit team included two qualified fire protection engineers.

Their review of the Gold Bond qualification fire test reports for 16 one-sided marinite board cable tray penetration seals [EIIS:PEN] concluded that the fire test was inconclusive based on the following:

• The type of the cable used in the tested configuration was not- identified and therefore cannot be compared to the cables installed in St. Lucie Unit 1.

• The test described burning cables but did not identify which side of the penetration the burning was occurring. It was probable that the fire was on the exposed side but it is not so stated.

• After the test was started, the 6th and 7th configurations of the test penetration were modified by adding Cerafelt. Therefore, a steady state condition was not maintained throughout the test.

• Per the original test criteria of ASTM E-119-73, cotton was to be at the penetration from the start of the test. On tested configuration number 7, the cotton was not placed at the penetration until 85 minutes into the test at which time the cotton ignited at 88 minutes (i.e. 3 minutes later) versus the required 180 minutes.

• For tested configuration number 6, the cotton ignited at 135 minutes versus the required 180 minutes.

In addition to the above, the Gold Bond test report for the subject penetrations indicated that the test was an "exploratory" test, not an official fire test. The 16 penetrations in question are designated 430-BW-159-001, 430-BW-159-002, 430-BW-159- 003, 430-BW-159-004, 430-BW-161-002, 430-BW-161-003, 430-BW-162-003, 430-BW-162-004, 430-BW-162-005, 430-BW-167-001, 195-S-003, 195-S-004, 195-01B-001, 195-01B-002, 195- 01B-003, and 430-S-004.

A condition report was initiated and, in accordance with the Fire Protection program, fire impairments were logged against the 16 fire penetration seals. FPL verified that no new fire watches were required to cover the affected fire areas/zones and the subject penetrations were included in the fire impairment log.

A subsequent walkdown of the Unit 1 Control Room floor penetrations was conducted in support of the engineering disposition for the discrepant penetrations. The walkdown results revealed that several conduits did not have any mastic coating extending above the face of the penetration seal as well as cases of cable in contact with conduits. Both of these conditions are deviations from the configuration tested at Omega Point Laboratories. The evaluation of these 34 additional deviations was completed on February 6, 2004, when this condition was determined to be reportable.

During the St. Lucie Unit 1 SL1-19 refueling outage FPL was performing maintenance to improve the integrity of the control room envelope by sealing potential sources of in-leakage. Of the 34 additional fire penetrations identified in revision 1 to this LER, 14 penetrations were within the scope of control room envelope maintenance and were located in the reactor turbine generator control boards (RTGBs). During the maintenance activities, FPL identified that two wire-ways within the RTGB fire stops were found to have the Transite damming board visible and were not provided with any FACILITY NAME (1) DOCKET NUMBER (2) LER NUMBER (6) PAGE (3) St. Lucie Unit 1 05000335 fire seal material as previously assumed. Other anomalies were found within the RTGB fire stops relative to small voids or imperfections created in the sealant material during the rework of these seals due to subsequent cable installations.

Cause of the Event

The one-sided marinite board penetration fire testing conducted by Gold Bond was performed in 1975. In 1976, as a result of the Browns Ferry fire, the NRC issued Branch Technical Position (APCSB) 9.5-1, "Fire Protection Program." On February 24, 1977, the NRC issued Appendix A to Branch Technical Position APCSB 9.5-1; "Guidelines for Fire Protection for Nuclear Power Plants Docketed Prior to July 1, 1976.

Appendix A to Branch Technical Position 9.5-1 provided guidance regarding cable and cable tray penetration seals in that the seals should provide protection equivalent to the associated fire barrier, and as a minimum meet the requirements of ASTM E-119, "Fire Test of Building Construction and Materials," including the hose stream test.

On May 11, 1976, the NRC requested by letter that FPL conduct an examination to compare existing fire protection provisions at St. Lucie Unit 1 with the guidelines in Standard Review Plan 9.5-1. FPL, using the guidance from Appendix A to Branch Technical Position 9.5-1, responded on March 31, 1977 to the NRC request. FPL described the Gold Bond test for penetration fire seals. FPL described the Gold Bond test as "conducted in accordance with the standard methods of ASTM E119-73, with exception of the hose stream test... The test demonstrated that, for the three-hour test period, each fire stop design contained the fire, i.e., the fire did not pass through the fire stop.

The NRC issued a Safety Evaluation Report with Amendment 33 to the Operating License on August 17, 1979. In the SER, the NRC recognized that "tests of the typical types of electrical cable penetrations were conducted at Florida Power & Light's Material Test Laboratory and at outside test facilities... The testing program included tests in accordance with the requirements of ASTM E119-73 with exception of the hose stream test nozzle. The tests demonstrated that the fire barrier penetrations would resist the passage of fire for at least 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> and no water penetrated the test wall during the hose stream tests.

Extensive search through historical microfilm was performed, but FPL test reports or insights into decisions and discussions involving the NRC were not found. FPL was unable to determine what the NRC reviewed or what additional testing was performed at FPL's Material Test Laboratory and provided to the NRC. Therefore, it is unclear what the bases were for utilizing the Gold Bond fire tests as the qualifying tests for cable tray penetration fire seals at St. Lucie. What is clear from a review of the Gold Bond test report, and other records from that era, is that the documentation requirements in the 1970's were not as stringent as are required today. In addition, the Omega Point Laboratories control room floor fire penetrations were not installed per the tested configuration.

Analysis of the Event

Per NUREG-1022, "Event Reporting Guidelines 10 CFR 50.72 and 50.73," Revision 2, Section 3.2.4, missing fire barriers such that the required degree of separation for redundant safe shutdown trains is lacking are reportable as an unanalyzed condition that significantly degraded plant safety. At the time of discovery 1-hour roving fire watches for the affected areas were already in place for other reasons (and had been in place for some time). Therefore, this event was not reportable under 10 CFR 50.72, but is reportable under 10 CFR 50.73(a)(2)(ii)(B) as an LER.

�FACILITY NAME (1) LER NUMBER (6) PAGE (3)NUMBER (2) L� —� - Analysis of Safety Significance The extent of condition review concluded that the fire qualification test reports for remaining St. Lucie cable tray fire penetration seals were valid. This review also concluded that there were no references to the subject Gold Bond test report as the qualifying test report for any cable tray penetration fire stops on Unit 2.

These penetrations are part of barriers that provide separation between redundant trains required for safe shutdown of the plant in the event of a fire. The affected fire barriers are the walls between the following fire areas:

• Fire Area/Zone B/57, Cable Spread Room to Fire Area/Zone Z/83, Reactor Auxiliary Building West Stairwell (Block Wall 167)

• Fire Area/Zone B/57, Cable Spread Room to Fire Area/Zone A/60, "A" Switchgear Room (Block Wall 162)

• Fire Area/Zone A/60, "A" Switchgear Room to Fire Area/Zone E/61, HVAC Equipment Area (Block Wall 159)

• Fire Area/Zone A/60, "A" Switchgear Room to Fire Area/Zone C56, "B" Switchgear Room (Block Wall 161)

• Fire Area/Zone C/55W, Main Hallway West (EL. 19.50'), Cable Loft Area to Fire Area/Zone 0/36, Main Hallway (EL. -0.50')

• Fire Area/Zone C55W, Main Hallway West (EL. 19.50'), Cable Loft Area to Fire Area/Zone D/78, "B" Electrical Penetration Room (West)

• Fire Area/Zone B/57, Cable Spread Room to Fire Area/Zone C/55W, Main Hallway West (EL. 19.50'), Cable Loft Area (EL. 28.67')

• Fire Area/Zone F/70, Control Room to the following Fire Area/Zones: B/57, Cable Spread Room; and C/56, 'B' Switchgear Room.

FPL originally submitted details of the fire endurance testing performed on cable tray fire stops to the NRC Staff per FPL letter L-77-102 dated March 31, 1977. In response to a request for additional information regarding the fire stops, FPL provided sketches per FPL letter L-78-315 dated September 29, 1978. The NRC documented that the cable tray fire stops met the objectives and were acceptable per NRC letter dated August 17, 1979. The basis for acceptance was the fact that the tests demonstrated that, for the three-hour test period, each fire stop design contained the fire (i.e., the fire did not pass through the fire stop). Testing of the horizontal cable tray fire stops was conducted in accordance with the requirements of ASTM E119, with the exception of the hose stream test nozzle.

In accordance with accepted fire protection practice, the combustible fire loading of a zone can be used to judge the adequacy of the fire area boundary barriers. Each fire loading increment of 80,000 Btu per sq. ft. indicates the need for an additional one hour of fire rating for the barriers. The relative fire hazard of a zone may be considered as low if the combustible fire loading is below 80,000 Btu per sq. ft., moderate if below 160,000 Btu per sq. ft., and high if above 160,000 Btu per sq. ft.

FACILITY NAME (1) DOCKET NUMBER (2) LER NUMBER (6) PAGE (3) St. Lucie Unit 1 05000335 The following is a summary of the combustible loading for the affected fire area/zones:

Area/Zone Description Fire Loading Rating F/70 Control Room (CR) 138.4 X 103 Btu/sq.ft. Moderate B/57 Cable Spread Room (CSR) 241.13 x 103 Btu/sq.ft. High Z/83 RAE: West Stairwell 84.96 x 103 Btu/sq.ft. Moderate A/60 "A" Switchgear Room 136.64 x 103 Btu/sq.ft. Moderate E/61 HVAC Equipment Area 89.38 x 103 Btu/sq.ft. Moderate C/56 "B" Switchgear Room 45.17 x 103 Btu/sq.ft. Low C/55W Main Hallway West (EL 19.50') 106.48 x 103 Btu/sq.ft. Moderate 0/36 Main Hallway (EL -0.50') 148.99 x 103 Btu/sq.ft. Moderate D/78 "B" Electrical Pen Room 174.88 x 103 Btu/sq.ft. High The CSR and the "B" Electrical Penetration Room have high combustible loadings. The principle combustible loading for the CSR and "B" Electrical Penetration Room is cable insulation (220.9 x 103 and138.50 x 103 Btu/sq.ft., respectively). The CSR has an installed automatic suppression system (Halon 1301). In addition, previous fire modeling performed for the CSR, without taking credit for automatic suppression, demonstrated that for the maximum expected fire scenario, the room does not reach "flashover" conditions (932 to 1,112 degrees Fahrenheit). Although there is no automatic suppression system in the "B" Electrical Penetration Room, there are few ignition sources in the room. In addition, all the zones are provided with ionization type smoke detection. These detectors will initiate an alarm in the Control Room and indicate, locally should a fire occur.

Based upon the above, FPL's initial assessment of the subject fire stops is that they are expected to be capable of providing a sufficient barrier based upon the functionality of the fire stop (i.e., not pass fire after three hours of a ASTM E119 test) when due consideration is given to the combustible loadings and/or ignition sources in the affected fire areas. Although FPL considers that the preliminary safety significance of this issue is low, the issue will be further developed as part of the corrective actions associated with the fire stops.

This LER revision addresses the additional discrepancies found during the implementation of repairs to the control room RTGB fire stops. Because of these findings, it is questionable that the fire stops could withstand the maximum expected fire scenario in the affected fire areas. However, the discrepancies are localized to small areas of the fire seals located within the RTGB panels. Operation procedure 1-0NP-100.01, Response to Fire, refers to 1-0NP-100.02, Control Room Inaccessibility, where the control room may be evacuated to safely shutdown the plant from the hot shutdown panel located in the "B" switchgear room. Therefore, while there were no cases where a fire breach existed between the CSR and the CR, site procedures are in place that ensures safe shutdown. Based upon the above, FPL determined that a potential failure of the RTGB fire penetration seals would not prevent safe shutdown of St. Lucie Unit 1. Therefore, this condition still has no significant effect on the health and safety of the public.

FACILITY NAME (1) DOCKET NUMBER (2) LER NUMBER (6) PAGE (3) St. Lucie Unit 1 05000335 111331=B

Corrective Actions

1. Compensatory fire watches were posted against the inoperable fire barriers in accordance with the St. Lucie Fire Protection program.

2. FPL will submit a fire protection exemption for the Gold Bond qualified penetrations by the SL1-20 refueling outage.

3. The Omega Point Laboratories qualified control room floor penetrations will be repaired during the SL1-20 refueling outage.

Additional Information

Failed Components Identified Various field-implemented cable tray fire penetration seals qualified under Gold Bond fire test report WP-387 and Omega Point Laboratories.

Similar Events Configurations," dated May 3, 2004, documents similar issues found on St. Lucie Unit 2 during the extent of condition review for the St. Lucie Unit 1 control room floor fire penetration deficiencies.