IR 05000440/1998005

Insp Rept 50-440/98-05 on 980209-0408.No Violations Noted. Major Areas Inspected:Portions of Licensee Fire Protection Program

ML20217J636
Person / Time
Site:	Perry
Issue date:	04/27/1998
From:	NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III)
To:
Shared Package
ML20217J597	List: IR 05000440/1998005 ML20217J592
References
50-440-98-05, 50-440-98-5, NUDOCS 9804300284
Download: ML20217J636 (13)

Text

.

.

.

U.S. NUCLEAR REGULATORY COMMISSION l

REGION lll Docket No: 50-440 i License No: NPF-58 l Report No: 50-440/98005(DRS)

!

Licensee: Centerior Service Company Facility: Perry Nuclear Power Plant

! Location: P.O. Box 97, A200 Perry, OH 44081 l

Dates: February 9 to April 8,1998 Inspectors: D. Chyu, Reactor Engineer, Region 111

,

M. Salley, Fire Protection Engineer, NRR

Approved by: R. N. Gardner, Chief Engineering Specialists Branch 2 Division of Reactor Safety l

!

l

! l l  !

t

..

9804300284 980427 PDR ADOCK 05000440 G PDR i;

.

.

.

.

EXECUTIVE SUMMARY Perry Nuclear Power Plant NRC Inspection Report 50-440/98005 The regional inspection reviewed portions of the licensee's fire protection program. Overall, engineering support to fire protection activities improved through the effective resolution of discrepant fire door seal ratings and fire hazard analysis revisions. However, a number of discrepancies were identified by the inspector with regard to the penetration seal program and the coefficient of friction used in the hydraulic calculation for the sprinkler syste !

l l

l

'

.

.

! Report Details IV. Plant Suonort l

F2 Status of Fire Protection Facilities and Equipment F2.1 Gaskets on Fire Doors Insoection Scoce l

The inspectors reviewed the fire door schedule, the installation specification, potential issue form (PIF) 95-1263, its associated corrective actions, and quality assurance audit 96-1 Observation and Findinas The licensee indicated that the rated fire doors were initially procured and qualified ;

without gaskets. The inspectors reviewed the initia' fire door schedule dated i September 1975 which did not require installation of gaskets. Some listed gaskets j (i.e., tested and approved by a listing agency) were installed later for heating, !

ventilation, and air conditioning (HVAC) purposes. Among the listed gaskets installed, I some were rated at 1%-hour and others were non-rated. The licensee subsequently :

revised the door schedule to indicate the need to install gaskets on certain fire doors for i l HVAC purposes.

l On June 21,1995, the licensee initiated PlF 95-1263 documenting the installation of gaskets whose rating was less than the fire door or lack of gaskets for some fire door in the cause analysis summary to PIF 95-1263 dated May 13,1997, the licensee discussed the material used in the rated and non-rated gaskets and the material property when subjected to heat. The licensee concluded that the gaskets did not impair the functions of the fire doors as a 3-hour rated barrier since the material used would not spread flames (a criterion discussed in an Underwriter Laboratory letter dated March 27,1985) and would not add additional stress on the door frame when heated.

The licensee concluded that installation of listed gaskets with less than 3-hour rating on

!

3-hour rated fire doors, did not degrade the rating of the fire door _ Conclusions The inspectors concurred with the licensee's assessment since the fire doors were l initially qualified without gaskets. It was acceptable either not to use gaskets on the fire doors or to use listed gaskets which had less fire rating than the fire doors.

l l

-

- ,

a

.

F Penetration Seals Inspection Scope The inspectors reviewed the licensee's Fire Protection improvement Plan dated December 1996; the licensee's operating experience review for Information Notice (lN) 88-04, " Inadequate Qualification and Documentation of Fire Barrier Penetration;" '

IN 88-56, " Potential Problems with Silicone Foam Fire Barrier Penetration Seals;"

. IN 94-28, " Potential Problems with Fire Barrier Penetration Seals;" potential issue form L (PIF) 96-3203; PlF 96-3243; Calculation No.' P54-205, " Fire Barrier Penetration Seal L Qualification and Application Basis," Revision 0; and Penetration Seal Drawings D-199-140, Sheet 19, Revision A and Sheet Observations and Findinos The inspectors reviewed PlF 96-504 conceming a horizontal crack in Penetration Seal No.1 SIB-0020 (rattle space between south wall of the room around the personnel air lock and unit 1 containment). The inspectors also inspected this seat in the field during this inspection. After the identification of the discrepancy in 1996, a fire impairment

! (96-1B-0063) was initiated until repair was effected on September 27,1996, under Work Order 96-812. Fire protection engineering performed a 10 CFR 50.59 applicability

!

'

check dated February 8,1996, for the temporary use-as-is condition. The inspectors considered the corrective actions to be adequat During a quality assurance audit in 1996, the licensee identified in PIF 96-3243 that construction gap fire seals and compartmentalized seals were not being installed as depicted on design drawings. The PIF also identified that tested configurations did not bound the installed configurations. The licensee concluded in the nonconforming condition (NCC) investigation that the identified condition was not a NCC and the drawings needed to be updated to reflect the as-built configuration. However, the inspectors did not concur with the licensee's conclusion based on the observations listed belo The NCC evaluation stated that additional testing factors for the penetration seal would not be included in the evaluation. One important perfonnance factor was omitted, which may affect the seal design, regarding the requirement for temperature limits on the non-fire (cold) side. The thermal performance (including burn-through) of the penetration assembly would have affected the rating and qualification of penetration ,

seals. However, the licensee did not confirm this key parameter when determining that i the as-built penetration seals were acceptable and operable. The licensee performed a i safety evaluation applicability check and revised drawings. Furthermore, the test report I referenced in the evaluation, Bisco Report #748-49, documented that the tested i assemblies exceeded temperature limits and failed hose stream testin I

,

!

T

.

The NCC evaluation also referenced a drawing for multiple cable tray penetration design l from V.C. Summer Plant. The licensee used the drawing to establish the acceptability of subdividing large block-out type penetration seals with M-Board. The inspectors concluded that the drawing was not representative of the penetration seat being evaluated for the following reasons:

-

The penetration seal assembly at VC Summer Plant was sandwiched between a minimum 6" wide by 1" thick section of M-board on both sides of the penetratio The M-board would provide resistive heat transfer around the edges and structural stability to the silicone foam. However, the licer s 'e did not install M-boards on both sides of the penetration seal as suggeste< by VC Summer Plant. The inspectors questioned the structural stability of the large block-out l penetration seals installed at Perry without the M-Board framin l

-

The tested seal at V.C. Summer had two or three sides installed against .

l concrete. The surrounding concrete provided a heat sink and structural stability for the seal during thermal and hose stream exposure. However, the inspectors concluded that multiple continuous penetration seals installed in individual cells in a wall opening would not be expected to function in the same manner, i

-

The licensee cited a test report of smaller scale which did not appear to bound the as-built configuration for the construction ga The inspectors also reviewed Calculation P54-205 which was to establish the basis that the penetration seal designs at Perry were qualified by fire tests in accordance with NRC criteria. The inspectors identified the following discrepancias in the calculation:

-

Section 5.6.1 stated that " Penetration openings of virtually any size may be accommodated by any listed seal types, by partitioning of the penetration opening into cells less than or equal to the maximum qualified size listed in l Table 3." The inspectors concluded that this assumption appeared unverified l and the calculation did not identify the upper and lower limits of qualified siz I Furthermore, Section 5.8.1 of the calculatien stated " Qualification of a penetration seal assembly of a given construction and size does NOT provide the basis for extrapolation of this qualification data to a penetration seat of a larger size (although smaller seals of the same construction may be adequately bounded)." These two sections may be in conflic Table 3 listed qualifying test numbers and identified the largest seat size to be 6.25 ft 2. However, Drawings D-199-140, Sheet 19, Revision A and Sheet 26 showed penetration seal assemblies of 141 f0 (18 feet 2 inch by 7 feet 9 inch)

and 51 ft:(7 feet 10 inch by 6 feet 6 inch), respectively. These sizes exceeded the maximum size listed in Table 3.

I

!

1

!

. .

.

The structural steel frame work which formed the multitude of individual cells in a wall opening would be heated and expand in a 3-hour fire exposure. This has not been adequately dispositioned L s the calculation and was not bounded by the referenced test l

. A partition wallinstead of a qualified seal size of 6.25 ft2 would be representative of the compartmentalized wall. ASTM E-119 required a minimum 100 ft2 sample for qualifying a fire rated partition wall. However, the licensee did not reference such qualifying tests for the compartmentalized penetration. It appeared that the licensee extrapolated the large opening from a maximum tested size of 6.25 ft2 This practice appeared to be unfounded according to Section 5.8.1 of the calculatio .

Section 5.6.3 discussed the percent of cable fills, cable types, and corresponding cable mass. However, it did not identify the maximum limit for cable fills. For example, the calculation did not discuss that either a cable tray with twisted pair 16 gauge 40% filled or with 300 MCM 40% filled tray would be the limit for cable tray penetrating fire seat .

Section 5.6.7 stated that in a wall or opening thinner than the required thickness of penetration seals, the licensee was allowed to install" collars" which extended beyond the plane of the wall or opening. However, the extended collar would experience greater heating due to the collar being extended above the wall's l thermal mass ar.d the additional surface area exposed to the heat. The l calculation concluded that a sealinstalled in this configuration would perform as well as one tested flush with the opening. However, this design has not been substantiated through actual testin c. Conclusions The inspectors identified one unresolved item concerning the penetration seal program as documented in PlF 96-3243 and Calculation P54-205. The specific outstanding issues include:

(1) The temperature limits on the non-fire side as demonstrated in test reports for the construction gap seals and compartmentalized block seals; (2) The bounding configuration for the compartmentalized wall considering the absence of M-boards on both sides of the wall, fire-induced deformation of structural steel frame dividing individual cells, and heat transfer characteristics of penetration material in relation to concrete-(3) Resolution of the conflicting information in Calculation P54-205 concerning extrapolation of qualified seal assemblies to larger sizes; (4) The acceptability of seal assemblies larger than the qualified seal size of 6.25 ft 2;

r

.

.

(5) The bounding cases for cable tray fills regarding cable type, size, mass, and

, percent fill; and I

(6) The acceptability of penetration material which extended beyond the plane of opening considering the increase in exposure areas.

l On April 1,1998, a conference call was held with the licensee to discuss these items.

l During the conference call, the licensee indicated that an evaluation of these items will

'

be conducted. NRC staff will review the results of the licensee's evaluation The above findings conceming engineering assumptions used to evaluate discrepant penetration seals identified in PIF 96-3243 and Calculation P54-205 are considered an unresolved item (URI 50-440/98005-01) pending inspector revie F3 Fire Protection Procedures and Documentation i

F3.1 Safe Shutdown Analvsis and Documentation for Fire Zone CC-1c Insoection Scooe The inspectors reviewed the licensee's submittal dated June 12,1985, and approved exemption for Fire Zone cc-ic, the safety evaluation report associated with Fire Zone cc-ic, PlF 95-2322 and its corrective action Observations and Findinas i

'

On November 11,1995, the licensee identified in PIF 95-2322 that Calculation P54-24,

" Fire Load Calculations," and the Updated Safety Analysis Report (USAR) had not been kept current concerning combustible loading. Design changes were made without updating design document The licensee had previously received an exemption for installing partial sprinkler system over Fire Zone CC-1c. The method approved by the NRC at tnat time was for estimating fire loading by averaging the total in situ combustible (i.e. potential heat release in BTUs) over the entire area. The combustible limit was approved at 20,000 Btu /sq ft. At that time, the fire loading was estimated to be 7,160 Btu per ft2 . Over the years, due to changes to the area and some unaccounted material from the original j exemption, the fire loading had increased. On January 3,1996, a fire protection engineer concluded that with consideration of additional fire loadings and their locations, safe shutdown capability was still available in Fire Zone CC-1c. In addition, the engineer re-estimated the fire loading and documented the results in Calculation P54-24, Revision 3, dated April 23,1996. The revised fire loading was still within the limits approved by the NRC.

.

The inspectors walked down the fire zone with the licensee. The licensee replaced originally installed Thermo-Lag with 3-M fire wrapping on certain redundant cable During the walkdown, the inspectors identified oil leaks on four air compressors and two air dryers. There were oil absorbent pads around the compressors to stop the oil from l 7

.

.

'

I traveling further. There was a tail pipe from the sewage ejector blowing oil unto the floor which appeared to have a thin film of oil. The licensee immediately took corrective actions to wipe down the noted oil leak and initiated PlF 98-0250 to document this deficiency. This is considered an inspection Follow up item (IFl 50-440/98005-02)

pending inspector review of the extend to the scope and corrective actions taken by the licensee for this and other areas. The components noted above, were located in an area within Fire Zone CC-1c where no sprinkler systems were installed. This immediate location did not contain safe shutdown related equipment. Other portions of Fire Zone CC-1c had sprinkler systems, including an area where an oil leak was found on the emergency closed cooling pump base. The inspectors considered the identified oil leaks to be of minor significanc The inspectors reviewed Change Request (CR)96-160, dated November 20,1996, for modifying the fire hazard analysis for Fire Zone CC-ic to reflect the change in the fire loading. A source document for this CR was Calculation P54-24, Revision 3, which contained an additional 300 lb of chiller insulation. However, the additional chiller insulation was miscategorized as motor insulation. Section 9A.2 of USAR designated the motor insulation to be insulation in motors greater than 50 horsepower. The 300 lb chiller insulation was on surrounding piping, not inside the motor. The chiller insulation was bounded by the miscellaneous category in Calculation P54-24, Revision 3. A design engineer initiated Design Change Control (DCC) 34 to revise the category of the chiller insulation from motor insulation to miscellaneous insulation and to correct combustible classification and quantities within the miscellaneous category. The miscellaneous category was revised to include file cabinet (5,016,000 Btu), supplies (2,232,000 Btu), and chiller insulation (3,000,000 Btu). The total Btu for the miscellaneous category remained unchanged (10,248,000 Btu). All other fire loadings previously included in the miscellaneous category were bounded by the in-situ combustible limit of 50,000,000 Btu. The inspectors determined that in preparation of CR 96-160, the licensee used the correct combustible loading calculation as a design input. The change in category for the chiller insulation was acceptable and appropriat Conclusion The inspectors concluded that the estimation method used by the licensee was within !

the licensee's original licensing basis and within the bounds of the exemption. The additional combustible loading and its location in Fire Zone cc-1c were considered bounded by the assumptions in the exemption. An inspection follow up item was i identified for housekeeping of the are I

F3.2 Review of Hvdraulic Calculations for Water Based Sunoression Svstems j Jaspection Scooe The inspectors reviewed Calculation P54-124, Fire Protection Suppression System Water Supply Calculation," Revision 1, P54-123, " Piping, Configuration input for the Hydraulic Model of the Water Supply Distribution System For Fire Suppression Systems," Revision :

i

(

.

b. Observations and Findings PIFs 95-74 and 96-72 identified inadequacies in the original calculations for the fire protection water supply. In Section 9A.5 of the USAR, the largest flow demand for a safety related area was identified as 1856 gpm. This value was supported by Hydraulic Calculation Sheet No. 4549-71 A-078, Revision 1, dated July 27,1980. However, the sheet did not show any calculation assumptions or design inputs. In April 1997, the licensee had modeled the underground piping configuration for the fire protection water distribution system using a hydraulic computer model and documented the results in l

'

Calculation P54-123. The inspectors determined that the calculation used original vendor drawings as design inputs and incorporated modifications made to the system since initial constructio On January 13,1998, the licensee completed P54-124, Revision 1, documenting the capacity of the fire pump and water distribution system to meet the demands of the suppression systems. This calculation included all of the sprinkler systems installed in i the plant. The licensee identified tha'. the highest demand for the safety related area was the Unit 1 Division 1 cable spreading room. The system demand at the riser base was 1434 gpm at 75.3 psig. With consideration of 500 gpm hose demand, the values

, required at the fire pump was 1934 gpm at 102.3 psig. The value required at the pump j with the shortest portion of flow path isolated changed to 1934 gpm at 112.3 psig.

l Based upon the fire pump performance, the fire pump was capable of meeting the highest calculated system demand.

l

,

Section 2-3.1.1 of the National Fire Protection Association (NFPA) 13-1973, discussed the acceptability of connecting fire system to a reliable water works system. Section A-2-3.1.1 of the code designated the water works system to be public water supply. For such a water supply, the licensee used the coefficient of friction, "C factor," of 120 and 100 for wet and dry systems, respectively, in accordance with the NFPA 13. However, the water supply source for the licensee's fire suppression system was not public water (potable) but Lake Erie (not potable). In Section 5-36 of Fire Protection Handbook, the C factor was adjusted due to age of and the medium in the pipe. The C factor for moderately corrosive water was typically lower than the one for mildly corrosive wate Lower C factor would have resulted in higher friction loss and reduced system flow The calculations performed by the licensee were within the guidance of NFPA 1 However, the inspectors considered the use of C factor without justification for system aging and the type of medium in the suppression system to be a weaknes The licensee had a surveillance procedure which verified the condition of the underground piping. This surveillance ensured that the fire pump through different ;

piping configurations could still meet the minimum flow and pressure requirement However, the tested configurations were different from time to time (e.g., the section of the piping tested may not be the same every time). Such surveillance was not an adequate tool to monitor C factor degradation. The licensee recently identified a microbiological induced corrosion (MIC) which could further degrade the condition of the piping in the suppression systems. Without consistent trending of the piping

.

performance, the C factors used in the calculation could not be readily validated. This is

i

.

.

.

! an inspection Follow up ltem pending review of the licensee's committed development

,

of a surveillance procedure that parallels the hydraulic design basis calculation (IFl 50-440/98005-03), Conclusion l

The inspectors concluded that the hydraulic calculation for the water based suppression system was performed in accordance with the requirements of NFPA 13. However, one weakness was identified for the lack of adjustment of the C factors for the age of and the medium in the piping. One inspection follow up item was identified concerning the development of a surveillance procedure which would validate the C factors used in the calculation.

l F3.3 Individual Plant Examination - Extemal Event (IPEEE) for Fire Zone cc-ic i

j- The inspectors reviewed the IPEEE submittals and portions of the supporting calculation for Fire Zone cc-1c. There were several areas identified in the licensee's IPEEE submittal which did not meet the safety goal of 1e-6 per year during the screening l

phases of the Fire Induced Vulnerability Evaluation (FIVE) program. These areas were summarized in Table 4-10 of the IPEEE report. Among the areas was Area CC1 which included Fire Zones CC-1a, CC-1b, and CC-1c. The initial fire induced core damage frequency for Area CC1 was 1.17e-4 per year. The licensee then performed a Level 2 evaluation for the areas. After performing the additional analysis, the fire induced core l damage frequency for Area CC1 was 2.03e-6. The licensee concluded that the area l was acceptable with no enhancements or modifications. The details of the Level 2 i evaluation for Area CC1 and the acceptability of the conclusions will be reviewed by NRR in accordance with Generic Letter 88-20, Supplement F6 Fire Protection Organization and Administration I The inspectors reviewed the scope of the performance-based fire protection surveillance reduction program and the recommendation for reducing smoke detector surveillance l frequency. The licensee has not implemented this program; therefore, there was no l actual reduction in any fire protection surveillance frequency. The inspectors noted l several weaknesses in this draft program. The criteria for surveillance frequency reduction was not clearly identified. The reduction recommendation was based upon the number of work requests generated from the total number of inspections performe The licensee did not develop an acceptable level of risk for potential failure rates. The reduction recommendation for smoke detectors was beyond the recommendation by the manufacturers and NFPA codes. In addition, the program did not discuss the consequence and increased risk of a missed surveillance if the reduction program was implemente F8 Miscellaneous Fire Protection issues

,

(Closed) URI 50-440/96016-02: The licensee identified in quality assurance audit PA 96-17 differences between the as-built fire seal configuration and tested configuration

f l .

( .

and documented the discrepancy in PIF 96-3243. The inspectors reviewed the Non-conforming Condition (NCC) investigation for PlF 96-3243. The licensee l determined that this was not a nonconforming condition. The inspector disagreed with this conclusion of the NCC investigation for the reasons as documented in Section F2.2 i

'

of this report. This item is closed and will be tracked by URI 50-440/98005-0 !

l l V. Management Meetings X1 Exit Meeting Summary The inspectors presented preliminary inspection results to members of licensee management at l the conciusion of the inspection on February 12,1998. On April 1, the licensee and the

!

inspectors had further discussions concerning the penetration seat evaluations noted in i PIF 96-3243 and Calculation P54-205. On April 8, the inspectors conducted a final exit meeting with the licensee via telephone. The licensee acknowledged the findings presente The inspectors asked the licensee whether any materials examined during the inspection should be considered proprietary. No proprietary information was identifie l

'

. ,

,

PARTIAL LIST OF PERSONS CONTACTED Licensee i

l H. W. Bergendahl, Director, Nuclear Service R. G. Collings, Manager, QAS C. Elberfeld, Compliance Engineer D. Haviland, Lead Civil / Structural

)

M. D. Makar, Design Engineer )

J. Perry, System Engineer D. G. Philipps, Plant Support Supervisor J. J. Powers, Manager, DES J T. S. Rausch, Director, First Energy J. H. Stegman, Fire Protection Coordinator i l

NBC

D. Butler, Reactor Engineer D. Kostoff, Senior Resident inspector ,

S. West, Chief, Fire Protection Section, NRR 4 INSPECTION PROCEDURE USED I IP 64704: Fire Protection ITEMS OPENED, CLOSED, AND DISCUSSED Ooened 50-440/98005-01 URI Engineering assumptions used to evaluate penetration seals 50-440/98005-02 IFl Weakness in housekeeping in Fire Zone cc-ic 50-440/98005-03 IFl Surveillance Procedure for validating C factors Closed 50-440/96016-02 URI Differences between the as-built fire seal configuration and tested configuration i

-

.

.- .

LIST OF ACRONYMS USED CFR Code of Federal Regulations l DCC Design Change Comment DRS Division of Reactor Safety FIVE Fire Induced Vulnerability Evaluation HVAC Heating, Ventilation, and Air Conditioning ,

IFl Inspection Follow up Item i IPEEE Individual Plant Examination - External Events MIC Microbiological Induced Corrosion NFPA National Fire Protection Association NRR Office of Nuclear Regulation NCC Non Conforming Condition PlF Potential Issue Form J URI Unresolved item USAR Updated Safety Analysis Report

!

l l

l i

!

13