ML20155K605

Forwards Info Requested in NRC Re Comanche Peak Response Results Repts for Isaps I.a.4,I.b.3,II.b, Iii.D & VII.b.2.Info Covers Concrete Compression Stength & Valve Disassembly

ML20155K605
Person / Time
Site:	Comanche Peak
Issue date:	05/02/1986
From:	Counsil W TEXAS UTILITIES ELECTRIC CO. (TU ELECTRIC)
To:	Noonan V Office of Nuclear Reactor Regulation
References
NUDOCS 8605290061
Download: ML20155K605 (139)

Text

{{#Wiki_filter:" w i \\ -y TEXAS UTILITIES GENERATING COMPAhT 4KYWAY TOWER e 400 N(DETH OLIVE KTREET. L.R. E t. DALLAS, TEKAs 75208 t C v v4E P ES DEsef May 2, 1986 Vincent S. Noonan Director PWR Project Directorate #5 Division of PWR Licensing - A U.S. Nuclear Regulatory Con: mission Washington, D.C. 20599

Reference:

Letter to W. G. Counsil (TUGCO), f rom V. S. Noonan (NRC)

Subject:

NRC Staff Request for Additional Information on Comanche Peak Response Results Reports for ISAPs (I.a.4, I.b.3, II.b, III.d and VII.b.2) dated April 28, 1986.

Dear Mr. Noonan:

Enclosed herewith is the information requested by the referenced letter. Should you have any questions or need further clarification, please contact Mr. John W. Beck at (214) 979-8646. w Very truly yours, A A John W. Beck for W. G. Council JWB/feo Enclosure cc: Service List W 8605290061 860502 DR ADOCK 05000445 o PDR l \\ A DIVISION OF TEXAS UTTLITIES ELECTRIC COMPANY

Pcgs 1 of 23 ENCLOSURE

REFERENCE:

DOCKET NOS. 50-445 and 50-446 REQUEST FOR ADDITIONAL INFORMATION FOR THE FIVE ISAP RESULTS REPORTS (I.a.4, I.b.3, II.b, III.d, and VII.b.2) AND FUTURE RESULTS REPORTS. QUESTION: 1. Address those questions raised in ASLB Memorandum, Proposed Memorandum and Order dated April 14, 1986, and provide appropriate documentation.

RESPONSE

The SRT expects to publish responses to the Board's questions, as propounded in its " Proposed Memorandum" and modified during the pre-hearing conference of April 22, 1986, in the form and time frame described at that conference. See Tr. 24353 (4/22/86). QUESTION: 2. Address whether the issues raised in the results reports had implications of deficiencies in the QA/QC program, design and/or construction and reference documents that will be provided to the staff that will address these implications.

RESPONSE

These issues fall into two categories: 1ssues relating to design, construction or testing identified during the conduct of action plans, and the evaluation of action plan results for impact on collective evaluations of the design, hardware, testing program or QA/QC program. For the first category, Review Team Leaders have and continue to formally notify each other of findings in the conduct of their respective action plans that could impact or require investigation in the context of another Review Team Leader's ISAP or DSAP. In addition, deficiencies identified during the conduct of some action plans may be evaluated for impact within that specific action plan Results Report. For the second category, the intent of the Collective Evaluation Reports described in Section VI of the CPRT Program Plan, though not explicitly stated, is to address the implications of any design, harddare, testing or QA/QC deficiencies discovered during the conduct of any Issue Specific Action Plan (ISAP) or Discipline Specific Action Plan (DSAP) in the appropriate Collection Evaluation Report. These Collective Evaluation Reports will be issued during the latter stages of the CPRT Program. NRC

~ REQUEST FOR ADDITIONAL INFORMATION (Cont'd) QUESTION: 3. Where an ISAP resulted in corrective action, address the status of the corrective action and identify the method you plan for communicating to the staff the corrective action is completed.

RESPONSE

Specific corrective action initiated as a result of discrepancies identified during the course of implementing ISAPs are translated to Project NCRs, TDDRs and TDCRs in accordance with the Project's Program. With respect to Results Reports I.a.4, I.b.3, II.b, III.d and VII.b.2 no corrective action beyond the scope of specific deficiencies has been recommended to the project. To the extent that the Program Plan might require third-party oversight of corrective action in any case, reporting of this overview will be done as set forth in Appendix H, Section B, Paragraph 3. QUESTION: 4. Describe how findings from one ISAP, which relate to a particular ISAP that is being addressed are considered.

RESPONSE

We do not understand the question as posed. 2 NRC

I.a.4 Agreement Between Drawings and Field Terminations QUESTION: 1. For the instances identified by the NRC TRT and Region IV, and CPRT where the drawings have not yet been revised, to reflect the existing field termination conditions, provide the actions you are taking to upgrade your as-built field termination drawings.

RESPONSE

For discrepancies identified by the NRC-TRT and CPRT, the drawings have been revised or the field terminations corrected such that the field terminations are appropriately reflected on the drawings. Discrepancies identified to the Project by NRC Region IV have been documented on NCRs and TDDRs. When these are dispositioned the field terminations and drawings will agree. To the extent the question encompass,es nonterminated spare conductors, the project drawings will not be -revised to reflect the field; because such conformity is neither a design nor project requirement. QUESTION: 2. What is the basis for considering terminated and non-terminated spare conductors as valid population sample items for essential Class IE Systems.

RESPONSE

The basis for including spare conductors in the population was as follows: Spare conductors could potentially be involved with functional deficiencies (e.g., a spare conductor reversed with a functional conductor, a spare conductor connected to an active circuit, etc.), thus information concerning spares should not be bypassed. Conductors that were once functional were often converted to spares by design change, and it was considered to be important to check these conductors for adequacy of the design change . implementation process. The NRC/TRT checked and addressed spares. One of their findings involved spare conductors that had once been functional and (af ter being spared by design change) were not lifted from their respective terminal points. J 3 NRC

r 1.b.3 Conduit to Cable Tray Separation QUESTION: Provide the following information: (1) Gibbs and Hill analysis report on conduit separation: (3) DCA-15917 mentioned on page 2 of the Results Repcrt which reduced the conduit separation to one inch (this may be included in the G6H analysis report), and (4) Gibbs and Hill memo EE-863, 1/17/84, which contained simplified analysis reviewed by NRC-TRT on site (this may be included in the G6H analysis report);

RESPONSE

The information requested in items 1,, 3 and 4 is attached. These documents are all contained in the Results Report Working File or Project Document Control Center. ITEM DOCUMENT ISAP I.b.3 FILE NO. (1) GTN-71266 I.b.3 - 8A.022 GTN-71284 1.b.3 - 8A.023 CPRT-294 1.b.3 - 8A.028 (3) DCA-15917 (from Document Control Center) (4) TWX #14,958 1.b.3 - 8A.001 GTN-69531 I.b.3 - 8A.002 Sandia Report I.b.3 - 8B.001 QUESTION: (2) Documentation to indicate that TUGC0 has approved the Gibbs and Hill analysis report:

RESPONSE

A FSAR change request which utilizes the Gibbs & Hill analysis as supporting documentation is being prepared. When submitted, it will document TUGCO's acceptance of the analysis. 4 NRC

[. b. 3 -6 A.co I SENT BY TELECOPY '-20-84 INDEXED I'30A.M. DATL m SEPTEMBER 20, 1984 TWX #14,958 ATTN: R. E. BALLARD / T. R. VARDAR0 / S. P. MARTIN 0VICH SUB: NRC REQUEST FOR ADDITIONAL INFORMATION THE NRC TECHNICAL REVIEW TEAM (TRT) HAS REQUESTED ADDITIONAL INFORMATION IN THE AREA 0F ELECTRICAL SEPARATION. THEIR SPECIFIC REQUEST IS AS FOLLOWS: "THE TRT FOUND THAT THE EXISTING TUEC ANALYSIS SUBSTANTIATING THE ADEQUACY OF THE CRITERIA FOR SEPARATION BETWEEN CONDUITS AND CABLE TRAYS HAD NOT BEEN REVIEWED BY THE NRC STAFF. ACCORDINGLY, TUEC SHALL SUBMIT THE ANALYSIS THAT SUBSTANTIATES THE ACCEPTABILITY OF THE CRITERIA STATED IN THE ELECTRICAL SPECIFICATIONS GOVERNING THE SEPARATION BETWEEN INDEPENDENT CONDUITS AND CABLE TRAYS." WE HAVE DISCUSSED THIS REQUEST WITH T. R. VARDAR0 AND S. P. MARTIN 0VICH. PLEASE PROCEED IMMEDIATELY TO FORMULATE THE REQUIRED RESPONSE AND TELECOPY IT TO US. A TIMELY RESPONSE IS OF UTMOST IMPORTANCE TO US. AS SUCH, OVERTIME IS AUTHORIZED AND EXPECTED IN ORDER TO GET THE RESPONSE AS 500N AS POSSIBLE. IF YOU HAVE ANY QUESTIONS OR REQUIRE ADDITIONAL INFORMATION, PLEASE ADVISE. W. I. V0GELSANG - ELECTRICAL ENGINEERING L. M. POPPLEWELL PROJECT ENGINEERING MANAGER CPSES J0BSITE 910/890-8660 TUGC0 GRSE LMP:WIV:ery 35-1195 CC: ARMS - 0 C C RECElVED FILE SEP 2 01984 DOCUMENT CONTROL

& 6, Z b.3-8A.co z f y Egtumume/187*,jPMartinovich, PNLalaj ' g6%ng, MMilam..Sleck ../ \\ i I i A Orevo Company l September 27, 1984 GTN. 69531 Texas Utilities Generating Company Post Office Box 1002 Glen Rose, Texas 76043 Attention: Mr. J. B. George Vice President / Project Gen. Manager Gentlemen: TEXAS UTILITIES GENERATING COMPANY COMANCHE PEAK STEAM ELECTRIC STATION GEH PROJECT NO.2323 NRC REQUEST FOR ADDITIONAL INFO ELECTRICAL SEPARATION CRITERIA REF: TWX-14958 (9-20-84) Attached please find the analysis requested in the referenced TWX substantiating the adequacy of the criteria for separation An advance copy of this between conduits and cable trays. analysis was telecopied to W.I. Vogelsang on Monday 9-24-84. (We have also transmitted under separate cover, one copy of Sandia Laboratories Report on Cable Tray Fire Tests Please advise if we can provide any (SAND 77-1125C). i additional assistance. Very truly yours, GIBBS & HILL, INC. Robert E. Ballard, Jr. Director of Projects T REBa-NLdPM:sca 1 Letter + Attachment cc: ARMS (B&R Site) OL W. I. Vogelsang (TUSI Site) 1L + Attachment ~ Orave - ~

m.. i am r.. ~ ~ ' Gibbs& Hill,Inc. } E l September 24, 1984 To: w* I Vogelsang Per your request to Sam Martinovich enclosed please find one copy of Sandia Laboratories Report on Cable Tray Fire Tests (SAND 77-1125C) and one copy of report entitled Separation Criteria as prepared by SPMartinovich and telecopied to you on September 24. ~ O S. M.Marano ,-/ - -~ ~....... l

v s s SEPARATION CRITERIA j The raceway separation criteria utilized in the Gibbs & Hill electrical drawings and specifications is based upon the require-ments of IEEE-384, 1974 and Regulatory Guide 1.75 (Rev. 1, 1/75). j Although very specific criteria is provided in the Standard and Regulatory Guide for separation between cable trays, no specific criteria is provided for separation between conduits and cable trays. In developing the separation details currently in Specification ES-100 and on Drawing El-1702-02, it was recognized that conduit provides a raceway medium which ef fectively isolates internal events (e.g., faults) from the external surroundings. In this regard, a conduit system provides enclosure integrity far superior to that of enclosed tray with covers and/or solid bottoms and splice plates between sections. Therefore, the same criteria required by the Standard and Regulatory Guide specif-ically for trays, need not be arbitrarily applied to conduits. In comparing rigid conduit to enclosed. tray, it was noted that conduit has: 1. Substantially heavier gauge body than tray - providing a more effective heat sink than equivalent cross-sectional area of tray. 2. Threaded connections providing essentially air-tight medium which inhibits internal combustion and effectively isolates internal events from the existing surroundings. 3. Size typically limited to 5-inch OD thus limiting both volume of cables (combustibles) contained and exposed surface area. 4. Curved surface providing radial distribution of heat and much less favorable heat transfer characteristics to or from an adjacent tray than a flat surface of equivalent area. Thus, in many instances, conduits satisfy the Standard's require-ments for a barrier *.

• IEEE 384 defines 'a barrier as -

" A device or structure interposed between Class lE equipment or circuits and a potential source of damage to limit damage to Class lE systems to an acceptable level." m m 31?tD BY TELECOPIER ?.2 V-ff i 2 _...,...._f.............f. =.

8

• Details 45 through 49, 52 through 55 and 57 on El-1702-02 identify the separation requirements between cable tray and conduit.

In general, these details require a minimum of 3-foot horizontal and 3-foot vertical separation in all general plant areas, and 1-foot horizontal and 2-foot vertical separation in the cable spreading room. This separation.is reduced to 1-inch only in those instances where the conduit is considered to be an effective barrier as discussed below. For the details shown in ES-100 and on Drawing El-1702-02, a conduit has been considered to be an effective barrier whenever it is at least 1-inch away from circuits or raceway of a dissimilar train and: a. It contains no Class lE or associated circuits or, b. It does not traverse directly above or in front /behind a horizontal or vertical tray, respec,tively, of dissimilar train. When a conduit contains no Class 1E or associated circuits, for example, it clearly satisfies the requirements of a barrier. It should be noted that the barrier need not limit damage to non-safety circuits to any level. Logically then, a conduit con-taining non-Class lE circuits can be placed up to 1-inch from the top, bottom or sides of a Class 1E open ladder tray since the conduit provides a protective barrier separated by at least 1-inch from the Class lE circuits (see Detail 49, El-1702-02). It is recognized that the converse is not true and conduits containing safety-related circuits may require more than 1-inch separation from open trays of dissimilar train depending upon orientation of conduit and tray. This has been considered in the separation criteria where in general, the minimum required separation in any direction exceeds 12 inches. The results of cable tray fire tests performed by Sandial , Laboratories for NRC (subsequent to issuance of IEEE-384, 1974), to confirm the suitability of then current design standards and regulatory guides, are supportive of the judgments used in developing Conduit Separation Criteria for CPSES back in 1975 regarding self-induced fire effects on IEEE-383 qualified cables. e 9 4 0 ._.,-------,Y-----,.--- ~* a

n j h i e Summarizing some of the more significant findings in the Sandia Report: 1. In electrically initiated fires, the intense period of the fire persisted at a particular location for between 40 and 240 seconds before die out began to occur. This is less than the time required to consistently ignite a tray of IEEE-383 qualified cables in the propane-fueled exposure fires (typically.300 seconds). 2. In the electrically initiated fire, cables in the tray 10.5 inches above the donor (fire) tray were exposed to a convective heat flux of about 6000 BTU /hr/ft2, which corresponds to a local gas temperature of approx. 1000 degrees F. The circuits remained functional and samples of the insulation from the bottom of the tray over the fire zone which were given elongation measu,rements, showed less than a 10 percent increase. 3. The luminous zone of the electrically initiated fire was optically thin which enabled immersed objects to radiate heat to the cooler surroundings. Thus equilibrium surface temper-atures of engulfed cylindrical objects varied from about 1200 degrees F just above the tray to 650 degrees F at a height of 10 inches. (Note that minimum vertical separation of 24 inches utilized on CPSES is more'than twice this distance and maximum temperatures are anticipated to be well below temper-atures successfully withstood during the fire tests.) 4. In the electrically initiated fire, heat transfer to immersed objects is convection dominated with radiation accounting for no more than 30 percent of the total heat flux, even in the luminous region. (Logically then, conduits beside or below horizontal trys are shielded from the major, convective heat flux.) i Probably the strongest evidence in support of CPSES conduit separation is the results of the exposure fire test conducted by Sandia in which conduits and trays were included. In these tests, 14 trays were stacked 10.5 inches apart. Directly above each tray within 10.5 inches, a conduit containing additional cables was located. No separation was provided between any conduit and the bottom of the tray above. Although all circuits in the conduits above the third tray failed during the exposure fire (the conductors short-circuiting to the conduit and each other), circuits in the lower two (2) conduits maintained circuit l

e o% i ~ _4_ o integrity throughout the duration of the exposure fire. Considering that the fire in the lower two (2) trays was more severe than an electrically initiated fire, being externally fueled and of longer duration, the results provide a conservative wors,t case. Recognizing that the Sandia1 tests are not plant specific, the following analysis is presented to demonstrate with margin, the adequacy of CPSES conduit / tray separation. A hypothetical worst case is chosen whereby an open horizontal tray is separated by only an air gap from a vertical conduit (note that El-1702-02 requies a minimum of 12 inches in Detail 47). See Figure below: '1F h 8 ( A 1 .l O air gap Since the conduit is vertically oriented, convective heat transfer is essentially negligible. Reference 1 establishes the time-mean height of the luminous zone as 5 to 7 inches above the tray and the radiated heat flux (for a cylindrical object immersed in the fire) as 7000 BTU /hr/ft2, Since exposed cables of one train cannot run within 3-feet vertically of another train per IEEE-384, it can be very con-servatively assumed that the minimum length of conduit will never be less than this distance. Assuming this entire radiated heat flux were transferred to 50 percent of the conduit circumference (facing the tray) over a length of 7-inches corresponding to the height of the luminous zone, the heat input rate is given as: q in = 7000 x .5( 7 d') 7" Btu /hr. 144"/ft2 Where'd = conduit diameter (inches) ~

~ 4 's Since the only heat dissipation considered herein will be via convection to surrounding air, the worst case value of 'd' is for the minimum conduit size. Per NEC, a 1-inch trade size conduit has an inside diameter of 1.05 inches. This will be assumed also for the outside diameter. Then q in = 7000 x .08018 = 561 Btu /hr. The heat dissipated to surroundings is given by: q out = hAA T (ref. 2) Where AT = difference between conduit surface temperature and surrounding air A = free surface area off conduit for convection h=C (A T)0.25 for natural convection of a solid surface in still air C = 0.4/d 0.25 for vertical pipes more than 2 ft in length with diameter = d (inches) Assuming: q in = q out q in = hA 4T or T = q in/hA and A = 17 d [36" .5 (7") ] =.744 ft2 144 h = 0.4j!d.25 0 0.25 (2LT)

0.395 ( 4 T) 0. 25 then tit

561 1.25 or T = 1908.2 (. 395 ) (. 74 4 ) AT **# l and 21T = 421 degrees F i \\ l l l l

i .' Even in a 122 degree F ambient, the maximum conduit surface temperature would not exceed 543 degrees F (122 + 421). This is well below the temperatures to which exposed cables were subjected (1000 degrees F local gas) in reference 1 with satisfactory results. The analysis herein is also extremely conservative in that conduit supports (and heat conducted to them) and radiant heat dissipation are neglected, a continuous 7-inch flame is assumed adjacent to the conduit, a conduit length of only 3-feet is assumed, and only an air-gap separation is assumed between conduit and tray. d Sandia Report No. SAND 77-1125C

2) General Electric Handbook 2nd Edition, C.

E. O'Rourke m O ,w 4

ha E b. 3 - 8 4. c22 Gibbo S Hsilo inc. 11 Penn Pf aza New Wrk, New York 10001 212 760-Teles; Domestc.127636/968694 Intematcoat 428813/234475 A Oravo Corepany February 28, 1986 GTN-71266 e as Utilities Generating Company . st Office Box 1002 Glen Rose, Texas 76043 Attention: Mr. J. B. George Vice President / Project Gen. Mgr. Gentlemen: TEXAS UTILITIES GENERATING COMPANY COMANCHE PEAK STEAM ELECTRIC STATION G&H PROJECT NO. 2323 CONDUIT TO CABLE TRAY SEPARATION REF 1: TRT ITEM 1.b.3 REF 2: GTN-70600 DTD 9/19/85 Enclosed please find Gibbs & Hill's Tray / Conduit Separation Criteria for incorporation in the TRT Item 1.b.3 results report. Mechanical calculation No. 800, Rev. I will be transmitted under separate cover on Monday, March 3, 1986 upon completion of design review. The criteria and analysis are in agreement with and support the FSAR change request previously submitted via reference 2. Therefore, no additional changes to the FSAR regarding this subject are anticipated. Please advise if you have any, questions or require further assistance. Very tru

yours, GIB S,& HIL Inc.

Q)) < +-. obert [ Ballard, Jr. REBa-J r lc 1 Letter Director of Projects CC: ldUiS (B&R Site) OL 7W. I. Vogelsang (TUSI Site) lL lA

TRAY / CONDUIT SEPARATION CRITERIA 101C9dWE1190 The raceway separation criteria utili:ed in the Gibbs Hill electrical drawings and specifications for the Comanche Peak Steam Electric Station (CPSES) are based upon the requirements of IEEE-384, 1974 and Regulatory Guide 1.75 (Rev. 1, 1/75). Although very specific criteria are provided in the Standard and Regulatory Guide for separation between cable trays, the same degree of specificity is not provided for separation between conduits and cable trays. This discussion will therefore present the methodology used in applying IEEE-384, 1974 and Regulatory Guide 1.75 (Rev. 1, 1/75) to conduits requiring separation from cable trays of redundant

• safety trains. Separation details are shown on Drawing El-1702-02

which, as stated

therein, apply when hazards are limited to failures or faults internal to electrical equipment or raceways.

Where other potential hazards from sources such as missiles, high energy line breaks, pipe whip or external fires exist, greater separation may be required. Such conditions however, are beyond the scope of the drawing and this discussion. It is apparent from the discussion in the foreward to IEEE-384, 1974 (and in the subsequent revision in 1977) that the minimum separation distances in the standard were based upon the potential effects of an electrical fire. Regarding the additional work needed to arrive at a standard wire and cable test to determine if lesser separation distances could be called out, the standard states "such a test should be designed to provide data on potential propagation to circuits above,

below, and adjacent to a cabl e fire. " In the 1977 revision, the forward states that "the distances that are given for separation between trays required to be separated in areas of limited hazard potential are based on current available data from actual cable fire situations and are considered to provide an adequate degree of separation."

In both revisions of the standard, the separation distances indicated between trays are the same. Consistent with the standard's intent, the most severe ha:ard considered herein will be an electrical fault of sufficient magnitude and duration to cause a fire in the raceway. The results of actual electrically initiated cable tray fire tests on IEEE-383 qudlified cables performed by Sandia (Ref.

1) will be used to provide

.the characteri:ation of such a fire and to evaluate a thermal analysis of a worst case configuration. The term " redundant" as used

herein, applies to different safety-related trains or safety and non safety-related trains.

1 1

r \\ Disswamino In developing the separation details currently on Drawing El-1702-02 it was recognized that conduit provides a raceway medium which effectively isolates internal events (e.g. faults) from the external surroundings. In this regard, a conduit system provides enclosure integrity which is superior to that of enclosed tray with covers and/or solid bottoms and splice plates between sections. Therefore, the same criteria required by the Standard and Regulatory Guide specifically for trays need not be arbitrarily applied to conduits. In general, the separation distances required by IEEE-384 between redundant cable trays is three feet between trays separated horizontally and five feet between trays separated vertically. This separation applies to open ventilated cable trays in general plant areas in which potential ha:ards such as missiles, external fires, and pipe whip are excluded. Lesser separation is permitted in limited hazard areas such as the cable spreading room where the minimum required horizontal and vertical separation between redundant trays are reduced to one foot and three feet respectively. The standard requires that where these distances are used to provide adequate physical separation: (1) Cables and raceways involved shall be flame retardant (2) The design basis shall be that the cable trays will not be filled above the side rails (3) Hazards shall be limited to f ailures or f aults internal to the electrical equipment (raceways) or cables Where termination arrangements preclude maintaining the above separation distances, the standard requires that the redundant circuits shall be run in enclosed raceways that qualify as barriers. A minimum distance of one inch is required between these redundant enclosed raceways. Regulatory Guide 1.75, Rev.1 is in agreement with these provisions of the standard and for the balance of this discussion, reference to the " standard" will mean IEEE-084, 1974 and Regulatory Guide 1.75, Rev. 1 as applicable. 1

( Ffgurce 2 cnd 3 in IEEE-384 depict arrangements of redundant cable trays enclosed with solid bottoms and/or covers which will satisfy the separation criteria therein. Applicable details in these figures are shown below. 4-loivision) scuo Tsay 1 seta [soucfeatcovsa Ivotvisionf heetvissouf l t secu i $ s* oivision f FIGURE 2 FIGURE 3 In the above figures, the standard provides examples of " enclosed raceway". It should be noted however, that in Figure 2 the trays are not totally enclosed as in Figure 3.

Thus, as would be expected, orientation of the raceway is obviously a consideration as is the degree of enclosure which'is commensurate with the hazard potential.

No examples of acceptable separation between a conduit and a redundant cable tray are illustrated.

However, a

inch separation is implicit per Figure 3 when the trays one are enclosed and conduits are considered to be " enclosed raceways". Separation requirements between~ conduits and geen trays must be determined by similar reasoning and analysis where required. The CPSES separation criteria are consistent with the requirements of the standard for tray separation and in addition, define conduit separation requirements which are intended to provide an equivalent level of protection for redundant circuits. The results of cable tray fire tests (Ref 1) performed by Sandia Laboratories for NRC (subsequent to issuance of IEEE-284, 1974), to confirm the suitability of then current design Standards and Regulatory

Guides, are supportive of the rationale used in developing raceway separation criteria for CPSES in 1975 regarding self-induced fire effects on IEEE-383 qualified cables.

Details 45 thru 49, 52 thru 55 and 57 on drawing El-1702-02 identify the balance-of plant (BOP) separation requirements between cable tray and conduit. (Detail 60 is a special case f or the Nuclear Instrumentation System (NIS) conduits which addresses specific requirements of the Nuclear Steam Supply System (NSSS) vendor. These NIS conduit separation requirements will not be discussed here, however in all cases the NIS requirements either meet or exceed the BOP conduit separation criteria.) 3

e \\ These details can be grouped into four basic categcries:

1) Safety-related conduits located above horizontal trays of redundant safety train (Details 46 and 48)

2) Safety-related conduits located adjacent to or below hori=ontal trays of redundant safety train (Details 45, 47 and 57)

3) Safety-related horizontal or vertical conduits located parallel to or crossing vertical trays of redundant safety train (Details 52 thru 55) 4)Non safety-related conduits located above, beside or below safety-related hori=ontal or vertical trays (Detail 49)

In general, these details require a minimum of 3-foot hori:ontal and 3-foot vertical separation in all general plant areas and 1-foot horizontal and 2-foot vertical separation in the cable spreading room. This separation is reduced to 1-inch only in those instances where the conduit is considered to be an effective barrier as discussed below., The orientation of conduit and tray in the electrically-initiated fire tests (Ref. 1) conducted by Sandia included all configurations in categories 1 and 2 above except for the conduit running parallel with and 1-inch from the side rail of the tray as in Detail 45 of Drawing E1-1702-02. Conduits used in the Sandia tests consisted of 3-inch schedule 40 pipe, whereas the minimum conduit size used at CPEES is 1/2-inch nominal ID. An analysis (Ref. 3) was performed to address these differences between the as-built and test configurations and justify adequacy of the CPSES conduit separation criteria. The Sandia tests also demonstrated acceptable separa ti on with only 10.5-inch vertical spacing between trays, far less than the minimum 24-inch required between a tray and redundant conduit on Drawing El-1702-02. It must be emphasized that in these electrically initiated fire' tests, " exposed" cables in overlying trays were subjected to the high temperature gases (approximately 1000 F) from the fire without damage. This provides additional assurance that cables in conduits at more than twice this distance above a tray will be adequately protected. The separation of vertical trays from conduits (category 3 above) shown on Drawing El-1702-02 is equivalent to that shown in IEEE-

384, 1974 for redundant trays and therefore does not require further justification, particularly considering the additional protection afforded by the conduits.

4

r i In'.compcring rigid conduit to an enclosed tray, it should be noted that conduit has a. Heavier gauge body than tray providing a more effective heat sink than equivalent surface area of tray b. Threaded connections providing an essentially air-tight medium which inhibits internal combustion and effectively isolates internal events from the surroundings. c. Si:e limited to 5-inch nominal ID thus limiting both volume of cables (combustibles) contained and exposed surface area. d. Curved surface providing radial distribution of heat and therefore much less favorable heat transfer characteristics to or from an adjacent tray than a flat surface of equivalent area.

Thus, when a -conduit contains no safety-related (Class 1E or associated) circuits (category 4 above),

it clearly satisifies IEEE-384, 1974 requirements of a barrier *. The barrier need not limit damage of non-safety circuits to any level. Consequently, only failures of the non safety-related circuits affecting safety-related circui'ts are of concern. Logically then, a conduit containing non safety-related circuits can be placed up to 1-inch from the

top, bottom or sides of a Class 1E open ladder tray since the conduit provides a protective barrier separated by at least 1-inch from the Class 1E or associated circuits.

It is recognized that the converse is not true and conduits containing safety-related circuits may require more than 1-inch separation 'com open trays of a redundant train depending upon orientation of the conduit and tray. This has been considered in the separation criteria shown on Drawing El-1702-02 where in

general, the minimum required separation in any direction is 12-inches or more.

The allowable separation is reduced to less than 12-inches (1-inch minimum) only when the conduit does not extend above the side rail of the open tray.

• IEEE-384, 1974 defines a barrier as -

"A device or structure interposed between Class 1E equipment or circuits and a potential source of damage to limit damage to Class 1E systems to an acceptable 1e. vel." l l 5

e Es'sults si Basirsis 1 Analysis were performed (Ref.

3) using finite element techniques, and computer heat transfer program HEATING-5 to determine the effects of an electrically-initiated fire in an open ladder cable tray on a 1/2-inch conduit located 1-inch away either beside or below the tray.

Key parameters taken from reference 1 characterizing the tray fire were the vertical variation of total heat flux (worst case from October 5, 1976 fire in Figure 11 of the report), flame and gas temperature, and duration of exposure of the conduit to the heat source. The model assumed the heat flux to impinge on an 8-inch segment of conduit located directly below the fire (This was considered worse than having the conduit beside the tray where much of the radiative heat flux would be blocked by the tray side rail.) The heat f l u:: was assumed constant in this region. This assumption is conservative since the report (Ref. 1) indicated that "the flame =one does not comprise a continuous line fire, but instead consists of one or more "axisymmetric" luminous zones which are on the order of 5 to 8 inches in " diameter" at the base". No credit was taken for the decrease in radiative heat flux with increasing distance (note that conduits located 1-inch below ladder trays are actually more than 1-inch away from the cables due to the height and thickness of the tray rungs which raise the cables approximately 7/8-inch from the tray bottom). No credit was also taken for blockage of heat flux by the cables in the tray or heat absorbed by the cables in the conduit. The maximum temperature calculated on the conduit surface was 357 F (180.6 C). This temperature cccured at a point directly below the

enter of the flame (mid point of the 8-inch conduit segment). Temperatures dropped sharply away from this point along the conduit to about 240 F at 4-inches ~,

and below 170 F at 6-inches. The maximum temperature calculated was not a steady-state value due to the transient nature of the event (approximately 6 minutes) as shown in Figure 10 of the report (Ref.

1) for the October 5, 1976 fire. The report characteri:es this fire as "one of the most intense and longest duration of those studied".

C90ElWs190 The analysis performed presents a comparative basis for evaluating the effectivenass of CPSES separation against cable tray and conduit configurations used in actual fire tests. The Sandia report (Ref. 1) referred to provides a characteri:ation of electrically initiated cable tray fires which, as stated in the

report, does not vary greatly from one fire to the next.

One of the objectives of the test was to use cables representative of those used in the nuclear industry. The report indicates that 10 leading architect-engineer

firms, 13 utility companies and 13 6

r l cable manufacturers were included in the industry survey which preceded the testing. Twenty (20) different cable types were screened on the basis of popularity of

use, small scale electrically initiated cable insulation fire tests, UL FR-1 flame i

test and pyroli:er and thermal chromatograph testing (which measured insulation outgassing as a function of temperature). The cable constructions tested are representative of those used most extensively at CPSES, namely XLPE,and EPR insulations with CSPE (Hypalon) Jackets. The cables used in the full scale testing

were, as a worst case, all XLPE insulated, with single conductor cables having no Jacket and 3-conductor cables having an XLPE Jacket.

Summari=ing some of the more significant findings in the Sandia Reports In electrically initiated fires, the intense period of the a. fire persisted at a particular location for between 40 and 240 seconds

• before die-out began to occur.

This is less than the time required to consistently ignite a tray of IEEE-383 qualified cables in the propane-fueled exposure fires (typically 300 seconds). b. In the electrically initiated fire, cables in the tray 10.5 inches above the donor (fire) tray were exposed to a convective heat flux of about 6,000 BTU /hr/ft2 which corresponds to a local gas temperature of approximately 1000 degrees F. The circuits remained functional and samples of the insulation from the bottom of the tray over the fire which were given elongation measurements, showed less zone than a 10 percent increase. c. The luminous =ene of the electrically initiated fire was optically thin which enabled immersed objects to radiate heat to the cooler surroundings.

Thus, equilibrium surface temperatures of engulfed cylindrical objects varied from about 1200 degrees F Just above the tray to 650 degrees F at a

height of 10 inches. (Note that minimum vertical separation of 24-inches utilized in the CPSES design is more than twice this distance and maximum temperatures are anticipated to be well below temperatures successfully withstood during the fire tests. d. In the, electrically initiated

fire, heat transfer to immersed objects is convection dominated with radiation accounting for no more than ~O percent of the total heat flux, even in the luminous region. (Logically then, conduits beside or below hori: ental trays are shielded from the major, convective heat flux.)

• The high currents required f or cable ignition open-circui ted the conductors during this period, removing the fault current.

~ 7

r I Computer analyses (Ref. 3) of the effects of the most severe fire encountered during testing (Ref.

1) on the smallest si:e conduit used at CPSES (1/2-inch) resulted in a

maximum conduit temperature of approximately 181 C. Actual temperatures e::pected would be appreciably lower due to the assumptions made in the analysis that the heat flux resulted from a continuous 8-inch line fire and the fact that effects of distance and cable blockage on the radiative heat input flux was neglected. All safety-related cables used at CPSES have an emergency overload rating of at least 130 C for 100 hours per specifications. In addition, the cables are designed to withstand temperatures up to 250 C under short circuit conditions. The fire analyzed will therefore not subject the cables to temperatures e::ceeding design conditions. Addi ti onal evidence which supports the' adequacy of CPSES conduit separation is provided in the results of the propane-fueled exposure fire tests (Ref. 2) also conducted by Sandia in which conduits and trays were included. In these tests, 14 trays were stacked 10.5-inch vertically and 8-inch horizontally apart. Directly below each tray (except f or the bottom tray e:: posed to the propane-fueled source) was a conduit containing additional cables. No separation was provided between any conduit and the tray bottom. Although all circuits in the conduits above the third tray failed during the exposure fire (the conductors short-circuiting to the conduit and each other), circuits in the lower two (2) conduits maintained circuit integrity throughout the duration of the exposure fire. Considering that the fire in the lower two (2) trays was more severe than in an electrically initiated fire, being larger in si=e and of longer duration, the results provide a conservative indication of the adequacy of protection offered by conduits during the less severe electrical fire even when installed as in the tests (with no separation of a conduit from the tray bottom, and conduits only 10.5-inch above an open tray) with significantly less separation than provided for in the CPSES design (conduits separated a minimum of 1-inch = from the bottom or side of a tray and 24-inch minimum from the top of an open tray). References

1. Sandia Report No. SAND 77-1125C 2.

L. J. Klamerus, " Cable Tray Fire Tests" - IEEE paper A79091-0 (SAND 77-1424)

3. Gibbs & Hill Mechanical Dept. Calculation No. 600, Rev.

1. 8

... ~. SPEC, HAH, 'IUGC0 (2), AM (4) A PAGE 10F 3

• '~

'~ CHANGE IflDEX:0EI COMANCHE PEAK STEAM ELECTRIC STATION

II

) DESIGN CHANGE AUTHORIZATION

III xx Jr//

(WILL) ('MXEG) BE INCORPORATED IN DESIGN DOCUMENT DCA NO. 15,917 1. SAFETY RELATED DOCUMEtiT: XX YES NO 2. ORIGINATOR: CPPE XX ORIGIflAL DESIGNER 3. DESCRIPTION: A. APPLICABLE SPEC /3%3/3333MEXK 2323-ES-100 REV. 2 B. DETAILS Revise the following paragraph and sketch details for ES-100: 4.11.3.2. Separation Distance for Conduits (2) Minimum separation between a conduit containing safety related cables and the top of an open tray having different train or channel shall be 2'-0" in cable spreading room and 3'-0" in general plant area. When it is imoo>sible to maintain this separation, the distance may be reduced to one (1) inch where a solid cover ~ is provided (see Dwg.2323-El-1702-01, detail 38). Minimum separation between a ) conduit containing safety related cables and the bottom or side of an open trav (solid bottom or ladder) havina different train or channel shall be one (11 inch. When a conduit conduit containing non-safety related cables is above. beside. or below an open tray (solid bottom or ladder) havino different train or channel. (30$ inj,e] g age 2) 4. SUPPORTING DOCUMENTATION: FO'R 0"mCC ,DC E I V E D l gggER3G BSEwes3 5. APPROVAL SIGNATURES: CE : C00UisENT CONTROI-25-83 A. ORIGINATOR: 9/J DATE /- 25 -E 7 B. DESIGN REPRESENTATIVE: A/LW DATE /,26-63 6. VENDOR TRANSMITTAL REQUIRED: YES N0 XX 7. STANDARD DISTRIBUTION: ARMS (ORIGIP.L) (1) Clark Conzatti EE (1) QUALITY ENGINEERING (1) Fred Powers EE (1) l ) TS FOR ORIG. DESIGN (1) WESTINGHOUSE - SITE (1) COMPLETIONS (1) l JERRY HENSON-PROD C0flTROL (1) DCA FORM 11-80 Admin. Rev 7 4f l

F,_ I -o DCA # 15,917 Page 2 of 3 y DETAILS: (Continued from Page 1) Paragraph 4.11.3.2 minimum separation shall be one (1) inch. There is no, separation required between raceway of same train or channel. See Separation Sketches "A" and "B". s / ,f* 4. f. 'M g W-9 l

DCA # 15,917 Page 3 of 3 Separation Sketch "A" Train "B" Conduit r 7 Solid Cover On Tray Required When Vertical Ladder or Solid Bottom Tray Separation Is Less Train A or Train C Than: (a)---- (b)---- (c) With soli cover acceptab e (Min.), 6" preferred. (d)---- 3) Separation Sketch "B" Train "B" Conduit ,e Tray Requires Solid Covers If This Distance Is Less I I Than: Train A Tray, Ladder (a)---- or Solid Bottom (b)---- Section "A.. (c) W th solid covers l' acceptable (Min.), 6" preferred. See Plan for min. length of cover. - ) P

s V y,y.g g y l Gibbs E; Hill. Inc. ' ~

• s-D3:a e

8 P.g. '#,ra 10001 ,v

• .e. :c*e 2T 763
• a<.

,.g s 4 GTN-71284 March 6, 1986 Texas Utilities Generating Company Post Office Box 1002 Gl'n Rose, Texas 76043 e P h Attention: Mr. J. B. George Vice President Project Gen. Mgr. Gentlemen: TEXAS UTILITIES GENERATING COMPANY COMANCHE PEAK STEAM ELECTRIC STATION ' ~ ' G&H PROJECT NO. 2323 ~, CONDUIT TO CABLE TRAY SEPARATION REF: GTN 71266 DTD 2/28/86 Enclosed per the referenced letter is a copy of Mechanical 'Oepartment Calculation No. 800, Rev. 1 for your information and use. Design review of this activity is. complete and documentation has been transmitted through normal procedures to duplicate file. Very truly yours, . o, GIBBS & HILL, Inc. ~ REBd-JIr :lc Robert E. Ballard, Jr.

l Letter Director of Projects CC

. ARMS (ll&R. Site) OL I. Vogelsang (TUSI Site) lL lA 9

oraJ3 y e . 0. r

e n DESIGN REVIEW RECORD FORM Te xas **t:1; tie s Se r n = e s. Inc. Comanc.te Peak S.E.5. 22:2 CLiEti! PR"'E;T G s d .'0 0.t. l

Title:

Cen.{u.t Te m peratures Due:no Cc h ie Terue F;co l[l Drawing @ Calculation l[l Specification %OO l I.;t - % DOG'JMENT NQ. M QA!T. COMMENTS ARE AS NOTED ON DOCUMENT SEEETS LISTED SELOW EXCEP" AS 3TATED MEREIN: 00h I k ' ,4 [, zk! &- !f m ~ _ _... em RE;lUIR$D ACTION SATISTACTORILY COMPLETED YES Cl NO Cl CCMMENTS 1 ,DE5IGN REVIEsi ENGI.NEER REVI U CA;E \\ e

Page 1 of 3 e MICHANICAL DESIGN VERIFICATICN CHEOF*.:ST CAL'".'LAT!!?;5 A?;O A!; A.LYIII Project Cranche Peak Stern Elec ric Station G&H Jcb No. 2323 Filing Code 300 Rev. No. I Date 1 % subject Cond uW hperxtures heiro Cable h Fire y Considered by Item Des. Rev'r 1. Appropriate Nuclear Safety Related designation marked on cover sheet s 2. Filing code, revision, and page no. noted on each page 3. Preparly signed by preparer and checker 4. Purpose of calculation properly stated e-5. Input data properly listed and referenced 6. Assumptions are, reasonable, properly listed and referenced-l l 7. Items to be re-verified, later in design, identified a 8. References listed, including revision no., page i no., letter no., zection no., ute. as applicable // A l, ?_, 3 t 9. Method is accepted practicer formulas applicable, j referenced and identified by equation no. or page no, etc. t l 3 l

Page 2 of 3 MECMANICAL DESIGN VERIFICATION CMEOTLIST CALCULATIONS AND ANALYSES Project Cm: e Pe."< Stes:- Ele:~r": 5'anc. G & H.*: No. 23:3 Filing Code - 800 Rev. No. l Date I.1 - ? (- Subject Co ndu;t Tem,perature s During Cable Tray Are Considered by Item Des. Rev'r 10. General approach and accuracy are reasonable; output reasonable compared to input 11. Spot check of mathematics or check by alternate method indicates accuracy is reasonable l 12. Computer program approved for use ,a /er,.m 5 l 13. Consistent with project guide s I i 14. Consistent with FSAR s 15. NSSS and other vendors interface requirements i complied with referenced v .i i 16. Codes, Standards and Regulatory Guide require-ments complied with and r.eferenced 17. All required modes of operation considered and ~ listed __. II. Safety class / Seismic category identified 4 19. Interface with other calculations and other s _, ,&isciplines listed and compatibility verified

a Page 3 ef 2 ..I ::n.'; ; ; A" O II: ::.. I :T :' ::- :: ;;f _:IT w CA:.C'J:.AT: 0N5 A' 0 ANALYSIS Project C ranche Peak Steam Electric Station G&H Job No. 2323 Filing Code SCO Rev. No. l Date / 8[- Subject Cenduit Temoecauces Durina co6ie Tras rice n a Considered by Item Des. Rev'r 20. Results to be used in design are identified and are responsive to the purpose of the calculation for sufficiency, accuracy, safety margins, and compliance with applicable Regulatory Guides, l Codes and Standards, etc. s. l 1 21. Complete equipment parameters are listed I NOTI: As a minimum, all items on this checklist shall be c:nsidered by the design review engineer. If relevant to the inpu ma-terial being reviewed, the item shall be check marked (/), otherwise the item shall be marked not applicable (NA) by the design review engineer. f/i W~ l / i j. _.; - .j.. ' y Signature of Design Review Engineer VA C a e'#'_ d S 1 E l 9-7 + w-- +wg.y ,-.v.w-- .---.w.%---.y.--- w. w.w -.m.-,a m s,. we.,-,ce-. e,. -,e- ---w e.--

Gibba 8 Hill. Inc. Calculation Cover Sheet 05 _ i ~_ - C, e : ~ Ca>ceiation Numcer 23O Number of Sheets in Originalissue 7 Subject c C NCUiT T_m PERATURES C u p,;n g c f. e. ,em s 3 NuclearSafety Related O Non-NuclearSafety Related-OA Program Applicable O Non-NuclearSafetyRelated Sheets Sheets Sheets Job Engineer Deleted Added Revised _ Signature l Date Original X'lXlX l./, I 5 5, Sa.356,5c,54 1,2,3, u,7 h Se,54,%,9 \\ h-ql'. l 9 6 I C ~ i n 3, i i \\ i i \\ l F-167. 4-81

Gibba S Hill. Inc. Job No. ' 1 T 1 Client 1 ~. - ~ Subject 00:1C J ~ P S p u '.' P 7 : C.: :.r G cp. ? -I r-Calculation Number 'a o o Sheet No. 1 . r,. f * " yg j 7 " fa~e ' ~ a, ;

as

s Cae Ae.

Date Aev Ca:e e ~ f7. t )<_~ Xi X X X ~ 3.. car., ca ; n m ca 12 ui e i i ! Checaer V L ~7/3: /C5. Af. 2/?:'/16 i l l 'l Y r,6 l P u Rpo s : To cat _c-) to E Tr> = n :r P '_ W J P ~:. 5 3F 2 L 2CTR "F-L C.C N C,t) 7 '3c.KTID N 2 Te A fs C.' 3 T P A f torm m tv2 re :nN 'N 1 GB fi (FRom AN E LE.CTRICA LLY T. M IT I AT E.D F IR E.'). C O M bu. t T S ARE t oC ATEb 5 s. a i: :s4 '5 'IM G TRAY 'g ; A E b 'l E 7'H6 TRA/ E5 LOW Tr' 'E TPA/ Tdn3 " A LC U L6-~ ic N L1 M t"5 1 5 ELF n A T% 'i 2 2 t I Atl4Ly Ot3 C O NSIST ENT Wt TH TM E F\\RE c.HA R ACTER \\STIC S PRGSENTED IN REF.1. TNE REsuLTs sacutb N o-C+. E L:s?.c i FOR OTHER SITUATio ns WtTe otK CormittAN e G.2 A P P t. t cA 2. t.. / a OF Tots C A LC U LATiON TU TH E W~ d. E R ~ ~ J 6~ C t i d. /.25 lt*P O N S: 1. HEAT f t. u 's U S t:. Q IS ~ Assumso ro o w s Equ c a E L'y l!! En vE tcP E TaE FLUX [:R O M 1'k1 M e&T SEVERE F i t= 2 i O E S C.R \\ G ED nN REF,' l. 2,. C O N C u m M 73 PA H oF corr e us' O tl '5-r, ~. 5 AP2 Co re.P - 3 E' / s u P.R o J NC. ED G '/ T',.e3 2 - c,, s 3 FOR TaE case e E m G A M F,1?/ ZED, Tx2 O'BlE Tit 5' IS -c ' t E t A14 C pet 4 A EE A, A LLOWING /\\ DEquATE S P A 0 3 FOR F RE ; co N " 2 C.~ ( A. Come.N a o H 9.-- 'tr.HOEE R ccErri : t t >*., Fo R A iP. F LO W (, CONVECT a I . S R ADIATioC ARE CALCULATED C O N S ERV A71V E L'/. Checidng Method # 1'W-m F-166, 7-82

Gibbs & Hill. Inc. Job No.

• . "; - 3 Client

^ o Subject ' c N c. ; - r - r-r., _e _ 3 t > c n.- u, ci _ - iv - n 2 r Cak:ulation Number oOO Sheet No. Y. ?-f 2;f C re aev fale ces ta:e sev

ae i a.

Cae e s':'. t 3< 'X ^X X 'X -r C KIE ' ' /1.!'14 t I i i P*eca er Checuer i t/ L 2 17/.? e/25i X 'Ut.In i i i l i l l I l l R G F ERE r'10 5 5 ' I.'CA3a2 "TRA'/ i tRE..:STE', S Mi f.. A LAEcR,. r.;3,, f,_E 1.P.J 2 R c, u '., r4. r/. R s p-7 9. 7 ri. si'c 7 7 -it 7,3 c, M Er,'r i rt, A N s F G R ', E r> c-2C, (. [G 2. -f, 1, ' PRINCtPl G S OF "iE. ' e4. 1 A TI N G 5' CoFPU"3R P R O '3 R A M O A4 I :., . ro A? : ! '.. 9 Lh mi A-~ C P /. 1 CM! TON 1. G EH ELECTRICAL DRAWIMG A'5Al-E.1-l'7ol-CA (Separat;on Craeda) I s. M _w 0 E E 9 5 6. C A- -.D 9/G/s 5, s. m r,n' N o r : s -: C.. r-c ( s G. MEMO. EE 99 D DATED i2./t3/ 95, g. g m ;1Hovica ro D. RAo. }I C A C U '_ A "~ 10 M : Tc. i s CA LC O L ATiCH F O LLCWS C L O S 2 L.'/ ~e l re. - - r' 0.0 U 2,'I t .N 'd ? E x P 3 R i M _t, S C E S C R S 2.0 1N R "2F A r l i'- .; ~ - - ~~.25 CABLE FIRE C H ARACTETh2ATio NS F%M REF.1 AcR e r t,7 r u s -s. (=Ltw.3 MC G AS i.:.rrp an.A u r : , A x: sas ci, - wx9s;:1R CcEFFICIEt4T. CM AGM.!!tSTics CF McGT 3.EGR. Fi -E ( is/s/ 6 Fir.1 !

y. :. a, nar.Q wsRE issse. (\\ tR HEAT 71t,AN*. FIR.

Co. 3 F r: e c t ENT E W3R5 T., x. Es F R. Oto R3F c 1 Ts 2 co re,p u Ir< P a e s a tiro F E A n s. 9 m i wis s u .1.. m

c. A :. f, 3

- Ere-m,- ut; 2 .rc 3 p 2 r: - - ,s _c n t.r ;, T a re. M.re.;te IMPsc<n JP2 0n n. : 2. - ~ > - '. ' t,.. OF ~~ ' 3 I r'r i C T. ~ '!! A E C 3 T TO M.9 i P., , ~~e' C. ~., ~~ '~ r s tM-omurr, Tsir,p =_r< r,TL r. 2 TaAT Tv2 cA e _ss i to s.. :, 2 CONC 3T wILL G5 D PC 5 GD To. THE LuMluouS FLME. HEIGHT } I Checking Method # 1' m-m F-166, 7-82 m .c,,

Gib6a E Hill. Inc. Job No. I .J_ Client -- ' 2-Subject _ce.:.; -

v-

~4 C'J3 :n ; ~ / 1 - -" r-a: CaJculationNumber g e Sheet No. 3 . = e. v.. f.- n .,e, t e. .., n e. ..are sev .ae W", i i D< X X ~X X . P ecarer +C<' n Si! D<R I /.l? 4 i ! C ec=er M Lu 7/1o/dS CM, //2r') e i l I I l I FLacTu ATES SETWE EM 4"AND10"IN l+EIGHT~ Pic 'T;F '. s-if A6OVE THE S MR Ni A)G 3" HIGH C.A S LE -fR Ay. ~ g ""- TTA/,'/ N f* ~. A '1.5 I I 3 D Q E. 3 d0G "?!i~A_. r F CP, '.r c E Q T* rte "o NC'J t "~ w, 0. " : ~ -) re - - -~; 'A ~~ a I Foot F gorit 73 ~R T-1, ORtEN EC V ~ P ' ". T._/. FCIR ;a s CC N C l-'1cN, FP err 1 R 3 F. I p.nT= 19 C o F 3 '7000 G J /.4t. i

• s

/ M Of i (, P. A c t A t~ t V Et. CON V EC '6 'J 3 H ST T F ' /J V n o J'_C G.3 N".3_ .5 IN raE lt E "n.O N '3 P P O '31- ~.:. Tri: i: J r.- E. m '.

r..i 3 l

t. /,i t /1~ F e t. A. P. .o 'v: _:;. 1 .2 x s. e> 7 ~'

s. -

AND -65 c.cscui7 c C M.8U S t c N A1 14 S O F ' ~.' ' .Y - t r. :. f (FORCED CoMVECTioM'i V R c r<

e'.1 R f.

C f i .: "i ~ - m rw :. P p .O N CF ~7 'l C /'. N C 2.9. 7A' f 1 ' '~

• I I ~ "-

~~' .i ' ~" t ' '". FL A rn ~ t eic 9 /. s- ' 3 . ~; '~ H 3 A ~~ ~7 ~ ~ ' - r -'- AT I 2.2. F f NATURAL CoMVECTiod (SEE e c_ t.ota, Fi 's. I') f de t o.m.ide d:a~ ate e I h AMBIEMT AIR / e 8 3 a I g' 4 i+ A: 7000 Gt-/kv't N au i g eo: r t @,, ). h e b, tJaturai C.,e_s*',, m ~ w.. ;.,, 3e.., ~ F.F. 2 @ h i he h v....ac.. m.. a,,d Rad:a t.,i I. g @ h k< %

• a *w a<..,,~,.

o 31. r. O naemm.d, g ;' ( g ggggg,gy 3,, 37,3L ,j ( PRcPE C 35: E

• 1 h - E. :~ 2 P-a s.r,2 c ern ) r - c.n a J/ 2 ~ 2 n.:

.e > g ^;,.,, }._,,,, y,_, ;, b.. @ cer oosm,,.if, 2 +. i ~ .\\ ,i .s g t g T. wems.: o e c w e.e c :c o c n e Ps.1:codouit e s.c5 ' c,2 r.we tAwnt er c a:.- gg mt (rta osia. c i t r s,.a, s,cw ;

1. ,.~._f c,, =..-

'm y c Checking Method # 1' m_e::_m_ F-166, 7-82 2. C R' T-I"' +) -- - m. _- - - -,

Gibba C Hill. Inc. Job No. 2^ ? Client ~~ Subject e_ont A-Tw F= Fn ro* 5 5 C.3 9 n c, 1/. E _ - P r,. = ': 2 s-~. Calculation Number 9 03 Sheet No. 1 7-- ,.a u.. a.e yre n 0 / 'g A / ~~ , P ec a e-c o. .p ; 4 ! Checuer V L.u i-/32,45 rM 7p.g i I I i ~ l l I CASE Q A55uMES WG WDWT Lo CA7C

2. 0

' H._ r' : s A a c ;- P 3 T11t A y AND P A R @ LL G t-To Ts. 2 Ayl S OF %.-

7. A j, p e g, T"ai s c o ra ctT"lo N, FRom AEF1, g _2, G,

op 30cc 3 p..jb.. FOR TH C-Ic]S/76 CASE,(7.5 is Ti1 m:,-

t.

Zela _ r 2 'p. 2.'- OF P - P. l ), Q G IN G W 2 P'3 7 ; C cd p...- 't h (, T-T [) 3000: 7 (T /D = 49 9. $ 'F. {- [2 ~ e ( c s

r.,- o c,. : -;

e - T-5 CASE A ssur 2 S W A~ >2 P 21 i s 8 INCH E S Lo % y.ig..c', /.; TkAT A F um c_R T4o F E.= 7-c,9: to u t,WT on EiTsEF 5 c. 2 iS L 0 cA 2D IN ST' t L A '. R AT i22 I - 24' _, B" _. -'- 24 _,@comBuGTow Q c.cnsus ion sn s orAs 3 Coo P_5 / .,t fj C U '~ '.-// \\ ) Q ce, $ = 20 p7,,,, 394 g g; \\faamesaur Am @ h

• b-he - 4t ed Co~ eel'.:r y

o e, q 4 33;c, I J v L \\ N TRM FL, AME FI G. 2. Conouir Anova TRAY s. 1 T - 70* O c'

- - ~27 v e

'. A,. I n 70 r. Y - ?:-- u h : 7 raTu., h -Fr2. p Checking Method # 1' =. 2:::g. F-166, 7-82

Gibbo S Hill. Inc. Job No. ', = ; L Client Subject C, r.4, - a rm e, 2 re - v.,,, 3 0 g e 7,.,, :,. 3 Cak:ulation Number 1oo Sheet No. 5 ~ .n -+. _r- - e.

re 2,.

ge

= e. lae 3: ", 74 3 X X~ X ^g . p ecaw . v t. 2.: : % 1 i Chec=or ! t I (7tE fim bei j I I l I l l l l i C.U e_b)_ O 55Ebkk.." Unit. COED

• Alt.__ICCa te:E. o n e i n ch._ b

. _ _. tho_ . _. _ g _ _. .c o.b) e t r o.y _.ct n d.. p a r s.!I e l. t o_._ Y h C d.e I C oA "k L_C.ic Ii. ~ icly. tee. condd+ add tU._cEEle fray .t e e G J :. E _ _ ~.. _ _ _C.M e&e bho rit.o nho.l ly.x Ecr-dis _Co nS0 v r o t-lc ry . (Sde.y-d c r.e nce. 13 p. AT)._.g. (ed.Tative1._ g ":. _~70 00 r g, p d a M5. _.he myeeatu re. ET[l930' 6."~ C: iee -Os :.h~~.: ~ b__Nou.ld be negilh}s.5n MLS h.2Se be:AA!.e c:v c4c-:N __.__7tr.__ wedd 5e.~ d.rlwnR4p. fed u.nder i r e tr >y.s-a. ~ !~ 4.cmiec2tst_.o_T" ta ; o F.. _ n ['~_.Ne, chte a.SSENes ka.t the' kire.is 3 i nc.hes lon,9 a v.d,

f. hat..-two__leitf4__ccMud.$on ellhc e Ude of the. @ra

~~ ~__ i_s Gcsed_w sP. t[_aicla.t _Mi*if, Z Z ~ ~ .LA ME ['.r_ N0N>Db W #>666 I C R#4 ANT

• A'T #

P< ._.__[, S/A - 7o co E u / v ? N1 h--TRM ~ @ h e be+h,- N d e a I @ 14 c o *F 1 ~7'1 hcounuiT. ?W/ i C.c mee t,*o n oed kad;de.e. f f--C ex:nr - ~ - - -s4r sw< ~ ~* y~{ 7ggG g, , n,J,/g g __@ walEnr A tR _ menWm) old cre cas hCA: ('s/.",' .. epals 0 37f,". There f.s re,% e "ksr.],i r .. _. _. __ h at leosf l' 07(o* -frew' 1ht bu t" ^] Ob' *

• CIG. S C04D.W T M L O W _. T R N f..

Jeq. T4 fope dewinjs GG 06SL-ta.06-CP, _4G-J85L-12-%-c P and GG-Job Q-06 -CP. ,.t., _ _ c

Cibba S Hill. Inc. Job No.

1;5 Client Subject

r.,,. A... +

h :.: c , e r :: T., e r C.; e i r3 2g Cak:ulation Number 2OD ~ Sheet No. 32 T.~. X: e X_ X X ~x .amw v2 g :., s. Crecner M !>Hr.%,i n i I I l l l 'i he_An6pra.UE _iEow n. k case._ b) has a l r e ad.. / ~ bese. a37c~e sc'e~d.._ii re f e~r.e de_.ll~ wit e ce3aci 5 'r_.o n d ui t s ~ ~ ~ . whi c.h~~ o.ri above ~ ~ -the b urni ng ~ca.b le tray b ut w i t hi n _.._. ~ ~ ~ .__oie Moot'..iF the ca6ie tray." The plot on page r,ts o f ' ' ~ ~ ~' . reference.I shoJs the temperature response of cLxpAedi_~ .___cabtei d~~heifhts of. h"l,'~6",4" and il " above Mi ~-~ ~ bur _nin3~31oi55,. Tne~casiwh'i h' resembles case b) the.' mest. li the. oni at hhill" heig'at. The raaxi w v m ~ ~ ~ ~6pestari_tecorded, inToe ~~6est wah.3 o o* F. The~ ~~ ~~ ~ ~~ _ _. -,._ t e m p e r a. ture ~At ao " Ill "be lower since the ~ ~ Tc~.omEu~stion_ g ses'iool 'as %ey rise,and, per reference i, ~~ d~ ~ ~.._ionveAttonh 5e doLinint ~ heat Wansfer mechadsm.i n -t Q s ~ ~ . '. _.~-c aw_.T_b_efe.fo.re _,_cde. Q. wi lll_nof I be' analy zed. ". _... ' ' ~. ~ 6:nce _ cases a) ard c.)..are some what (;mitar, ler u s ~ .___corisUer..ihe fotiewin[. A's the dis ~tance. between thec. ail.w;Ti ~ o _ __ and _We~ftii)nTr~eafis', o. s,ma Ee c. petcentae]ee +!5 c : c7,..+ ~ ~ ~ ~ _..__. Evf_h c__orco._. edieivei 45 rf.O of.h c. e r 39 In add;t;or, __ Uddia4;on' - ~heait 41G.virici inverse.(y with d.istance spJ ~ ared. .51.nce._the. co ndait in cas,e a) is farthe r 4,h a n % o n e _i.n Ia.ie K. the ' ede a). cond2t_ui.11" rec 5; ve, less rad. c. _'F es. _ ~ 7efehre sose..c.fl.h hoh. severe than case. :-) 3 a nd. _l~ ~ . sis e__o ~' Checking Method). TvJiiI'.'_not. be_an2yiecf.~__.i~" - ~~ i 'm_e_ m., p.166. 7-82 O Ca71WW 8D2 W The 3 2F1hw e@ M rpm pg 6 cf N m

Gibba C Hill. Inc. Job No. 2 ; :. ; Client Subject %,1.- To n.:.o r, , e a :. % c ,1 G c ' e '~e,.; : re Calculation Number 2oo Sheet No. 5S

• -.:. -'.- - "; f--

n 2,.

ce e,

a

e A e.

e new ~., a e a X.- X X X X X . P e.a:e e.: an t C. ecuet . ti i _ t M 4. i i t I i 1 i o n22.it s.._a.t-the 2 P555 ~~EejT..i n si z e' & a m V:. _de ~~ ~~ r 'o ". Eo d.~~A..d favr ete r. ~ '~or.d e c +o selec.? th4 coch t _ ~ ~ In ~ ~Tiz.elbiel_ wiii...eii u ltTi n maim.u E te -per2 u e<, le F.... ~ l'"__uCcadsde Cthe fe liowi~ g..~ Be pre se n ti the co rd u.it.' ~~~ r ~ a.s_a A eZG nain~3_KT.hTa6 o u ec e._

m. s o reo r_ _

Z:

e. ~ +b e E ~. p to u em.15 pen..in rik<i s C e.T.Js4T-~~~~-

~ ___ p =.. lh Pk A 'l.Tsi.To. ) ~ ~ ~ ~ ~:: Nc w. ._ P .2 r d a...- .-. A = 5 (d?_ d ') -{(do-di)(ded;) = [ t (do -dd t xhere.: k : cand. sit. _..outside J:ane'ee ._ di :..cond <it ins;de da%e tee t = conduit wait thickne sc ALsaL.p.,M.da._..._. a.s. d. h * (f,) ~ ....__ n e c. d o c. we..ca n :sa/:~~ ~ J J 4_._. p.(p, )(do)(da_+..d )4 7 T. (L T ) i ~ Corskat (/4 h() (47) ~ ~ ~ Z_IheicchinaSon. ~o~4. c[., d.;..Sd..t ~ ~ i.o bic h mi r i ~: z e 1. i ' e. e Z he~E b'l.'o e3j sa7e. N~et Erac.ket wi II re:.u l & i n O o_ ~ h;o.be st +eniper.6 ce.. h sed.on.the s in eru.aN cn id i f ~~~f re. e raince _ o, lhe i P2."I.condEh i s the horit ici.e. ~ Checking Method #~.!' m s - L --~ ~ ' ~ p.5, p.h i ~~~ --- -. ~ -

Gibbo E Hi/I,Inc. Job No.

-'i Client

,5'_- Subject ro,. L.. Te,:., s. 3.,,,. 3 (_o se

c.,, ::,. e Calculation Number 9oo Sheet No.

h .2 e. .a. -e. .n n. 2 3., e .~. tn 'X X1 'A' X X" o e.xv . % :!: in ~ t cn cm, i + 'M f re.iA 1 I I _._._ _.Fo r _c a. s4._c.). _.w e _.ra u. it. c.o m wh t're Pe rM o c.e o f.. 1he..Co$d.wi{- db u Me ef Ec.c.3,ic h. w

• Il rec ei se

"'7E ._. T1} %% hibY e.~S.*r* yN ?. .A. ~ 4 7 rrs 6:ameter .q. . i,,. _. _ j r From reference 6i c.o n d a it __..... ~ s 4'. s _. d.w5 0.haA in.=D.r.s.0.3fl i. : A.O.159 A. o s i _.,, /-_ s n s g* g'8.. a N 'do 0.3 4 0 in. -O r = 0. %.O irs. = 0.0 350 :: a o .-,,, c_o n..d 4 f. - O,0 0 9 / f t. __, 1 6: r - r g C, I C 9,*n. o u. J A ~ -- f a n of a--/. 2 + :. a:.17 2? I. S,., y

• <, ~,b.:,

p ,,,,,... 0. ;r,.:I .c. . o r.' d l t The hoh l a m u n t of co ed a i t to. .-.. j._ _ . l. _...\\ /. . receive. radi. tnt heat en'e]y. 's 9.- % _. . 3lo 0- a 9 4.7) '= 5 l % 5

• ty-

.V ~ ? Ain$* The ei~ls. c)e dbe. Ij'n Ae A r'j, \\ct ^' .s. s '.o e. '. ;.4 r.:e e . I'~ Ell _To G e 5 ond.4if.~ .2 ~ ~ ~

1. c:2 0 * '.4 9~ ~ ~h t. ~1 ' ~'.'(J_~hearenput)

~ ~Mi* <~9 < ho *-.. raat. ant heat-i ut ~ Checking Method #' '1' WTnCL ~~ ~~ -~ F 166,7 82 4 Cartsre rum aus m s ars.us ne m ruas se ' unde a, s.us meus

GSbba S HsII. Inc. Job No. .=:3 Cliem Subject ',. d ;.. To,. 3 c c 3 , r e. C.ir:, e c., ='c2 Cakulation Number ' EoO ' Sheet No. 5 d. = _n ..n ..n -n

. e.

a. y.~. X X , 'X 'x M _~ ~ ~ ~ ~ ~ o na v + v _, :. ; o r:

c. ecw r 7e.unw.

i e i i l i i i i i l l l _ T ~ T he l;7e~._.ya~r'ie.s M.t h tNe. a. s~ s ho wn in refe re n ce ~l, ~ %r. ~2is e c~), I.A s i nh.cu~e ve @ i -he co lari~e te e _. ~ -_.T heOp.eeAhri risss Ie:m 19'Fat t=0 su ed.: ~^ ~ c ~ . O___l ;L.o_o F__aV._.t :~ 3 0 seco$d.s._ ihe te-pero N ee ~~

~ ~ reN1Lns at il00*5.vp~^koTa.u.I se 5 ds] o.e'd deo71.

~ ko..&oo*1..' i~i~ t?=~5 '3 se~c~on'd's. ' Ge' rad;at;<e heat ?Ius she e 0no . _. h.--_=,Ix o o

• P _is 70cc na. T,e.

opp ro < ;ma tely Yn__o rd er $fo liteLinT_.t.h e. Ee s.t @ h.tx a t T,.:, F 2

• F

~ '~ s

TO eh e?e s 7 c.d ; e.o.

S O s' tenperab<re. ^ ? .vatt T.1 dc+ E ricd.~4;dt. . _ _ _ hear c' alo nce shown in ce cer ece ' hing foe. .. ;< 4 y . _.. h.LT.g _T ).__t s G"( T,.Ts ) = (1 - s) c-(T.- i m,) s g .-Che_:Slaes :tempe ra t ure.. con be de terwe +d. .. _G : v e v. : 5.<, .-.& - 9.!5,_.h .7 n,.g,.m ~ ^ _ f G '; _T.: 40oF= $6a"R. 'i~s ~'f.= = E 5 o *C s 1 ."i$$(c6)*(o.)5)(C.ni+sio ) -(26o7 =(1-0Jfj(,0.p1615)l(2' sit *(~:$j 5 3 = .___J T, - Goao ,;:,. 5 It x icJ,oT,, - ; +]. 6 + - 6 21 42 4.. . r ).57f.xto*'* T.s4_? 1 Ti = 62 +2 6 L ~ ~ ~ ~~--" .Z

l. ~ T.~ 9 +3 R

'48 3 F (Fio-e ie,.p,~ ot re ) 4 r g g . { _g __. _ ~ ~ _

G.ibba S Hill. Inc. Job No. ,1. : t C'.em Sub;ect C r. 3,. a % u r a., e e 1 3,v a C, c.e =: r a Calculation Number 930 ~ Sheet No. ep - i+ 4. n e. a:e L e.

a

e nev

a J' ~.

'X X X X 3 (~- ~ ~~~ av.arer '/ a ~ ' i i t ;s t . C. ec=or . M. I 2/n./MI i i i I I I l l l i ~ ~ ' FroJ ue 4la e _ ter3palaine,. Ik4. beat flux !e a vi r o, ~~ % E' 4la Je..wiil 5~e' _c a~tca{itri. ~ .o o,,: 4cr (T q T.34 .f - gr.. f. (0*I5).{0*17)4 H O) NO ~ [EM)[ L =. ..-.o, . S.Ttt ._f r _E-b.7,.O.hr.-ft.__.._ @...T:=VooY) h _E_. 7U -GIfowing_i npu.t wIIIIe~ ~dEd._7 __. ~ __. .. Z~" . y ~ ruj.\\ . _.._ _ l i m C_ ( i t Co n d 1).. ... 41 eat..MIWy. ( nc. 41- / _..O.C.. _ 0.O . 50. O__ - .. __..._.__ MO O. _... _ . _~1 o o Q. O.. 1 'o 0. 0 1000 0 . 570 ~ The. hee. 46 M assmed to i-Pige on the S inch e-e3me. t of the. conduit which is directly below _tk T_. 4i7e.. 'Th< heat f u i s a s su m ed h be constut i n 6.i s ~re a/o n. Ajo c.redit is taken~ 6e the d_ec. r ea t.e. in radi atNe. heat flu win. increasing d.istance. .._ Nje_ _ heat b is applied b the re mainder-of the. $. _ L o n d u.i 5. It-is more. conservaNve with' redard to %e . _..__Jna g i.m u.m. condui t...+.empe ra t w e e to l agp y a cons %.e.t - ~ ~ .heifiIcb._.6 tkES Ech ~~cefion a e. d n e eg e c t tke ed ~ .lGectfi 4._the..'hea+ f (a i n ike c+ hee veq;cr.. No._ Add,1.t' is, tak.en.for..biockage. of.. heat $b.u d u.e to .Checkin.lbe. di.steice of. odlacent c.aees in__ the_ tray. __ _. p.1Ei6, 7-82 ~ g Method # !' = m m., 4 Car 484 '94 WW'undt 3 ermas m@ W #pAB WW1 mas Of res ammt

r Gibba C Hill. Inc. Job No. 1: 1 Client Subject C c,,. g,, + Ter,._ r3.;c t % c,, 3 c y,,e ,,.s..

c.3 Cakulation Number 4OO

~ Sheet No. 5: _a - e. _ a.s r u. a >cc ~X X X ~x ~

o. car u

=. v, t C. ecaer i ! eld frygf.9 i i i I I l l The.__c.oncliit.i.s ish7.ed. to..Tose ~ heo.t by con /ec.Wo'r.. ~ ~~ ~~ o.ncC~ridIi.ifo~n ho 3hi.es.2ronier.tl.. . ' _ _~ ~ ~ a.)_doku.cs.\\.Lo5ic tion Sarn 4orizoe.ti\\ C.y % d e r s. ~ h[d - o. 5 3 (C,qF.r_)2... _. ..(re;.4,.759. .. W .whcf.4_.Pr_7 d.',5 _a.nj..._ i #EGe'~<l0 ~~" __T ~ ~~ t l , FM reE.'A]. p./o3 /e :. (dir_.ySpe.rtEs [Md$F) ~ ~ ~~ ~~f=0.isxto'.lvp' .k: 0. ol1+ $,.$.@~l 3r_= b 7A, h_ o, 5 3 -k _- 9 $f *d//4D \\ [p")l 4 i ... _d o. _., _P a_ __. /....J_ y -- (o.53) (0.0/ N),-(0,E54 !o") (0 0V Q. d. ~!

• f(AT)-

3f . __ h : / ( o.07)..__ .6.:. 0. 5 ( AT_) *. _. $_ ___8 lhd io tj eqZ.l.' _1 1. ~. ~. [ ~~ _s E_o;3._ tia S a oT:ud sheer (re..,.7.c.i7'. r .___C :.. (0 3). (0._1 I I + x_.10-g 1_._ __5,1 +A x io",*,,gr., = . y.,e. .. ___ ( Fo r._conservatit m... the shape factor is ascu, red to be I.0) p = Checking Method #' 1' c 55t.C ~ ~~ F 156,7-2 ~ ~ ~~ ~

Gibba S Hill. Inc. Job No. ; 325 Cli:nt ~ 2 :: Subject C=D-~ ~~D9 2 R A ~u m

R%

3. s 3

-O< = '~ r-Calculation Number 2o: SheetNo. 6 "b.i l-lA Ee. Ca:e Aes a:e Ae. Ca e i Aev

a e

.i 7.~, 3< X X X Mc P ecater 3Xt ~;;si L<R l ' m 9.,, I Checner '/ a !1/ i oj rst .'hk. i;.j26/r6 i l I I l l N P ' ~~ C A ~~A O 3 ~ FOR Us.3 3 C //MLU$ECTF JOB (C48.'11320'.'002323006','5076'),'M.LUBRANO 15FL -1019', // CLASS =C. NOTIFY =MLU,MSGCLASS=0 // EXEC HEAT 5v2 //FT06F001 00 SYSOUT=* //SYSIN 00

• CPSES ELECTRIC CABLE TRAY FIRE, 1/2" CON 0Uli 1" BELOW BURNING TRAY 2400 1

3 2 1 2 2 3 3 1 9 500 -1 0.00002 0.0 0.15 1 1 0.0259 0.0350 0.431 3.14 2.0 2.33 1 1 2 2 0.0259 0.0350 0.431 3.14 0.0 2.0 1 2 3 2 0.0259 0.0350 0.0 0.431 0.0 2.33 1 2 1 STEEL 1 22.0 490.0 0.11 /' 2 S1 EEL 2 26.0 490.0 0.11 (* 1 122.0 1 1 122.0 5.14E-10 0.5 0.25 0.3 1 ,) -1 2 1 122.0 5.14E-10 0.5 0.25 0.0259 0.0350 1 0.0 0.431 3.14 2 6 0.0 2.0 2.33 24 4 1 5 0.0 0.0 0.008 7000.0 0.067 7000.0 0.1 67.0 0.2 67.0 /* // i l l l s, i I l Checking Method # 'W i _ M-~ F 166, 7-82 l

T Gibbo e Mill. Inc. Job No. *r n.2 Client ' Subject.:nt.. - -Eru 5+r, 23;

c. _

. 3 _Aa_3

_ g..,

.- 2 r Calculabon Number 8oo Sheet No. 7

.... ~ ~

e. I

~ re c. ae

. e.

.r. ~~ e 3."; ', n ZX' X_ _ X ^X >C e:a e-0O C 4 R ' */., $ e. 6 t C ecuer 6/ L ; 7/1 ::;r i & Jn,/f

• l i

7 l l l } l-l I R E S U LT S *. FOR A CA BLE TRAY F:lRE, THE TEM PERATURE OF REAREY C.o M D u )T S WILL INCREASE W IT4 TIM E U N T IL. THE IMCtDENT H EAT Flux. DECR EASE S APPREC.l A B LY. GOR CASE,C) THE M AX I mum TEM PE R ATuR E O F. Tt4 E I h." coM Du i T. IS 357 F (SEE COMPUTER. PRtNTouT MLu$ ECT F JOB 9 39 PEER #R'i At 1%). THIS CCCURS AT A PolMT 3 DIR E c.TL'l BELOW THE C. ENTER OF THE FLAME AEcuT Mo SEco@S AFTER FIRE IMITI ATioM. PLOTS OF MAxlMUM l C.oMDu. tr TEM PERATu RE NERSuS TIM E AND coNCulT s TE MPER ATuRE PRO Fit.E AT THE TIM E OF M AX t Mu M ~E M PE R ATu R E ARE OM SHEETS S AND 9. I N CT'E TH A T THE TEMP 2 R A i'Jii 3 G L.G T"II. A E. 0 / 2 0 0..; e i N ~ 4 E PtRST C MIN t,,T E s FO L Lo mHG iniTi ATIO N OF W3 F'.ft2, r Tr* 2.'l Oo NOT R. E P R E S Eff" STEA DY ~5. TA T3. Co t \\ C n c N E., Tr4 E HE AT Flux FROM THE MOGT SEVERE .iL20.W i!. ALLY I4nTiAT3.D FiRG WAs USED tri mia cA tc JLAT ou T. Mare-CcQPL35 3/3" AB OWE SU R NIN G CAEL23 Q E 00R C. G. A c 1 '.rG 1 ~ { i )N INT *i H 3 t T'/ I': Roto % I toos To 440cF wim N S t y. MN w i. (,R E P i, p. l.[) ; eJ.

• r4 FIR 5 WIN G l"* 't D E'M 1A p 'a AN ER FCu A M inu "ES,

EEvoNo (=oVR M I N U '" 3 { h F I f' $ C I-7: 7 NATJR3 WILL f7 2 -3 0 ' T g r4 f G M F'E t A i J R I J.: LOA 2R T*rAN L Tr' S M AX tMUM ""r 2M P' E MA T U R EG C.A L C U L A-~ G o I N T,4.5 AWA LYS LS. Checking Method # j WZ:::2.-- F 166, 7-8? .T.:. M ';='. M m -a- -

T' Cibba S Hill. Inc. Job No.

=.. h Client

~ 1 '_. '_. Subject e,-o.: e-

2. c 2

._., c z c,, Cakulation Number ?O Sheet No. Si i i =,... =,. -=. .s. TJ:5 TX 'A: eX ^X >( a e re-u a _:.: n i i ! Checuer T ?!.c /g i t i I i i l i i I i i OMe c.).C:r d i t. Me l

• wm Te m perd wr e vt eu T; m e.

a.t_:the_ poi nt of.-oie. -te-pe ratu re (z.-2.13ft) . TA _ _31,. X.:*OO s* .$.w e.. .e... b .g p_ 4 f y__. g _y - ..s.. __ _.m.._ 98 ~, M.. .g ggo, .-1 L.. t ,y. A.._ _ C g.. _ yzy / -.. g. g. . O _.. _%) _ 60 ._--O - e-t0.__. .ito.. I30 .. 2i40. 3oo _ ho to 4Eo -- N t Timt Ott) Checki5jMethod #' i,d..., 62. L~ ~~ ~~ ~~ ~ ~ ~ ~ F-166, M2 e rse eman ow wems s eram, e%,mm re suas a w me===s

f. Cibb3 E Ni/I. Inc. Job No. "575 Cli:nt Suoject

t.,,.4. - -.

,,. n =,, Cakulation Number '2 9 Sheet No. 4 I =,.. =.

r.

=. ~.5:. DL< 3< X ~x 3( ~ _= e :,- v_,1 v, Grecter i 2.^M., . i _,,91 l i i i l I i i i C.o.s.e c ).C c e d u ; + bpsmv ee. Prcfile at the i a me o~ Moanue Temye eau es (z.19osecord:) i W) ^ ..f/g.,. l I s u u I I I I i .. goo. i i i l i gay.y i .s ~ /;, 'J Q t E-95 10-- l ~ ~ M l i v i i n W t. t m,20- ~~ t i r -"2e 7 ,= Cc r.l, -

5...

, % ie.!ted, 5 ; n., z.,, V..

7. e. # :

' 0 -.. w ,,e-e = ' =v.p..~. t e 6.. 4 s it. 16 tc 2+

.?

n D;t.bnes along e.onduit I I i hocking ethod #, i __. ;_%. _, F 166. 7-82 O N M Tb4 W WW"D F 85W@ 84h.S eng thd5 # grung,3i,gs +

.aa..... (* CPRT-294 INTEROFFICE MEMORANDUM TO: Howard A. Levin FROM: John J. Ma11 ands DATE: March 12, 1986

SUBJECT:

Action Plan I.b.3 Design Observations During the implementation of Action Plan I.b.3, " Conduit to Cable Tray Separation", the Electrical Review Team noted two design observations that by themselves did not indicate an adverse trend. However, I believe these observations, since they involve design criteria, should be included with other findings generated by the Design Adequacy Program (DAP) to determine if an adverse trend exists. The original issue as identified by the NRC is: "The TRT found no evidence that the existing C&H analysis for establishing the criteria for a 1-inch separation between rigid conduits and cable trays, as stated in C&H Electrical Erection Specification 2323-ES-100, had been evaluated by the NRC staff for Comanche Peak. This analysis. should have been referenced in the FSAR." Upon investigation of this issue, the Electrical Review Team noted the following two design observations: 1) No analyses existed when the original criteria was incorporated into design and construction documents. The basis appears to be engineering judgment based on experience with other nuclear projects. The one inch separation between safety-related conduit and cable trays was originally sent to TUCCO via Gibbs & Mill letter GTN-2441 dated February 19, 1975 which included the document, " Criteria for Separation of Class 1E Equipment and Circuits". Additional criteria involving conduit above cable trays was added 'to the Electrical Erection Specification 2323-ES-100 via DCA-6132, Revision 0, dated November 16, 1979. Again, engineering judgment appears to have been the basis. 2) The Gibbs & Mill analysis eventually used to verify the adequacy of a one-inch separation between conduits and cable trays contained inconsistent assumptions after design review was complete. The latest revision of this analysis is attached to letter CTN-70439 dated August 20, 1985, and the Design Review confirmation was transmitted via GTN-70614 dated September 23, 1985. e

I s 4 CPRT-294 Page 2 Two assumptions that were considered inconsistent are: The analysis states that the smallest conduit size is the worst case since the only heat dissipation considered is convection. However, the equations presented indicate that the largest diameter would give the highest temperatures (worst case). Subsequent analyses indicate that the smallest size is indeed the worst case. The assumption that a three foot section of conduit would be at the maximum temperature is inconsistent. Subsequent analyses indicate that the maximum temperature is at the point in the middle of the flame region and temperatures die away rather rapidly as the distance from the flame increases. Several other assumption were considered questionable. For example, the analysis assumed that a one-inch conduit was the smallest size. Specification ES-100 indicates that 1/2 and 3/4 inch conduit were used at the site. A walkdown has not been performed to determine the smallest conduit routed one-inch from a redundant open cable tray. Attached for your information and use are copies of the documents noted above. If you require any further information please contact me or Bob Bizzak. Please let me know what your conclusions are regarding the above design observations. frL1.kM A JJJ.}jKlanda JJM/1s cc: T. G. Tyler (w/o attachments) R. J. Bizzak (w/o attachments) CPRT File I.b.3 (w/o attachments) CPRT File (w/o attachments)

z.b.3 e e.cc r Spd 7 7t/D M ,t. -.-s..~-,n x =v_, m f o cy-Yn c) T Q CADLC TRAY TIRC TESTS ... - + n n-y f. b.\\ 9 \\bY q. 2 g d-4 y 1N96 qQ L.'J. Klamerus, 5432 N' M R. II. Wilson,12G2 4 Sandia Laboratories j Albuquerque, New Hoxico h

s P

d Ml l 1 P (SAND 77E1125C) 4 / k 1 1 f.i / '4 Abstract h( Funds were authorized by the Nuclear Regulatory commission h to provide data needed for confirmation of the suitability of current design ntandards and regulatory guidos for fire protection 1 and conttel in water reactor power plants. L This paper summarize: tne activi' ties of this program through March 1977. It describen a survey of industry in order. to determine current design practicer. The adequacy of cable tray spacing designated in Regulatory cuide l 1.75 was chosen for evaluation. I Using electrical cabic types i currently being selected for new nuclear power plant construction, a secconing tant was designed and completed to select two cable g f constructions which were used in subsequent full scale tests., 3even full scale toots were run and resulted in no functional Ij danago to cables in trays adjacent to that cable tray in which d; a firo was electrically initiated. Characterization of theco fires was mado and reveal a margin of safety in the separation q criteria of t.he regulatory guide for electrically initiated fires, in IEEC-383 qualified cable. ,4 ,9 b / R.. 5 .m. ~ aft f; :_~.,.,.% m. -- ~

/

Y ".::":: W.:::.=::.f (b A Y,h Fdar t $'di . 5 T W g.,~ E E y ,mmeuncu er rms wcuutur,s ununu N '(* ... w

e s ij ll. Introduction I g

• l The Of fice of tiuc) car Regulatory Research of,the United St.atcc I,

Nucicar Regulatory Commission is conducting con'fitritatory research in arean considered important to protecting the health and safety ,; { nf the public. Fire protection, as established by NunEC-0050, !l; " Recommendations Related to Browns Forty Fire," is one such I critical aros of research. /

t

/ The objectives of the Fire Protection Research Project at n Sandia Laboratorios are (1) to provide data cither to confirm the suitsbility of current design standards and regulatory guides for firo protection and control in light water reactor power ~ plants or to indicato areas where they should be updated; t (2) to obtain data that will provide improved technical basis either for modification of the standards and guides or for new I ctandards and guidos if necessary. Such changes are to be made where appropria'te to decrease the vulnerability of the plant to j fires to provide for botter control of ficost to mitigate the effcets of fires on plant safety systems: and to comove unnecessary j design costtictions (3) to obtain fire offects data for water r; coactor safety system equipment and to assess improved equipment, ,d .iesign concepts, and fire provention data and methods that can tj be used to reduce vulnerability of plant safety to fire. u j il. <4 Osckground 0 0 f{ When the project was initiated in July 1974, thn only task .g ansigned' was to provido the experimontal and analytical information 6 to evalusto the adoquacy of cable tray spacing designated in l Mogulatory cuido 1.75, " Physical Independence of Elcetrical Systomo, Section 5.14, conceal Plant Areas." This section of tho t i j .) j

l ~ ','r 2_ / ~ / guide covers separation of protective cystems in areas of the plant where power cablec are included and the only source of fuel is that provided by the cable materials. All evaluations were to involve the testing of equipment and configurations representative of those going into now nucicar power plant designs. It was decided that a survey of industry should be made to determine current design practicos. The cooperation by.mombers f' of the nuclear power industry was outstanding. Either personal visits or correspondence elicited responses from 13 leading architect-engineering firms, 13 utility companies, and 13 cable manufacturers. Three nuclear power plants were also' visited, although design practices of existing nuclear power plants were not included for evaluation. Information obtained during this /* survey has proven very valuable in determining cable constructions, cable tray constructions, cable loading, and types of cable assignments in cable trays. The survey also solicited information about previous incidents and experiences including the cable 'y tray fire at San Onofre 1 in 1968 and the subsequent investigation to determine the cause.2 \\ I A primary concern was to insure that the test facility truly represented the reactor plant area. The discussions with I architectural and engineering firms were particularly valuable { for improving the realism of the proposed tests. i N Since we had been warned of the difficulties of electrically initiating a fire in power cable it was decided early in the pro 3cet to conduct the test with 12 AWG, the smal' est power Q' cable normally used in nuclearlpower plants in order to minimize (/), the amperage demands in the test setup. A preliminary heat transfer analysis was also 99tformed at that early date. A rough analysis 8 was all that was considered necessary to determine the approximate current required to raise cable innulation to a combustibio I temperature and to dotormine if the conductor temperaturo is at its molting point 11003*C) when the outside of the cablo insulation is at its comb tion temperature. The analysin showed that \\ l ) .'a /9r/ *F 1

/ currents in the range of 300-120 amperes would raise the cabic .{ insulation to its combustible temperature. This agrecq with subsequent testing. e i With the results of the survey and the preliminary analysis t as guidelines, a test facility was developed to perform full scale testing of cable fires of electrically initiated origin. E Although it was originally intended to test all known types of p cable currently specified and acceptable for use in nuclear power plant design and construction, the largo number of cable types coupled with budget limitations precluded such broad testing. Therefore, screening was indicated that would lead to selection [ for testing of two typical cable types that would be most likely of propagating.a fire and would present a conservative approach. Cable Screening Tests A survey of utility companies, architect-engineering firms, and cable manufactureres, ascertained their p.'ferences of insulation and jacket materials. The inquiries stipulated that the cable types must be those currently being inctalled in or would be included in the design of nuclear power plants. As a result of this constraint, all cable types suggested were capable of passing IEEE Standard 383-74.3 There were thirty-nine rep 1'ies from industry which cited 20 different cable types that were being considered for use in new construction. Screening was necessary to cut thic list to manageable size and allow full scale te' sting to proceed. The first' cut was made.on the basis of popularity. The leading types were crosslinked polyethyleno with or without some jacket material (3I percent), EPR with a Hypalon jacket (23 porcent), and EPR with a Hooprone jacket (19 percent). 1 '? ..7 .. ~ t / ~~ ~ n ..y. ~ V s

s. -4 Considerations of the cost of filling esb1c trays in a full scale test prompted a further screening test-to obtain two dif ferent cable types that were "most likely to prooagate a fire." The screening tchis.:re performed merely to rank the various cable types in some manner. The rel'ative differences,botwoon results were small thereby subjecting the conclusions to dispute, especially if proprietary interests were involved. When burn length differences . are measured in millimeters, as they were in one of the tests, it-in difficult to attach true significance to those differences. The relative ranking of the cable types was based on three ,/' different evaluations. They were chosen to complement other evaluations, not to duplicate them. The oxygen index test which has bcon dono on all of tt.c cable insulation types under con-sideration is a case in point. The three types reported here i., are a small scale electrically initiate'd cable insulation. fire test, Underwriter Laboratories FR-1 flame test,4 and a pytolyzer and thermal chromatograph test (messure of insulation outgassing { as a function of temperature). Electricallv Initiated Cable Insulation Fire Test To determine the amount of current needed to produce a flame, five small scale tests were performed on five different electrical cables. The c.7ble types were: 6 Cable il - Single conductor 112 AWG, 4,,5_ mil (1.14 mm) EPR, 30 mil (0.76 mm) flypalon jacket, 600 V. Cable 12 - Single conductor 112 AWG, 47 mil (1.19 mm) e chlorinated rubber (proprietary), 47 mil (1.19 mm) chlorinated polymer (proprietary) jacket, 600 v. Cable f3 - Single conductor fl2 AWG, 47 mil (1.19 mm) EPR, 15 mil (0.38 mm) Nooprono jacket, 600 V. .s /

e

-l 5-Cable 14 - Single conductor 812 AWG, 30 mil (0.76 mm) cross-linked PE, no jacket, 600 V (Suppliec B).

j P i Cable IS - Three conduc' tor 812'AWG, 30 mil ('0.76 mm) cross-silicon 9. ass tape, 65 mil (1.65 mm) linked PC, 1 crosslinked PE jacket, 600 V (Suppli,,er A). rigure 1 shows how the cables were arranged in a cable tray for each test. Current was increased in increments of 5 ampercs every 10 minutes u'ntil a flame was observed. Cable 81 flared at lia

Cable 12 flamed at 130 amps, Cable 13 flamed at

-12[ amps (whileincreasingto125), Cable 14at120 amps,and 1 i Cable 15 at 120 amps. The spread of currents measured and ) ,/h observationr. of flame extent (flaces e,xtinguished shortly af,ter the conductor open circuit,ed) make all resu'.ts appear ci,ose, but relative positions were assigned with the better cables teing the ones with the highest current for flaming to occur. g.: FR-1 Flame Test Underwriter Laboratories FR-1 Flame Test was chosen as another screening test. It was not intended to be used as a pass-fall test (for which the test was devised) but to establish a rank based on length of burn and burn damage. It was expected that all cables tested would pass this test, and they did. In order to fail, s s s the paper flag 10 inches (254 mm) above the flame impact point must l burn. See Figure 2. ,7/ ,i-The test was conducted in a three-sided metal enclosure a under an exhaust hood. The metal enclosure was 12 inches (305 mm) wide, 14 inches (356 mm) deep, 24 inches (610 mm) high, and the top and front were opsh.- An 18-inch (457 mm) specimen cut from a sample length of cach cable was secured with its longitudinal axis vertical in the. center of the enclosure. Figure 2 shows the test configuration. ~ r,. f* \\ /

I ..'s.= t,. \\ J' e

• -j l

l o )) n l l l I i l o i t i l l l n l l l g I la 1 l l l 1 l l l c i r t c l l l l l l l l \\ e l E I 4I ro -b f n / o N i ta \\ r % g u % :s g i f n o 1 l i l 1 l l l C 3 e l b b, l 1 l l l 1 l I a C '\\ l l 1 / I I l 1 1 er u I l l 1 I l l 1 g (( F 1 1 / t s

• \\

/[ V / ~. .A =_ 4 l. /./- l'. / f s \\

f !

. !ltfs1 I }. : Iil.6* l.

u ( 4 4 Eneiosure 3 \\. 1. t a f I Lont;itucmo' exis of specimen [ Specimen [ _ 34.- Krof t poper (19 m ms Vertical atene parallel I30 9 to 'Jdes of enclosu.*e [ and cor.toining longitudinal .* i i oars of specimen ond y/2 ~ [ longitucanol omis of torrel B (O mm), f 1 s 10 i 1a- /q.T(254 h h l (3S mm) .h Longitudinot [)3 't exas of berrel / I i i 18 J I I (457 t< 4..

• s i

i nun) 1 I i l s i / si ,',1 l l 3" { Pione of tfw tip l j / i Min-I i a, g/ e I (76 of the accret Pilot LA 'N h \\ 9 Y2 i h) Max. c (241 ,i g Dorrel mm) l f. l Wedge l u I O: Front 'N f t i / e 7 ,/ s__. i i Loterol ~ Cotton / 1 .~. i Figure 2. Eccen':ial Dimonnions of Apparatun and s 3 "7 Specimen of Vertical Flamo Test l, 4 ,l u ~ t ' ~~' _,. I e

I, A Tirrell gas burr.cr (which differs from a Bunson burner in that the air flos as well as the flow of gas is adjustabl ) 5 supplied the flame. e The barrel of the burner extended 4 inches (102 mm) above the air inlets and its insido diamet'or was 3/8 i, inch (9.5 mm). While the barrel was vertical, the overall height of the flame was adjusted to 5 inches (127 mm). The blue inner core was 1-1/2 inches.(38 mm) high and the teraperature at its tip was approximately 815 *C (1500 'F). A wedge was secured to the base of the burner to provide I a sloping surface of 20 degrees from the vertical. This wedge was positioned to place the point A 1-1/2 inches (38 mm) from the point B, Pigure 2. Point B is the point at which the tip of the blue inner core touched the center of the front of the specimen A '..alf-inch (13 mm) wide strip of kraf t paper was attached around the specimen with its lower edge 10 inches (254 mm) above 'B and with the paper protruding 3/4 inch (19 mm) to provide a flag. See Figure 3. s l The test procedure was to apply flame to point D for 15 seconds, turn it off for 15 seconds, on again to point B for 15 seconds, etc., for a total of five 15-cecond applications of the gas flame to the specimen with 15 seconds between applications In no case was the specimen flaming from the previous application of tire flame when the 15 second "off" period had ended. The duration of flaming of those specimens af ter each removal of the i gas flame never exceeded five seconds. After the cable specimens cooled, burn lengths were measured beginning at point B. i l f ,1. !\\ !\\ \\ 'l \\ R &_. i '* s. ..s \\ /_ J n

_ =. =.. .n -9 m. e f b; r .i v- .v g .,.r ,e. .~ q y ~ m. ..4.. g i ac Y h* W s 'I . f.... e, ., =. - .g : ; ..t L ..r.. r ' - '.( -,,...... .t ~ .f -. y g .g. '. '. g.

• . *./... '.=... :,s..M....

., i. t. 6

• s.., %..

...c,. ~ s.,.. z s *.* .. =,y., ~ *../, :.../ * -.. r =. 8 =i 9 ,: a..

p.-

.t g W .i 8 u..- u.;..,.,... e :....*.s, a ..*. 1l, i

• e t* ',...

,,1, j o ..t 7., _...._ 3'. .).,.

4.,

'ml ., f, ? : g y [ t p *, 4. hl

• d...

l t (.s s. i g " gr ?, d st t. - .. a 8 a p a, r!, ; f t.;g;* f.. .)..,';,:i..' f M 4 c

. p.

• ].

.g...

..s

'.r?'

i O }...-* ;s. ', - l '. s e . s .

• tj n,

) }"* *= *. '. 6, a,P '..'. .g] .h ., '. (.... 'o. ( t e.. ". ; <r ; %....a

... a

..i... J W l 3 .i.. r - ~ >,

y.;..;.."..

. y, - =. s

• r 1::. *...[. t =

e. *i..;

.l = y, k . l + y.t,..,.....,. s.. ,,.s..'. * ...i... ~9 , :.p

v. ;. '.r..

..Q" '..).Ca.g.: r...f.,;..J;;,..,R. n e. .e .g

• s -..*

a.. e [.. a,n....? ~ ?.. '. 4 e ,~.,.8.,, ..>y .::...- ) ' ", gd ,t.. .. + ~, :. .q.. eq.. s;( ,.... t..1 r..,n...g *.:. ]. r.,i e :..s y, ..<.......a 18 .....c .s .T

a. e p s.

g. ?...

.. ". 's. s ...p s. ... - { ~ ~ * '..:.~. }l:'.f.j. s. .s.3.. )..:.* *.2 f ~.:m. '..: 4:. . '..V.

• q :. 'io

>. i'. j :f '.,'.. l..~,:e.? : .:. 1. a n . ' l.., M s- 'i . w.* ::..?.

• ;~;

.sf.4..v. :.. ;. g,.'.' s ~ ...r.c;

h.,.:q.* ::..r ' v',:. ;n.*.

• s..;..t :t i. /.

1 ..,:.;.y. r..-;*s y~.. e.,. ~< ' T.: ..W...,,. . '. t,. .... c..r.. c .w ., -.f ;-:'.:..s3.. .,:.:.. :.. g. 9 . g. s. -8 \\ s i 1

e

...g......._. . c.. ..*e. y . :q '.. (. :..... .,g.f.*n. ; ~ ).p.

9

..s. -

.S- ,,. ei,..f.* " .'.. '.r, g-l..;s...' i c ~.. <. :.. e,:.... ;. W.

• a.

9 .J e. -. \\.r..,.. J G s .1 n .l' 3; ...t. 'ir... g...,s, g. 'a.. w:n. .l.., *.;., '. q . e.". h,,., - e. %.,~g.'. l / - *.,,,'.v.7,,.- n . +

,,,h '..?"',....*S. *, s.s..p.f.*

.,,s. . f. .e,*,, ,,, ;,,,8 ., ; E.,. ' = A.'r.. 7,. ,,.g, ,,<s,- o i , =: -..'... ; ' 2.' g. a..

f' g.:a....o. r..c.:.

.. s '.:. ....r. r.* ,' s .!,5 ~. e s .r ., g,. s.,'; rJ '1.. i' t.j .. g,.*at..

.. y a

. ;g, C r.y.

• f. ;~.,.........,

n. . d..: 7.,.,..m .g .. f.

• l

..+,...,;, r , e,,'."se.' .,g. a.....e *. gr.,. i =..s....* s,,.9 fr @t e. g

s. 8 =, s,.s :.

.~e.. is..

• la.,;n

~

e. ;*.

i.;. - 3

• y:f.a a

....;.; c.* .~ e s,...a;"e.[ ,4.;g. .v e "pr .f;,f, .. t..;.,*,,,., s .'. ~. V l ; '.'. ,.'E,A',... ..;^.*'~.,...., s,.,5 ' '.ri

; o

• f.+ 1
• s

(. l e. y...',.::s. e.d'..g****. ?.%. s,. T.,,.L.:y./{' iC. n ;r : l** 6* l

c. s*

'. ~g ?t *. {.. '.&s *.gt. ,.p- \\ *-

• .. J. *. *. 1
• 1

. '. :..* *T;* ..;'l;.f.. d ?'..,.e~" ];,.., a a , '..;. tg.G &.. n ~'.d. ~. I e" ,,.....f...,..,,f,.i..,.. 4H ,,o ','. n j .....t..... .i .:. _.. :. a. \\1 . g. ..a i k,,..h e,..d.,:.26.,t"..I,s.m., [h' 'd.E....h. ';. 4-l 'h.IN;h.3/ T 1 1,. . a,. n . 5 ^.n....:..... +,y: e,, ~

, g.;..;..

,.....,,.j, .....,. e. 1 ,... n,.,. ..: A... ..a.... r w.r - -.: n : - [m,.o,jn $:% W %n,$ ? + ~ M @ u $. M..;.'.l.i l 8 re.$.:Mr.=2-M;;449 M.6.?.M : 10:.,: 'u. W .G \\ &s.b 53M = ~ .k.t.w..s.m.n..:.w.w5s. x.rt.:.; l FieJuro 3. FR-1 Test l l l . i g N .\\ s. s s. .4 \\. j 8

==e .-= .+ -

4. i Eight cable were used as test pecimens.' k1 Cable il - Single condactor #12 AWG, 45 mil (1.*14 mm) 1-:Pn. 30 mil (0.76 mm) Ilypalon jacket,.600 V. I l ~ Cabic #2 - Three conductor 112 AWG, 15 mil (0.38 mm) EPR, GO mil (1.52 mm) flypalon jacket, 500 v. i I Cable 83 - Single conductor $12 AWG, 47 mil (1.19 mm) chlorinated 'ubber (proprietary), 47 mil (1.19 mm) r chlorinated polymer (proprietary) jacket, 600 v. i i Cabic #4 - Single conductor #12 AWG, 47 mil (1.19 mm) chlorinated rubber (proprietary), 65 mil (1.65 n.m) j i chlorinated polymer (proorietary) jacket, 600 V. {< i i {j Cable 65 - Three conduer.or il2 AWG, 47 mil (1.19 mm) i chlorinated rubber (proprietary), 65 mil (1.65 mm) fJ chlorinated polymer (proprietary) jacket, 600 v. jI i1 l' Cable 16 - Single conductor #12 AWG, 47 mil (1.19 mm) EPR, t < 15 mil (0.38 mm) Neoprene jacket, 600 v. i 9

• i Cable 87 - Three conductor 812 AWG, 30 mil (0.76 mm) cronslinked i.

fI PE, silicon glass tape, 65 mil (1.55 mm) crosslinked PE jacket, 600 V (Supplier A). i 8-q f Cabic 10 - Single conductor 812 AWG, 30.311 (0.76 mm) crosslins:cd ?r, no jacket, 600 V (Oupplier B). lr l 'Eight cables were used l'n'the two screening tests requiring chort [ samples while five were used in the electrical test requiring [ longer sampics. If those three which had not coon all three i tests had been marginal performers additional lengths would have bocn obtained and given the electrical te,st. 4 _~ I^

• hM I

g ,4 Comparative results from UL FR-1 test were: Cable Type Burn Lenoth (mm) Comments

• _WT.'.

s 3,, jacket opened l ', '76.2

1. 2 4.5 jacket not opened

$3 '50.8 jacket opened 04 63.5 jacket opened f5 63.5 jacket not opened $6 61.0 jacket op.ened 87 69.9' jac!.et opened ~ 08 73.7 no jacket a Pyroliter and Thermal Chromatograoh Test .b. The last screening test used a pyrolizer on a thermal chromatograph a interfaced to a gas chromatograph / mass spectrometer. Thermodecom- .j position chromatographs were obtained as a function of temperatore !+j and the area under each curve was measured. Approximately 50 mg M of jacket material was used in each test and the temperature of W the specimen raised from ambient to 600 *C at 20 *C/ min. The material driven off below 300 *C was analyzed to test the hypothesis k that large a.nounts of material driven off at lower temoeratures was an, undesirable characteristic. Since outgassing of combustfblo j materials or fire retardants at these low temperatures was tiicorized kj as being undesirable, larger areas under the thermodocomposition chromatographs were assigned an ur.dosirablo rating. Figure 4 shows a typical chromatograph. h 6m M

t ~. a. s n., ~~*~*""I""~~' _y_,,, ._.._--..-~.---v------*- ~

• ~~~~ ~~

I [_,; _, - .e-..... f -~ -- = --

• l I

i I l 1.O_ _ l EPR-Hypalon j 5 4. 2 m g. l Q8_ 20*C / min. i a .e-6=> I e Q6_ 3 + c 1, .e - ac 'l " O.4 ~" g i t D s s I O I i = 0.2 - w I O.O, I i e i e e i .i e .i .e .i .s i s 100*C 2CO*C 3OO*C 400*C 500*C TEMPERATURE m a i Figure 4. Gas Chromatcgraph

l l .~ s g j The normall:cd areas on the chromatographs for the same cable types previously described in the UL PR-1 test are: l Cable Tvoe Normalized Area J Il s'- 1.2 s' 92 1.6 4 83 4.3 l 1 u< F' 94 4.6 .l', t i 85 1 - 1.8 P1 96 1.0 ? 97 i. 4.9 88 7.7 .a .?' s.'

5..

]_. Screening Test Conclusions

v.,

? Although the small scale electrically initiated cable insulations g .j fire test and the UL PR-1 Fire Test indicated none of these cables b would be capable of propagating a fire (in support of IECE 383 quali-t fictrion) cables 37 and 48 in the last two tests (come as cables 44 and 95 in the first test) were designated as the cable types to be I$ used in the full scale tests by a relative figure of merit. performe,d in Europe in 1975 5 Work on radiation and fire resistance of 7 cable-in'aulating materials was recently brought to out attention and v b is in good agreement with our ratings. .-)f x II id 1 0,. u / gp / ~. /

i l>- o Full Scale Testing Three phases of full scale testing have been completed. All involved electrically initiated fires in horizontally oriented E cable trays._ The first phase was intended to evaluate the adequacy lof cable tray spacing as designated in negulatory Guide 1.75, i e ,l " Physical Independence ot' Electrical Systems, Section 5.14, General [ Plant Areas." For this phase vertical separation of independent E divis' ion is designated as 5 feet (1.52 m) and the horizontal a } separation as 3 feet (0.91 m). The second phase was concerned with varying the separation h distance between cable trays. Phase three requir'ed a stacking ,,-l or matrix of fourteen cable trays as one division with cable trays p representing the second division separated by distances,as specified t r in Regulatory Guide 1.75. The vertical and horizontal separation . F~

• ', 's I in the first division was 10.5 inches (0.27 m) and 8 inches

'. ), 5 (0.20 m) while the separation between divisions was 5 feet (1.52 m) ' g.) and 3 feet (0.91 m). All testing involved equipment and cables , \\[ " representative of that goina into new nuclear power plant designs. See Figures'5, 6, and 7 depicting the three different test setups E for the three phases. l Coupons of aluminum, galvanized iron, and mild steel were hung in the building and periodically removed for corrosion analysis. A profilometer is used for this purpose and has not shown significant D I corrosion products. s. E 'n s An oxygen analyzer and gas sample manifold were installed an ' s, gas samples were taken before and during the fires. There was no depletion of oxygen found in the fire' area. Flamo retardant antimony bromide and an organophosphate were found in the gas samples as well as a high, molecular wax material. u O = E O g \\j%. ^- L j - ,t f., ,2, /. \\ / 1

n.,

-1s-h rr- --- - me v-x- - m- .m.-.-. n:: m c...:-.a :. f l'.- I.; il S. .i.

• j-

.?,i..^

i v

• l;.iI. * \\.'\\.1 '.

.W - c +,. ;,P. .O ...,. ~.. -s.. .. s.,... . j s,i, e \\ . i ~, 'Q,s. \\. .l: o 8 I ~. r .' -l 'l '.. *.

• l
• p
  • I Q,,l

I.

I'q .....,.. } {: 4 .o ..:t .,s...,,.. .s, .~ l.r: ...I .... =.. .s -. - f \\...,. T '. '...' :, L.;. .,'t: : I:.,.: Y. .I ':). .n.. sh '.t' N.~- :- '.'.:,5:..

. ),.. Q ',%, ;t

\\. ,o, l l l;I j. ./,.; f h. *. . ~......<,.., ,f.). '../ ,e,. t. t,. e . :. 8)s. '. -.g

*,1

.....,f l..

s.. t ,1 g,.. .,; y, t. s.i ; ,j~ .,I.

r..,i 2s..

g e t g c.. Q. u .. g... ~... w..u). * .n \\,. ,1 .e a

. w.;

,. u....; ;.r..c:...v:,g

n. {../..,

s z. .. v .p

w.,.i. e.. :

s. .t. ,...y

. l

.w. .. s. :,.7...

n)

> a....i. - l

s.. I M.i.; ? r.t J

e #, y. .,.;,.. v... r g

. e.,

p' c.; m:.i t. 4 ;.A :.,,.,a, .. ic v ',.y ,t .p.. .r / 1 ' w ::;. n ,.b.:l1.z n'l.':. h.v.1:. :j$. '. .;, ji.qi..y.. ?.'.'* : '... M. . L'.~... ,,~.'J.. 1 ~- s .n .... l9., ';l l.. 3 *b,,r.?:';:,;] i, l . L: Y. t . / 7 K'l,:! s. 2,, M,,.. 1 :'~. P,y;. ,. ",j,1.F. ; p.5 'l .g,h' g..., .L .t. /2 3, t.i. *.n! it s. %.. :e,e4+5?.,.:;.v.::- p. n'.G;...,,. h. i.,:.. n.

i

,J: v ';:;:>. 5-~., 'p i...., n. .s1 ,./.a.. ;. /

==

C..

... s. g.. A.. f.. f, <: *.. / r. ..= ". ~ . a. :... ?;,..,... ..,. :.....~.. e..x.. s. :c. '. %. A @_s_..D...:.y..a. a:: ~ .c m. , s. : j. '.

.:T.

R...:. '. ~.. . ?. C.. - * ^ ..g.,... . - 3....t y : E'.*. a.: .i. .') p e.\\. . v.,..c,',k=:'. ft %.2 i: 7

. A.. '.,<...

.-.,-....'.',0.s.4. E ~..,c: :.'+w.c...w%.y >.n.7 c\\. s. ::. -t-m ;&;..... '.::s y fi u,#lil u:.; L.,i.a,,q .,i >v . /.. / ' :.h ?:,.. h< \\ i .....:..M.. e . ~ .._u

n.. r,.. y i..

1. g.. a ::. u'

.t..

~. )

. :9 p."'. .ge i.,f.) p'g., "Jur==.-* '. m '.. : ".,s.w.., d'y .r. .r.: n g (*.m==, 1 h.:li?.0F 'ar#.g * '_ /.;.. I;g. *,.. ;.

• .1 e

... ' T,3.,P. l J s. ,.;l. ,.8.,%. *.,g .l % ?-. 2-. ,/- .=..*g:u r. a u t. - g s a.s r. :y.:t,.:,. -. g i.. e

r. f f. ? -

/ j (' i:.* 1.-.s '.:ny.f.y. "... t ?. m. :. - ./ 4 6 / .i!, '.. -lI ;.C. :1. _. 4;.': ..... \\ :.. ',./, . %.*.7,' ' el ;.. e.g L : } t .,.d. l a .l. y.,t,\\, ,. ~. s --a ? .. g f..a... 1,. :. lf.:, l-w p.g .f.. . /,, .r,.--

t

. ';.. y j,., 7- .,. i t. ..+...v.. s.,.. .M.. w..:n 4.y3 96. ...,a .{ ..#:i.. // ;,.s. .-:...e..*..... g

n..

s...,,.

n. ...:.... :.:.= ...,:-.,.. i,:. 7"..,, ',. i......i.'...

....\\..

" > P. f',. y.. .:..^z... 4 ?. '.i 1 c,. ..u,>... ......,..,.....,c. e,. t. ' /<. 1;.y. .. o.c.... n.. v... f f.. ..... s ;,.g.- ...w.. 1.. .o ..:.n.... , ~.........s., e. :.1 .y.. :.:.:.:. v.ni.:. :. - y. .....,..9... .,.., c. t '.. i. ..s. ..,. is. . v., .. s. ; e,. ......, r. 1..... }- ..N.. . q,/.., .... -,.. l.. .... '.,.. '. O . ;..= ~.y.j,;..' .< ~; .p.. . s.. .u . \\, ~...

.s ss.*

"...o..t",,f>#.... ,...... r, I.. w .g , i. .. '.-...,1 1 . g. ,t... '.. :; : : ~ o g - ...... ~ * ;>..v. /. /.

'.e s.

.f > ! : 1....,'., . ?;.. s, 1 ... s. m... . e...

g..

.p, -r. .;c*.... J. 2. . ' \\.

, j,.g

.4 c

l. M,..

s >, .,../.1.. g s.. .[

g.. -,,*,i.,.,

Q.:. , y.*, l s

.. e a

g-g . t. 61 " ; 9... *

• y * :,3.;....= k f.l..

J,/.

• ' " l.,-,.",

a L :.... y. !,.. i I.

• b.-

t

• A e

. * ?, g. 1 ..58.... l. ; ;,...,..... sk.. ;... ....,;.. e. 4 ig i... ..i.. .. /.f.,.s.,s.. s

s. ;.. { s

.w

t...H.,...,.w_,.

. c.. Ml',. , ;.t .._ ~,...i: c......>.:.ts.'*... 1 .1.u. -s v. )n_s s pe> -.w w .e e g I.m*--*.* -*-M.8 W-*******

y \\ s. i f i -16 s r!vi.- r-I;I ,i n;;.g.: . > p r.. ' y. t.

f. :

.h 'Iln. 9 l'.' ' ' ~ ' ::- ;6,[' :d f::y.1.,:p. :..*.. ~' ? I:

1..% :.

)-

-) : ' ' : ;<t v.)' :p.. :!I l.

t.

=:: .. ?- .........y i .. e, /... ]a... s .t . :.i!j ,.* i *? ....'. <.......*. :.,.4, ,,l...j ?.;.,. .t ..**:n-,, .,. ;. l.,. ... tr !y s' .,::.:[..;..h....i.i.,;,p'.

j ; ;'!v Y /....,..

p.c;. ~ ~

. s.,1: y. 2j l.. e

:l.l..;j. i,7l l. a.p! y' /...

'1 'r... ; 7:... ~. ^.:.. . 9...,.f CO>.:... . u.. 1.~ ,. y ' : .. i.

!.yL n.....:a

... <.:L...: '..vM y ' y g.. y j.:,;: . l. . ?"r.' ::'i.t..E,.. W. H n.'r'. .* /.".<l.v!' - :s rv .~.i .. f. t.

c. m...,....y.

.r.::.:. :,i / >.. n o.n$,,:,. . ;. };; i p, - l,l,.c %.. '/.:.;L'_. 4,/..y;, .:.? ~. :

: n.;

a ..-... I. '.it ;/.:.,. l!..:q7 ',..y., :/,'y:.x :;m@1 s..&.: lJ....:. ~r,s pt. ,u b}-.., = .f; k..)4.. ;.-. I:., l,.,. . 4.; M,.. ' '.: p !.1.I'., l g ; k.., u,_ qa @ b,.i.y. .. y( 9 2 9 +) S y 3.. .,d. n..x, y.; q . n..~ :.i /2.q. ,,p. . ;;l ...m;;h q.._.,>y,.4, p[ :.g.,:.l}N' :..; ',. .'f,. Q...,' 'y.: .m. 4 :'

qa yi.:.

n

..~

$ u':.r[Yl'.?g .,g s s. o..... [ UNM ..;, 7M 24 I k,* 1 ,. g:y{ Mgh.;p,.o.wr(%.'%Q^>g% o. , 'f' W* Nr.- m (. ~. e e-i i ,? .k }U $,k',Y . ' j. ' f.{0*y :' .e~E {'.>.$2.* .' p 'k}$ If*; . l l .W M*ang'. 42.. I'.". S ' .Y.j l }[ t 1l1,j,,f;;,g;ew.g.iff'%},,7Mwa= 1(' ' 4 {' g' ( {.j'i. Rsf.ffClQ,Q,j. l 1 '-?~ ggf".'

• 4

. 7 /1 8'..

s.g.

n y. c t,1,*:,Q,,flW y 2,e;'e>f::fE'M:-:+

fW:'.19MMT&@p%::

r;.:;n

).

Ilt..::rt.y ~,.

. y

,:m! iqi& -f. ll.NllIp[;j 'l. I.../.A \\1&W 3.'\\'?'it h,I : TN f/h ';fif ,)% s'ii.1 ?'<yf/.;g..//.*..*f/er,.g1#.,.,.R. Wg\\l.yg,

• {*

4 e., I

m,.

}. j '. *. 4..,,o >9. :, y: :,',"e ..r

q. p.. y:. m : ~.1. (.

.. - s l. I r J. - 4

1. 1

.'sh,......j.. s.;. 'is.: e .f ! ?a

.,),..

I 5 i 'j .,.,;P,.~. ;,: ':. :h.A1. ;g.!.:. . c.< ;,I.%,,;..': h?. t.t .. : y..,,),,,. . w.y :. - # t4..<..s s.<r'.,.. :.4;,c. l .s . t, g r<o. .s ~ ?, s s... :r e ',. t.* :* y.;7.:h':y:. s -.. s.: m,- .,s.',.7.".' s..z u/ 0.L1,i.&'r s,<1..t..;.Y.

b,./.

g :. :

n;,..;; o .:;;;'a*.... WM'_ :.U.@$. U.-QsLW. 1, ' _,. ".,~.;.*.:.:.y % ',f.. ?w...( : 4. C;L.N Wi k

1 i

'1:- 5. ;..:.,.,.: - ^.. yli. t ;. A. o t ,..c .. : i. n... e / -- s- '5-9.i.;P. '.i '. .yh IfYC: ~ g p. .p- .e s.

• • ..=*.oge.mm,,,

---7.-----

s t ,s .. s io l $7-3 E'

• h l

4,;.,,,ss. 3.r. s. u... ~ 1 e,. >... u:s; r..%.>. s-rg>. g,, Q.e... ~ u.. .,e 7 9.', e.;;r/..'. 4 .x ha.y. !"s. .d i;..i.....n 4r..y sN.g.... .. M, C. ./ v. f;; b.. G Q9.*;:?:g..,. p.,.., f 2,.

j..

,. 7% ,t0,' '.3. * }G' e, KqT. 9 f&:. e.p. j:n..-[.y.. !).i ? f: y', e,.%m 58 s

e.:a.:v 4..' #, s.., 4' v;.;j.fr r ;

- -.3.3 .,g.p:.&'I;. e ..-Ar-.,,'u O f.('%,,, - ; .t v. %:6. -. g.1 t ..-c* ;. -~,Q,Q t . VQ /.7:. :. A;:.,%. %m.a T' /s.hr . :.:d.,tn 4,.) QID. % 4: ',.;...n(. w .p- ? o,,r....:.. ~ . m,k',:::n.<-.r.. Q.t.f: i .....;~.. m e. 4.. .~ ....m ' * *9.'.;.;C:-: &.y t '. : Q.9

er m...

Q}..}..',y&.&n[l%,. L,R 'N W I'* q. z,m..d &.:.F/':.;/I'.r.,.,.&. ,. L:pu.m: l.f.,.g ::: j..,.... 'm:o. : >n '/.-$ ,.-;::3 . '...,, -r.s q%, t: %,;~v.. .n,gs. 1. .v. - r.%. $:s:a. w@:qp, # '%. s.' <.. e'N '-'$ : v. , % :. 3 M:.q? - c n... /h:. r C. ::. dd,I;-$'""a Lo,, ~.;;k.f.;rga. ..,,. "L..n . =',;w:C ~ 4 . %..; N :p .r.. - l',al,I '., u.*D;-m., t~;.,. k- )h +1~.: ;;'n~xch. "~w.::f.Q<..%*

• s

's ;t. .:~r., y..:.... .....k. + " '/.. /* ' h.. M M.-[ % ~,.,g.t g a 2 4 *f e t -=%f . M's *..r-R M".Nd * ".i ..c - ' + rl ?.=: Mi' '.' ' 7 N:

.s-Q. s G,..:T-i 2.-

t=:r,P. . <? s c,:,;.;;44 :. &: <>:\\ - 0 V s.:

s q' w.

...M-..YW..l 7 ..., f,.; M.. T.~g'.:t.'d.4 ,.' L~p >.$, e.'.m.; '. ;...'.'.W. m.'m%.,. ..bg.. CL:: d G g.f. p l. r.,. - ~::.c. :. s i t .::' c4 ::.p _ ep s. n 4 t p,t ~: 1 1, s

re:.gn::c N:.; =::===>~g~~.?qN

. ~:.5.-?%u.

W... JM 2 M'- G(-i Y N hJD2iBu% ~ $4hh..M$r ?"

y dk-h, k ' f.

i

1. p?,.h -

~ ~ ~ ~ ~

• [ ',ec;is gi. %.j M

p,.

r

• p.,.- $. -

t;- ,j l,. N.,,' h, [ YU 3' )., %g . 5'.g* !'.y, '/ - v.: p h.7 k 9:.--f.J;;7"$;..'.iNPuc: 1 3,. ~ 7 M y,,;e.g.WW..,m.;.s p,. . i. ' 3,.,.z.,.".,,u, g,p .y- .gyg.,.

6g g.

v: ,2. j Y

), f j

:y..pQ &,31.4%U

.4g&:s.. ,,:y V,;7.tr' .i W y: 't ';*.1.a. iq~t ,t.. v s ,n, e.e,- -A .:r .... sJ.m... m.=. =

• a W h. _.,,.

=,. .:; ;.1 <.. ' i;n,Q:.;', 'N'.&. o..,.*;*, 2 ':: e. I .. :h ) .r.. .,9. :.*. .c'. ] m..Q t..,...

t.M:.w.-;( 7:.4.P.. s..

' : %:..#.T.'Fd D__ .,g. jap'. '.y'8.;. *- sartrM. jr4-- ds 1;;,...'., r' f. $ ' -Q,.;I.,. ?. - 4 I r.. .. a. ,g r.L. : w. &.9.r.Q.,: . r% 4 .. 7M... a. ........:,, 4;;. ;.-l',, r.. ....:, a.%,-....*@u r.in.n..s,e 4 p .... :--.c ,. 4 tr ...:.... :. i .y:: w. w.w. ......y- .. m. ~,. el ae r. g.l 1 .1.... N..:g y. i,f 3 5.m -.,cy.; : ' I :+.. aa ...s . -) 4 u a. 3 <..'.s. e * :.. s...... 'M'$. m%.."L.l?$. ,t,g ..4

.u.

,,<w, m..re. 1 ..~. .W m l a N.; G 2 % m ...!.. <.,7J;(7 1 ar

N);G.M 4 : &.. e ' p. f' ml n

n.++.w.,&n....::,:&+:R.: eg$: >. A. i w:. ....v .y -....,:....m.

f...r Tv c..

.... w. t n.. . a..y,. . w... . c. t,,. c.. s. -.,s

• s

~3,3.~""I;g-' ?.d**/.E '. M, .Q( 3k" ~S*fdDi I,t:$$.M... p':.-.M : '..%.

T'"A..

...., r.*. 1 ~,L, - 1, 1,4 ;. l .. ~.. Figure 7. Phase Threc Test Setup ,,e . N. 1 >s.

'\\

\\. s /

,/ i 13 1 i nemote controlled cameras were installed for closed circuit television, color movies, photographic

• thermometry, and infrared I

thermography. Television was used to monitor the testing and l i,. In determining the proper time to at. tempt gas igni. tion (explosive bridgewires and electric matches were spaced over t'ho ignition o i point and simulated arcing), to take gas samples, and to start movie cameras. The movies not only provided a record of the event but gave information on the ignition mechanism as well as measurement of flame velocity. Despite a lack of success in i_gniting the gases with simula*ted arcing th'e movies show the combustible gases do indeed ignite as the flame producing mechanism. Measurement of flame velocity war needed so that the convective heat cransfer coefficient could be calculated. The photographic thermometry and infrared thermography were to supplement the discrete spatial measurements taken with thermo- ' couples and slog calorimeters. i On each test a minimum of 31 thermocouples and slug calorimeters were placed in the test setup and connected to recorders. Results of these measurements are discussed in the following section on the characterization of the fires. Air velocity was varied somewhat during the tests because of conflicting opinions on worst case conditions. Ooinionc varied between zero flow, which might be encountered in a cable spreading roon, to high air velocity providing abundant oxygen, which might be encountered near an exhaust fan".in the open plant area. An a compromise, air velocities for the different tests ranged between 2 ft/ min (0.01 m/sec) and 30 ft/ min (0.15 m/sec). These measure- ~ ~~ l ments were made with a hot wire ar,emometer before e'ach testr only fan exhaust 'relocities were nonitored during the test. Seven full scale 'tists were run in the three phases previously deceribed. Spacing was reduced in ph0aa two to 10.5 inches (0.27 m) In all s'ven tests vertically and 8 inches (0.20 m) horizontally. a l t 1 i .r

• ~ -

3

.v I 's all circuits other than the ignition tray circuits remained. a functional. This was determined by operation of these cf rtuits for some period of time after the test. In addition, samnlen nF the cable insulation at the bottom of the tray over the fire zone were given insulation elongation measuremenis to determine mechani-cal change. These measurements showed less than a 101 increase in elongation due to the fire. Quite often this small increase is attributed to a small chance in crosslinking due to heat. Characterization of Cable Tray Fires Characterization of the cable tray fires is based upon a review of the data that were collected in the full scale testing described above. The sources of data include: 1. Color Movies 2. Radiation Thermometry I 3. Slug Calorimeters and Thermocouples 4. Thernovision (infrared detection) This information is used to investigate the following ~ characteristics of the fire: 1. Size and Duration .~ 2. Flame Temperature - 3. Gas velocity 4. Optical Thickness tapparent emissivity)

• '/

consideration is also given to the thermal response of simple cylindrical objects which are engulfed by the fire. Approximate calculations provide estimates for: 2 4 1. Convective and Radiat'ive Heat Transfer 2., Equilibrium (Steady-State) Surface Temperature _in no a'ttemot to use the data t_o evaluate the likelihood There of fire anreadina_to an overlying trav. hecause this requires consideration of the geometric arrange, ment of the exposed cables and the kineti.cs of decomposition. 6 s ?.. I s 2 -~~ __ ~_.._-

~ 1 *. ( g-g

s..

It is emphasized that the measurements ard analysis techniques are approximate in nature, and are intended only to provide an overview of the gross characteristics of the fire. Within this framework, thd data are found to be self-consistent and in reaso::able agreement with theoretical expectations and corp,arative data. Color Movies Observation and analysis of the 16 and 100 frames /second / motion pictures of the cable tray fire tests have proved enlightening in characterizing cable fires. (Figure 8 is an illustrative sequence shot at 16 frames /second.) For example, the following observations tend t'o characterize the pictured fires. ~ (1) The flame zone does not comprice a continuous line fire, . /. but instead consists of one or more "axisymmetric" luminous I zones which are on the order of'S to 8 inches in " diameter" ,i^ at the base. \\ (2) Although migration along the tray may occur, the propagation f .= i is quite slow. M ,/l (3) The height of the luminous zone varies rapidly, ranging '/ from 5 to 10 inches _above the burning tray. I l (4) The time scale for variations of the luminous zone extent l [ is on the order of 1/10 second. I (5) I The flame is turbulent with luminous eddies clearly visible. i (6) By tracking the upward progress of small luminous eddies t which are shed from the flame, the gas velocity (time-mean) is estimated to be in the range from 3 to 4 feet /second (0.9-1.22 r/s) Variations from this range are quite small, ,,( even over a large number of measurements in different cable tray fire tests. Also it does not appear that velocity is decreasing substantially in the ver,tical direction, at least in the first foot of rise. These characteristics of the cable fires do not vary greatly from one fire to the next, even though significant variations in the duration are observed. .i l 'l 6 e j Og g ,/ -. ~....... .s.

1. .N r i l ....t......,,,, .?,*. .. s. e,... .s. 1 s

st

,..'js s ll t,. '.* ' ~,. '~. a f, . i.. s. ', ;. 1N..it , e V..lV. .1-j t f..: / *.

• r, a.

o i,,. i ..gs 1 ..g. ...b... f ..,s... .i.,... w.. ...o., .e,.. .s.. ~ .g, .v .s.. . ?f * ' ':' ' Y.. ~~.<... s *..s = 'l< .'. / ~ v.... .g- .......r

• .,.t..

,.*-s, A., .l.]r '. e~, ;:.,!p;.'.' ~f1,5:'.$d *.%".* l..v~.. l ;p * !. e.-.- ;c-,. . ~ - - ;.:~..- e s. v.~.*r.

r..s.. ;..

F ~i., , ' O. a.. s. .1 ..../* ..r g ,.I e.- ,,,;,a s. 5.. ..... ~, s.s 3.,, ;

.=. r 1.

e C k

f.,

.es::*, :e,x'o s. '..y- ... n..-.c '.r.- = . p f t 33 '..* *'t. p. r * # %.:.. s .g.s... f,. '.. i .. -t. ..-*t J,,,i f,.... - e*,. % *t.5.g.,v. ss . el e g i .. 4,..,;.s; t.*. *...

• *,'., :.., J

.s k{ se..o . s.;.. .lg)(o.V = s*j. s *. R e. -.#. r ' Y. ' .. ' t. g.

• W'f =. :...

. *. v,.,.. t J.. r. j..

• * *%,s

... V;*. *;, 4 },

• i.
• l. '.p**.., e, ?l' *... *

,,,'s / '. ~ *. i c, . #'t ..s ...,.. p,.I,.. e .h.., *54;.*.

• ! l 7.:.,. *.w
  • , ? 8',

6.,i...,* i,*g,' .t

• j=4,,..,..

. J ,.g. 8 i.'.',4.,Y. M M..& V,O, l i ' j...... $.,.{.., O /. *'..& Y*..'.,I gi f.. r Q:..'4*.g D.,.(t. g 's. $ ..A ,,. q u.J. W G l * '*1:.: 1 G &. eaj t . 'l Y 5."'5 w~~s...v . g'5' . fl f ?;,;.. ).. .a

f. ;g.. y, y.

[. .o i.s. ..t .: s .s. . ' *'~m I,

..~~.,

s t. o y... q .. * *;V a j s... . J... ..i, W i. i,.,.s.,. *=, i ,,y e,. .,.g L .s .. 4 ,.i.:.* ' n'.~. y. *.. c. ?.o g.. --s. ..s ...,._ _.. :... u.u _ _ _.,.s v. ,13 _y a I 2. g 1 [k Figurt 8. Cabic Trcy Firc (1/16 Second betwccn framos) h h1

f. a

\\. g

1 ./ - Plame Temperatures, Padiatior. thermometry is used te determine the temperature distribution in the fire. At chosen times, photograp'hs are taken through two different, narrow band filters 0,iA .03u) = which are centered er A =.55p and A =.65p. *The negatives are scanned with a microdonsitometer to determine the expcsure dis-tribution. The intensity of radiation received along a particular \\, line of sight is found by a comparison of the exposure at a j particular point (small area) on the negative with that produced by a calibrated lamp which is also in the field of view. The " brightness temperature" or corresponding blackbody temperature for each point is then calculated from the Planck function. ',) A typical plot of the isotherms (brightness temperature) I obtained from the radiation thermometry is included in Figure 9. All area.s enclosed by the isotherms are at temperatures above '1260*Khthelowercutoffonsensitivggyofthefilm. Makimum s 224 [ temperatures are roughly 1500*K. Figure.9 also shows the varia-

g. ~

p tion of temperature with horizontal position, taken as the hottest vertical location just above the tray (Section A-A in isotherm plot). I since the flame zone' is not optically thick, the apparent emissivity is less than unity and it is necessary to correct the temperature measurements. However, the magnitude of temperature corrections is relatively small. For example, a five-fold reduction in apparent monochromatic emissivity (rA = 1. 0 -+ 0. 2 ) only requires a.correctior. of about 100*K between the true temperature of the flame and the above brightness measurements. The measured flame temperatures are well below adiabatic flame tem; ecature, and are in agreem' nt with theoretical exp'ectations.8 e l i-Thermocouoles and Calorimeters f b i The array of thermocouples and copper slug calorimeters s above the lgnition t, ray provid.s two types of information: (a) heat fluxes (combined convection and r,adiation) that are determined from the transient temperature i I response of the calorimeters; .i l ,A. A ' I h. 1 l /. j \\ .s ~~ ~ ~ ~

'\\ lr., i i e i i e i 4 i. Cable Troy 6, g en O ,e e e t>A e e n s \\) gy, - 6 -2 v O Isotherm f ' e,p 'e D 8' Contourc 0 3 s .9 C s 0.m 6 p Ne f J f f 9 h x. Caeremmae e o

\\

a i . e.. Temperatura g Profile g j Along Tray f I s im.. at Height A/A [ ] in Above j r s isse.. - m m k w l s gg. f f Itseenemme reseman e 9 j s J

T -24_ I, lL (b) steady state temperature which may be significantly ( less than the local gas temperature due to radiation through tbc flame Fi_gure 10 chows the tempetature renpense of selected t: 1 calorimeters (Noc. 1, 7, 9, and 11).and a sheathed thermocouple a (No. 2) for e.he fire test of 5 October 1976. The' separation between cable trays is approxicately two feet. This particular i fire is one of the most intence and longest duration of those !5 '-- studied. It is sece that the intensity of the thermal environment __ (d { 3 f_ alls off very rapidly in the region from 5 to 11 inches (.13 to .28 m) above the fire. This' height roughly corrc pe;.ds to the

s the upper edge of the luminous zone.

1 In view of their relatively slow time resronse, the calori-W meters and even the thermocouple rarely reach a qua:isteady temperature level. However, in the fire test of 5 October 1976, (r thermocouple No. 2 reaches and holds ll50*F for a short period R H at early and at late times, and in the intervening period the temperature is clearly steady at 700*f. These quasisteady j temperatures are confirmed by similar data from calorimeter No. I which is also located about 3/8 inch (9.5 mm) above the burning w tray. It is noted that t' hose temperatures do not represent f local gas temperatures, but rather the temperature of a surface [ immersed in the flame. ? ,a Figure 11 shows the variation of cold wall heat flux with 9 height above the burning tray for several fires. Each of these i data points is calculated free the initial slope of the tempera-ture vs. time curve for a particular calorimeter. It is seen that p a significant reduction in heating rate occurs from the base of the flame to the upper reach of the luminous zone. Although these U are significant variations in heat flux distribution from one fire te the next, the two more intense fires (October 5 and November 15) are very similar, as are three lesser fires (July ( 21, August,13, and December 16). It is likely that some of the dif ferences are due'ro unintentional changes in position of the instrumentation relative to the flame zone because the z' exact location of the flame could not be controlled, ri N l l Y b O ~ s.

• • • "M seno

-m- ...e..m.m...

~25-g i t, Q E ,,-l g i.. k v I I i i i I i i i 1 Y b 3 0o - h 'N 1000 / r,f., N s \\ C / t N

- soo

/ 1 /,o \\.~ [A \\ gJ e'f \\ 3 / n p' / -fs_ ./ i a0 s s f e's /,,, N s< - a 400 // g / \\ l = // d~ ***,,,, #, an.****'**m.,*%.,,\\

• • m

/ / / p'j,j'" ~g 200 ~~ b# I f I f f f f f f f .? Ij 0 10 80 120 1GO 200 240 280 320 360 400 p Etapsed Time (sec) h Ileight Above Burning Cabics @ Tc 3/8" @ CAL

• sa c

w ( @ CAL 3/s" @ CAL 11" O CAL *" g s \\ Figure 10. Temperature Responso of Thormocouples4 m and Slug Caloriar.aters (10/5/76) Fg 4 LL n ,( n- ,I l / I, d g W h s-N i G M e epamp. ge m e sm g ee a e .n=

• • .e i--.-

5 F' -2G-L 8 = ~. y- ? l t e i s ' 28,000 6 3 6 -l l l g e 101G11(, I d A 12/6/1C g 24.000 - I k B 11/15176 k S l 8 20,000 - M ag -08/13178 j o a a1/21/TG is,000 r f 8 m M g g [ En. e 12.000 3 l y e a 5 [ U A g 3 E 9 m 8,000 -A i c4 'l e4a A g i l A 4.000 - i 5 A a E. O 8 M O ? " U f I f f f f e I f I e I . -s O 2 4 G B to 12 14 IG 13 o ~ Distance Abave Troy Un.) ,/,, e l t g Figure 11. Vortical Variation of Iloat F3ux i I' ? i. 11 I.j l 8 i I t e e i I-f.

i. J I

Lt.J L i / 1

I .I N / M

4 1

Thermovision 4 An infrared detection system marketed under the tra'de name

]

"Thermovision" was used to monitor the cable tray fire tests. f The field of view is continuously scanned by a,mir,cor system, and for each point in the field the amplitude of the voltage H ~ signal from the detector is converted to a gray " color level" (intensity) which is displayed on a black and white monitor. A movie is made from the monitor to provide a qualitative overview of the development of the fic,e, and at later times particular 1 frames are extract,ed for quantitative analysis. Selected frames from the thermovision movie are scanned by l a microdensitometer to obtain a quantitative map of the degree I { of exposure. The exposure levels are then interpreted as i levels of IR radiation intensity using the calibration charts /. i.; - provided by the manufacturer. ip Since the broad band (thernovision) measurement of IR inten.sity C is fairly sensitive to the effective flame emissivity, this IR ./ h intensity can be used in conjunction with the previous estimates r) of flame temperature to calculate the flame emissivity. Based on 9 the procedure described by Sato and Matsumoto the total emissivity of the flame is found to be on the order of i = 0.15. When this 3 result is compared with the theoretical calculations of relske and I Tien10, it is concluded that particulate (soot) concentrations in ,[ flame are on the order of.10 cm /cm, whien falls with'in the expected ~0 3 3 range of concentration.11,12 p LiH Analysis of Fire Test Data d Heat transfer from the flame to an engulfed object occurs si by both convection and radiation. Although the calorimeters Provide' a measurem.ent of total heat flux, it is also of interest q to know the relative importance of convective and radiative ? contributions. The following paragraphs outline snee approximate IJ calculations which answer this question and at the same time e show that all of the measurements (flame temperature, total heat flux, velocity, infrared radiation, thermocouples) comprise u.3 2 k ,/,n. ,.,_: m i

a g /,' \\. -f i e i f, / / a reasonably scif-consistent characterization of the cable tray fires. At a location just slightly at,ove the burning tray we have the following measurements of C' lame temperature, total entiscivity, and" flame velocity: Tg = 1300*K, i = 0.15, V = 3 ft/sec. Using this velocity and properties'of air,' the mean convective heat transfer coefficient for a small cylindrical object (e.g., 3/8" 13 2 colorimeter)'is approximately h = 7 BTU /hr/ft j.P. The ' convective and radiative contributions to the cold wall heat flux can then be separately calculated as follows: p(ET g - T,, = 13,000 BTU /hr/ft, q'- -Tf,'i=7,000GTU/hr/ft 2 / q" = to Tg / / This shows that convection accounts for about 67't of the total flux. Note that the total heat flux (convection and radiation)- is in good agreement with the calorimeter data shown previously - in Figure 11. In view of the above calculations, it is useful to reconsider the vertical variations of cold-wall heat flux.hown in Figure 11. It is seen that the heat flux is roughly 13,000 DTU/hr/ft2 (tta nominal convection rate) at a heiglit of 5 to 7 inches (0.13 '.18 m) above the tray. From the color movies, this level also corresponds to the time-mean height of the lumi eus zone. c is therefore ,t expected that' convection dominaten above this level. In the j due 'to entrainment of cool air and turbulent a :ing. At a height upper nonluminous region the gas temperature fallr. off r apidly \\. of,10 inches (0.25 m) above the fire the cold-wall heat flux \\t 2 is only about 6,000 BTU /hr/ft, which corresponds to a local s. ~ gas temperature of 1000'r (900'K), assuming convection alone l \\ and a velocity of 3 ft/sec (0.91 m/sec). 'Since the flame is optically thin, a 9vlindrical object placed in the fire (thermocouple, calorimeter, cable) will, if the fire continues long enough, reach an equilibrium temp-orature which is well below the temperature of the surrounding 7 s / .s p p"

"1 t' ! medium. This teady-state surface temperature T can be .i s estimated.from the folacwing energy balance in which heating of the aarface by convection and thdiation is equated with the cooling afforded by radtation from the surface which passes g through the flame to the cool surroundings at T,: 1 E(T-T,)+ fo{T - Tf) = (1 - ()e (Tf - T[) 4 'i At a point near the tray, T 1300*K,F = 0.1, and h = 7. These 4 !f values give a steady surf ace temperature of about 1100'r (870*K), ~ in good agreement with the quasisteady tempera.ture recorded by thermocouple No. 2 in Figure 10. ' Note that calorimeter No. 1 also approached this temperature befers t:e fire began to die out. l It is interesting also to calculate the equilibrium surface temperature at a height of 10 inches (0.25 m) above the tray. Daned on the measurement of cold-wall heat flux the local gas I temperature was estimated as 1000*F, assuming convection alone. 1 Using the steady energy balance with T = 1000*F, thw equilibrium surface temperature at the 10 inch (0.25 m) level is approximately 1 650*F. The above est an.ites.of equilibrium sur f ace temperature are indicstive of the n rady state surface temperature of a single electrical, cable w t.:n is subjected to fire. In an overlying tray, cables are c osely spaced and the details of the geometric configuration bocene imoertant. Thus, higher surface temperatures probably are attair ele because radiant losses from the expoced bl'c are blocked uv adiacent cables and convective velocities i may be higher than an the single cable configuration. On the other hand, tae duratic. of the fire may not be sufficient to scalize equilibrium corm ttions, as was usually observed with thermocouples and sluo calorimeters in the test fires. In any case, the temperature of exposed cables ca_nnot exceed the 4 temperature of the surrounding medium which is estimated as ji roughly 1000'r.at o. height of 10 inches (0.25 m). .I 4 s.

l

.s y!: ( .e ~ 2 m n..

h ./ r_..,, l, II h Summary of Characterization S Essential features of the cable tray fires.are outlined f below. Although based on worst case conditions, thesc observa-l tions are generally representative of the entire sequence of fire tests. i (1) The intense period of the fire persists at a / ~ particular location for between 40 and 240 seconds before die-out begins to secur (e.g., 240 seconds in l Figure 10). 1 (2) The luminous flame zones fluctuate rapidly between 4 and 10 inches (0.1-0.25 m) in height. I (3).Cas temperature in the luminous zone is roughly 1900*F (1300'K). r ] (4) Gas temperature at 1_0 inches _'(0.25 m) above the burning tray is esti ated as 1000*F. (5) Velocity of rising gasses is approximately. 3 to 4 l feet /second (0.91-1.22 m/sec). I (6) The luminous zone is optically thin with an apparent emissivity on the order of E = 0.1. j (7) Heat transfer t'o immersed objects is convection j dominated with radiation accounting for no more than 30% of the total heat flux, even in the luminous region. (8) Equilibrium surface temperature of enquired cylindrical l ~"^j objects varies from about 1200*F just above the tray l to 650*F at a height of 10 inches (0.25 m). Although the above measurements and analytical estimates are 3 i approximate, they are indicative of the gross characteristics of the fire.- i It is noted that the present cable tray fires differ greatly i I f' rom large fires which are often considered in safety studies. Due to the small physical dimensions of the present flame, radiation fron'the flame is less than 20 percent (e < 0.2) [ j l of that encountered in large fires, and convecti.on therefore l dominates. In large fires convection usually accounts I less \\ . ~ ~

= ' ;.:',, ~ ~ ...,.q. ~,, ~~ .a. 14 than 25 percent of the total heat transfer Also, objects immersed in a large fire will eventually reach temperature equilibrium with the flames..This may not occur in the optically thin cable tray fires because an er$ gulfed sur f ace is able to radiate through the flame to the cool surroundings. Thus, the cable tray fires comprise a considerably less severe { ' thermal environment than a large fire, even though.the flame j temperatures are of comparable, magnitude for the two cases. Summary and Conclusions The first objective was to obtain data through experiments to aid in evaluating the effectiveness of cable tray separation as a means of assuring functional integrity of redundant safety systems. The first task undertaken to meet this objective was to survey the industry in order to determine current design practices particularly with regard to the materials used. Of these materials primary interest was focused on types of electrical cable constructions being used in new nuclear power plant design. A screening test was applied to these types in order to concentrate on two electrical cable con-structions representing a conservative approach. The evaluation .h covered separation of protective syst' ens in areas of the plant where { power cables are included and no, source of fuel exists except that -~~'- provided by the cable materials. Thus, all fir _es in this oroiect have been electrically ini'. lated, i l 21 22 l Seven quick-look reporta and a progress report have [ been issued describing full scale tests included in the period l covered by this paper. Separation distances between cable trays of 5 feet (1.52 m) vertically and 3 feet (0.91 m) horizontally 'were used in phase one tests. Four tests were run in phase two I with spacing reduced in stages to 10.5 inches (0.27 m) vertically a 4... i \\ i j l s /y - .\\ %~- I

and 8 inches (0.20 m) horizontally. Phase three involved three tests of a large matrix of trays arranged in such a manner that 14 cable trays closely spaced represented one division while 3 trays separated 5 feet (1.52 in) vert'ically and 2 f'est (0.91 m) from that matrix represented the redundant division. In all these i tests an overcurrent in one or two 12 AWG conductors of an electrical

• cable in an open cable tray was the source of fire.

Trays were filled with electrical cable to the top of the 4 inch (0.10 m) siderails. i = Fire initiation appears to be from combustible gas initiation as seen in pictures taken during that time period. Typical of this initiation is the sequence taken during initiation of a fire on November 15, 1976. This is,shown by Figure 12 where the gaseous ignition appears beyond a photometric calibration lamp. t The maximum duration of any fire obtained was 29 minutes with _ the mean time, approx.imately 6 minutes At no time did the cables 4 in trays displaced from the ignition tray begin to burn. All circuits in these trays remained functional and elongation measure-ments taken of insulation closest to the fire showed no major (<.10%) change. ~ N

t o s', _33 UT. O.T.F. s '.v.. " Y.,..* '*? D..'.=...s 7a m..~:.r..., .n 5- . : a.- t: -.N* U ,,. a '!.y.,1 ,,.l..

.n -

~. ! 6 **.r.".. . %$.1 b i

• ...*e

. q.q....

c. . p=.s: ,r..........,.;.... s.. n.

., m, v. -l.

..s :n: = .:. t......,.,.~n.,...., o,.w a 8 ;,.... ...,a..... .y W:.,. J \\ z. r .) g, p ......'. :4 ,,~s.' +N.a.,W!;. g.-. .%'4..s a +.j/:: p \\ %. ;~..: ?........ P: y. N...\\ ... Q. 4:.- 4 h:sj'.,.C.a)..'j*...O: r....tr. :.:~w%., ' $,2P/' *T.u..' %*. ..Y*. W 'U ~ s . i.t.:. : '...W.B. in "- 'i ~'Q - .:m.,.. ".9 ., m.,;n.., -.a. . <.,. r .x.,.

y..... s

+ ..~ m.m. ~,..g .....r

r.,..

y. ~...C.Q. .. ;y.C::

d

.. :. 2 2,.. *s.5.:... i .> a :<.....: y.. P <:.. '.. . (..W*...:...,,,,. ~* *... ;: 4c;.4 .c , ~:,.;.... s;. :* :.,p. 7 ..*..s,, r '.. . if.'n'.

• _. ' i. *.* C'

. SY.

• 1 i?. '. d i'.,%,, N,( :.* i ",
• 1 g..

1.

.a.

, '...f.h-l;,. - - -- . -- a,. NG.(b:.:t:l,5*:..... 3..,'t...**:y*iO;r - ;g 1.. '.+; e*u.;'s

i. r:: *a.n..K. &,.4 k W. ::.e.g maa-c1.E'. s..9::

s :; v

• Y-et,

9.i.'.. ?.ii s/t.%

.'....~.

..

). ~ - r.- PMh ^ .%Q>.....Q]lQ .:. N..s i .a , ~r th e.g,," w% = (.4 a %.;. '.: *. M 1.~;.M 0 a. F ': 3.;......<. d t *,f g p.W?.D. ~ S., g$ Y.. ' N e. '. i E'w t p. ~? N* ?' ??I:"5{$':h"Y-l .g. 1 b.g - 6 - $.*Y 'l?i'O-. m M's. u*lI. ,~*,,.h.

7.. ':. 1 k :N'Y'; ~.

C t..y.., .W..;y'*. '. -. 4.%r.<.: . A. ~.

n:. : :.

':e..n..-.... 1 ,s.. e ..s. > :.\\w.* ~...., ; *. ~* . 'l. ~n....-. s., l, g,'*:;.,*g ; ;m*.* ;..

• di.#.,t 3.g l. gr e.

r a e ~.: . a.: ~...:., M * * * , *"m*.*",**' l a.i.. ; a :., :. i s- ..ie.. g $..~,.... ;. s..W d;: 5 ir. .G 'i:E y 6. ]>;l0%:2:, _ _.;,:#WY'.V..:&..i'.*:. ':$:::.1..,: ...M :,., _._.~ ;nl. :.s '. ".t

G:

- V..? : ;, v w c. ...x -2, " 3 2. :'n.,w m w c. m.x...,.. n;s.. t

. T wl;.

.. ' t :r.,.-.p. ;y< 3,j... ..o u g. fu. x.;; r;2 s. %]., c y. y.. =:m %.. *

.,d. :. i '..Q', m.4 f,q

..g. $k'y..'{;'.; SW

cdyt A

?. O :sd.:;;;.W ;%.:.$

l'.

i <.n . c. m..:.,r v.. ~.;, 7,pg.a.,.a.,., y].,,,W nh.Q~.% . w%.h WT;., . a, u - n..,..;.. . :w w:.2:w %. t,vs...

.e..

-- s ... v.,~, ..a.... .t >..c.~,.,. ..q. s 4,

g.,,................y,.

u..

..h,~:-: c 2 i, .. g ;. ...,..... e..v.. 4. c.r..a...,;. m.,n.:. e..w.. .n,.s,, w m. ..,.:r-:..:c .. m.n. n > n.. r. ~. . w. ; a...c... ,.... a, t......,:.......... . w *T m g, g_d_ a.. -_ - ).h...% L 1 ._ s t_m _A_ gjg g,, 2 g J. Figurc 12. Ignition Mechaninm (1/16 cecond betwcon framen) -r- = = * " " " " " " .. s,. a g

,/ '\\ ' l [ 6 s ?

• e 34 6

t-REFERENCES l

l. '

1. " Recommendations Related to Drowns Ferry Fire," HUREG 0050, February 1976, NRC Staff.

2. ' Report on Cable Failures-1968 at San Onofre Nuclear Generating

's Station, Unit I, Southern California Edison Company. 3. IEEE Standard for Type Test of Class IE Electric Cables, Field Splices, and Connections for Nuclear Power Genersting Stations Std 383-1974. 4. James Gaffney, "The Significance of the New FR-1 Flame Test," Wire Journal,' October 1973. ~~ 5. H. Schonbacher and M. H. Van deVoorde, " Radiation and Fire ,f Resistance of Cable-Insulating Materials Used in Accelerator Engineering," CERN European Organization for Nuclear Research, 15 April 1975. 6. K. Annamalai and P. Durbetaki, " Ignition of Thin Forour. Pyrolysing Solids Under Normally Impinging Flames," Combustion and Flame, Vol. 27, p. 253-266, 1976. .. r _-. 7. H. J. Kostkowski and G. W. Burns, " Thermocouple and Radiation Thermometry above 900'K," Measurement Techniques in Heat Transfer, E. R. G. Eckert and R. J. Goldstein, eds., Carca Publications, N. Y., 1970. 8. D. durgess and M. Hertzberg, " Radiation from Pool Flames," Heat' Transfer in Flames, N. H. Afgan and J. M. Beess, eds, Scripta, Wash., D. C. 1974. 9. T. Sato and R. Matsumoto, " Radiant. Heat Transfer from Luminous Flames," Conf. on Int. Dev. in Heat Transfer, Part IV, p. 804-811, ASME, N. Y. 1961. 10. J. D. Felske and C. L. Tien, " Calculation of the Emissivity of Luminous Flames," Western States Section Combustion i Institute, Monterey, October 1972. j 11. M. W. Thring, J. M. Boer, and P. J. Foster, " Radiative Properties of Luminous Flames," Proc. Third Intl. Meat Transfer Conf., Vol. 5, p. 101-111, AICHE, Chicago, 1966. 12. R..Echigo, N. Nishiwaki, and M. Hirata, " Study on the Radiation of Luminous F1.ames," Eleventh Symp'. of Combustion, Combustion Institute, Pittsburg, p. 381-389, 1967. 13. W. M. Kays, Convective neat and Mass Transfer, McGraw-Hill, ,. ! N H. Y., 1966. ~~, 5 / 1

(,

• 14.

L. II. Russell and J. A. Canfield, " Experimental Measurements of flest Transfer to a Cylinder Imr.orsed in a Large Aviation-Puel Fire," ASMC Jout nal of Heat Tr ansf er, 95, S, p. 397-404, 1973. 15. L. J. Klamerus, " Quick Look Repor t on Fir ~e Protection Research," July 197G. 16. L. J. Klamerus, " Quick Look Repor t on Fire Protection Research," August 1976. 17. L. J. Klamerus, " Quick Look Report on Fire Protection Research," October 1976. 10. L. J. Klamords, " Quick Look Repor t on Fire Protection Research," November 1976. 19. L. J. Klemerus, " Quick Look Report on Fire Protection Research," December 1976. 20. L. J. Klamerus, " Quick Look Report on Fire Protection Research," j February 1977. 21. L. J. Klamerus, " Quick Look Report on Fire Protection Research," March 1977. i 22. L. J. Klamerus and R. H tiilson, "Progrecs Report on Fire Protection Research," SAND 77-0303, NUREG-0206, June 1977. f l I i i F i i \\ / 3 l L i l i E i

1. "

e 0 t II.b Concrete Compression Strength QUESTION: 1. Paragraph 2 on page 13 of ISAP II,b Results Report refers to errors in the Schmidt Hammer test program identified by third-party review, and refers to them as "not significant." Provide the basis for your concluding that the errors are not significant.

RESPONSE

The statement in the II.b Results Report regarding the " error rate" relates to our initial use of an incomplete population (concrete volume established by identifying truckloads poured during each of the periods under evaluation) and subsequently determining a more complete though not exactly complete population. In the Results Report we conclude that the " error rate" in not exactly determining the population size was not significant.. The initial evaluation of the hammer indication data (101 data points for the CAI and 99 for the CC) was conducted on an incomplete truckload population (see transcripts of TUGCO-NRC meeting of 3/6/85 and F. Webster, " Additional Background for TUGCO-NRC Meeting of 3/6/85," CPRT File II.b.4a-008, May,1985). The missing truckloads represented approximately 20-30 percent of the total number of truckloads. To complete the population determination, an attempt was made to identify all previously unidentified truckloads, and a proportional sample was selected from those additional truckloads identified. This augmented sample was then added to the original sample. The resulting evaluation of the hammer data (presented in the II.b Results Report) included 119 data points for the CAI and 132 for the CC. The added data did not change the conclusion that the CAI hammer indication is within 5 percent of the CC hammer indication at the tenth percentile level. If the population could have been completely determined an additional seven samples for the CAI population and two samples from the CC population would have been taken. If these additional samples were randomly selected from the remaining truckloads excluded, these test values should be dispersed among the other data (as was observed during the effort described above). Therefore, the distributions shown in Figure 3 of the Results Report would be changed very little and the conclusions not at all. QUESTION: 2. Review of Figure 1 of page 20 of ISAP II.b Results Report shows that CAI compression strength is approximately 9.4% less than CC compression strength at the 10th percentile level. It appears that this level of deviation was judged by applicants as not "significantly lower" than CC compression strength to trigger a need to implement calibration of the Schmidt Hammer test. Discuss the technical basis for the judgement. 5 NRC

I II.b Concrete Compression Strength (Cont'd)

RESPONSE

The design compressive strength of 4000 psi is 18.4 percent lower than the tenth percentile cylinder strength of the CC. If one assumes the CAI cylinder data is valid, then it is seen from Figure 1 of the Results Report that the CAI strength is only 9.3 percent lower than the CC at the tenth percentile level, and is well above the design strength of 4000 psi. However, the validity of the CAI cylinder data has been questioned and the CPRT investigation was established to determine whether or not the CAI strength is not more than 18.4 percent lower than the CC strength. This is done in the II.b Results Report through the use of the Schmidt Hammer tests in association with the CC cylinder data. A difference of 18.4 percent in compressive strength corresponds to a relative change in hammer indication of approximately 10 percent, based on the slope of the hammer indication vs. compressive strength curve (see Operating Instructions Concrete Test Hammer Types N and NR, copyright 1977, PROCEQ, Zurich, Switzerland; and Attachment A of F. Webster, " Target Tenth Percentile," CPRT File II.b.4a-003, February, 1985). The tenth percentile CAI hammer indication reported in the Results Report is only 2.5 percent lower than the CC hammer indication, and when we evaluate at a 95 percent confidence level the CAI hammer data is determined to be no more than 5 percent lower than the CC hammer data at the tenth percentile.. This is less than the 10 percent difference in hammer data that would be required to signal that the CAI tenth percentile compressive strength is at the 4000 psi level, or lower. This provided reasonable assurance that the CAI tenth percentile cylinder strength is well above the 4000 psi design level. Therefore, it is unnecessary to further refine the relationship between hammer indications and compressive strength in the present application. QUESTION: 3. The resolution to ISAP II,b as presented in the Results Report may not be able to identify localized problems where the number of falsified records is small. Discuss potential safety implications on overall adequacy of the concrete strength due to such localized problems. 6 NRC

), II.b Concrete Compression Strength (Cont'd)

RESPONSE

As discussed in the II.b Results Report, there are two general types of potential falsification. The first, and the focus of this discussion, is the masking of out-of-specification concrete by recording it to be within specification. The second, and of less concern, is the false recording of concrete test data for within specification concrete when tests were not performed. Neither of these two types of falsification appear to have occurred in any systematic way. There is a potential for not detecting specific examples of the first type where the number of falsified records is small; however, as discussed below, the engineering significance of such situations is limited. The methodology of our investigation was constructed such that if this type of localized falsification occurred, it would have been detected unless it had occurred very infrequently. Thus, our discussion is - focused on evaluation of the engineering significance of a very small volume of concrete that may potentially be out-of-specification. ACI Standard 214-77 addresses the implications of out-of-specification concrete as related to the ACI criterion permitting ten-percent of cylinder tests to fall below the design strength. Specifically, the following excerpt from Chapter 4 of this Standard is also applicable to evaluation of potentially lower strength concrete due to falsification: 6 7 NRC

1 II.b Concrete Compression Strength (Cont'd) "4.1--General The strength of control cylinders is generally the only tangible evidence of the quality of concrete used in constructing a structure. Because of the possible disparity between the strength of test cylinders and the load-carrying capacity of a structure it is unwise to place any reliance on inadequate strength data. The number of tests lower than the desired strength is more important in computing the load-carrying capacity of concrete structures than the average strength obtained. It is impractical, however, to specify a minimum strength since there is always the possibility of even lower strengths, even when control is good. It is also recognized that the cylinders may not accurately represent the. concrete in each portion of the structure. Factors of safety are-provided in design equations which allow for deviations from specified strengths without jeopardizing the safety of the structure. These have been evolved on the basis of construction practices, design procedures, and quality control techniques used by the construction industry. It should also be remembered that for a given mean strength, if a small percentage of the test results fall below the design strength, a corresponding large percentage of the test results will be greater than the design strength with an equally large probability of being located in a critical area. The consequences of a localized zone of low-strength concrete in a structure depend on many factors; included are the probability of early overload, the location and magnitude of the low-quality zone in the structural unit, the degree of reliance placed on strength in design, the initial cause of the low strength, and the consequences, economic and otherwise, of structural failure. The final criterion which allows for a certain probability of tests falling below f' used in design is a designer's decision based on his intimate knowledge of the conditions that are l likely to prevail. ' Building Code Requirements for Reinforced Concrete (ACI 318-71),' provides guidelines in this regard, as do other building codes and specifications. L To satisfy strength performance requirements expressed in this fashion the average strength of concrete must be in excess i of f', the design strength. The amount of excess strength l depe8ds on the expected variability of test results as expressed by a coefficient of variation or standard deviation, and on the allowable proportion of low tests." i l I l 8 NRC t-

a r n II.b Concrete Compression Strength (Cont'd) It should be noted that the criterion of allowing 10 percent or less of the cylinder strengths to fall below the design strength of 4000 psi is more than met by the CAI truckload population, which means that the frequency of potentially understrength concrete (regardless of whether it is masked by falsification or not) is very low. A supporting consideration is the fact that, with age, average concrete strength asymptotically increases above the 28 day strength on the order of 24% at one year (ref: A. M. Neville, " Properties of Concrete", J. Wiley, 1975, P.258-9) and continues to increase thereafter. Therefore, based upon the II.b results and general structural considerations, the chances of a potentially understrength concrete being coupled with a critical structural element are even lower. 6 9 NRC

f,) 6 l III.d Preoperational Testing i QUESTION: 1. Section 5.4.1 of the Results Report stated, in part, that System Test Engineers (STEs) "...did use current design documents in the conduct of preoperational and prerequisite testing activities." During an inspection of documentation related to the 60 preoperational test samples that were evaluated by the CPRT, the NRC inspector identified 26 preoperational tests that were performed where the STEs failed to update the revisions of design documents referenced in Section 3.0 of the test,,rocedures. The documentation clearly showed the CPRT's awareness of this discrepancy, but it was not identified in accordance with Appendix E of the Program Plan. The NRC inspector informed the CPRT that failure to identify the discrepancy was deviation from Program Plan commitments. The Results Report should have addressed this discrepancy. The staff needs to know what actions were taken to determine whether this was a DCC problem or an STE problem, what impact this had on the objectives of the ISAP, and what assurance exists that other tests of safety related components and systems, not evaluated under this ISAP, were condu'cted using current design documents.

RESPONSE

CP-SAP-21, " Conduct of Testing," contains the requirement for the review and update of test precedures. The administrative procedure was not explicit as to how the STE review and update should be documented. However, the SIE was required to update the test procedure to be in accordance with the latest design information, therefore, but was left to his own discretion as to the method of documenting the update. Close examination of the specific procedures revealed that they had, in fact, always been updated, but that sometimes the updates were recorded only in those sections of the test procedures containing the action statements (i.e., sections other than Section 3.0). The procedures had been updated by the Test Procedure Deviation form in accordance with CP-SAP-12, " Deviations to Test Instructions / Procedures." The CPRT third-party concluded that the absence of specific notations to the reference section (Section 3.0) of the test procedures was neither a deviation nor indicative of a DCC or an STE problem. In those cases where the reference section had been updated, it was easy for the RTL to verify that the STE review and update had been accomplished. In those cases where the reference section had not been updated, any design change would have to be verified as being implemented in the remaining sections of the procedure. In all cases, it was possible for the RTL to confirm that implementation had occurred. Each design change requiring a response by the Startup organization was, in fact, incorporated into the test procedure. 10 NRC

V III.d Preoperational Testing (Cont'd) Based on the foregoing, the objectives of the action plan were met and there is reasonable assurance that the document control problems which existed prior to 1984 did not adversely affect the testing program. Reasonable assurance regarding the extrapolatability of sample observations derives from the facts (1) that there was a start-up administrative procedure which required such revisions and (2) that in all sampled cases the procedure was followed with. QUESTION: 2. During the inspection of documentation related to the 60 preoperational test samples that were evaluated by the CPRT, the NRC inspector identified an unresolved issue regarding twelve screening checklists that were not completely filled in. Three of the twelve checklists failed to show the CPRT's review to ensure the associated preoperational tests were conducted using current design documents. This issue must be resolved before the staff will be able to accept the Results Report.

RESPONSE

It is believed that the requisite data to demonstrate the adequacy of CPRT's review is available on 9 of the 12 checklists. The other three checklists were overlooked during the final file review. For these three, all the information required to perform the evaluations is contained in the various files, but the checklists are not completed properly. The project central file will be amended to correct this discrepancy. 11 NRC

e j' 1 VII.b.2 Valve Disassembly QUESTION:- 1. Section 4.1.2 of the Results Report states, "in addition to proper matching -of components, the procedures were reviewed for (sic) damage during the disassembly, storage and reassembly process." Please provide the results of this review.

RESPONSE

As discussed in the " Procedure Review" portion of Section 5.2 (page 14 through 16) of the Results Report, the procedures used for valve disassembly - CP-CPM-6.9 or CP-CPM-9.18 - have always contained provisions to package disassembled valve parts. The purpose of this packaging (in a heavy duty plastic bag or wooden box marked with the valve tag number) as stated in CP-CPM-9.18 is "to prevent loss or damage and to maintain traceability." This practice was found to be adequate to identify damaged parts.* Additionally, the operational travelers and QC checklist for valves (QCV's) reviewed during the sample reinspections all contain a sign off by the craftsmen, QC engineer, or in the vast majority of cases - both, verifying all internals have been cleaned / prepared for reassembly. This constitutes a final check for visible damage prior to reassembly.

See, e.g., action plan working file Section 5.0 (Item I-M-VALV-122).

QUESTION: 2.. Section 5.2 (page 12 of 20, last paragraph) addresses differences in non-ASME and ASME manufacturing processes for the bonnets. The Results Report states that physical and chemical properties identified in the material specification would be the same for both and also that post manufacturing testing would be the same. Please address how you considered the differences between ASME Code and commercial requirements such as material identification and traceability, welding and weld repairs, personnel qualifications, and nondestructive examinations.

RESPONSE

The conclusion as stated in the Results Report is that there is "no substantive effect of interchanging a ASME bonnet with a non-ASME bonnet on ITT Grinnell diaphragm valves." This conclusion was based on discussions with the manufacturer's QA Division Manager as documented in the action plan working file number 9.0 item 9.0-25 (copy attached). It was recognized that there are differences in the quality assurance programs under which the ASME and commercial grade bonnets are manufactured, but this was determined not to be significant in this particular instance since post manufacturing testing is identical for both ASME and non-ASME (commercial) bonnets. 2 12 NRC

e e. VII.b.2 Valve Disassembly (Cont'd) QUESTION: 3. It should be noted that NRC Inspection Report 50-445/85-14; 50-446/85-11 identified an unresolved item (Appendix E, paragraph 6.j) pertaining to the differences identified between the Westinghouse and Gibbs & Hill (G6H) Lines Designation Tables, and differences between G&H Tables and Code Data Sheets. Please provide the necessary information for resolution of this unresolved item (445/85-14-U-15). This question is in no way related to the conduct of ISAP VII.b.2.

RESPONSE

TNE is currentle performing an extensive line by line comparison between the C&H and Westinghouse Line Lists. Members of Gibbs & Hill's Design Engineering Department", Westinghouse's Design Engineering Department and TNE's Mechanical Engineering Department are involved in this review. The objective is to identify and reconcile all differences between the two lists and to determine the correct condition in each case. Site system flow diagrams and Westinghouse design flow diagrams are also being reviewed to insure that both are in agreement with one another and are consistent with both Line Lists. Following this review, TNE will compare the questionable Valve Code Data Sheets to their respective line number for final assurance that the valves are acceptable for their applicable conditions. QUESTION: 4. On page 1 in second paragraph under Section 3.0 reference is made to a valve testing program (a) Identify the program and/or programs and clearly indicate the scope i.e., how many and what type of valves are included, what types of valves are excluded, etc. (b) the loss or damage of valve parts is a QA programmatic concern when it'.s repetitive and uncontrolled, even if its documented. Explain how this issue is addressed in your implementation process. Section 4.1.2 the third paragraph addresses an evaluation of the adequacy of present procedures. Was there a sampling inspection of valves (and documentation) installed under the present procedures? What are present procedures as opposed to past procedures?

RESPONSE

The system test engineer is required (by CP-SAP-20) to walkdown each system. The valves in the system are inspected (Section 4.4) for proper flow direction, accessibility, bolt tightness, stem travel, operability (smoothness, etc.), packing, etc. This is required for all valves in the system; safety-related as well as BOP. 13 NRC

o-VII.b.2 Valve Disassembly (Cont'd)

Additionally, some valves are checked / tested for operability over and beyond those in CP-SAP-20, such as: All Motor Operated Valves are tested in accordance with XCP-EE10 All Air Operated Valves are tested in accordance with XCP-EE11 The Main Steam Isolation Valves are tested in accordance with 1/2 CP-PT-3401 All the Steam Generator Relief Valves 1/2 CP-PT-3402 All valves used for containment isolation are local leak test .(10CFR50 Appendix J) to 1/2 CP-PT-7501 The RCS Boundary Check Valves are tested to 1-CP-PT-5709_ and 2-CP-PT-5706 In the sample of 106 valves, there was one instance of a lost valve bonnet and one instance of damage sustained to a bonnet requiring replacement. Both had been properly ~ documented by TUGC0 on Nonconformance Reports (NCRs). As less than one percent of the sample items indicated a loss of valve parts and less than one percent damage, the Issue Coordinator and Review Team Leader do not consider this to be a programmatic concern of repetitive and uncontrolled loss or damage. Had this condition been determined to be a programmatic concern, the action plan would have been expanded or corrective action would have been recommended to the Project. As stated in Section 3.0 of the Results Report, the action plan focused on the undocumented interchanging of parts. It should be noted that the ISAP, as it pertained to damage, was only concerned with damage sustained during valve part storage as per the allegation. Other cases of valve damage were found in the sample items. This damage had nothing to do with the valve disassembly / reassembly process. Our review revealed that repair had been accomplished satisfactorily. The sample included valves which had been disassembled and reassembled under past or " earlier" procedures, valves which had been disassembled and reassembled under "present" procedures, and several valves which were disassembled more t!.an once and so were dis / reassembled under both past and present procedures. Present procedures as used in the Results Report means CP-CPM-9.18 issued in mid-1983. Early procedures were*those used prior to that date. Section 5.2 of the Results Report discusses the details of both procedures. QUESTION: 5. Section 4.1.3 second paragraph states in part an evaluation was made to define potential code violations. - What are they? They should be identified. 14 NRC

VII.b.2 Valve Disassembly (Cont'd)

RESPONSE

The evaluation for potential code class violations mentioned in the second paragraph of Section 4.1.3 was done as part of the analysis discussed in the first paragraph of this section. This analysis is contained in the action plan working file as document no. 6.0, item 6B-6 (copy attached). Revision I dated 11/25/85 of the analysis was inadvertently omitted from the action plan file and has now been added. QUESTION: 6. Section 4.1.4 first sentence states that reinspection of valves which were disassembled was performed to provide assurance that the valves were reassembled using the correct components. It is not clear how, or from what documentation, the correct components were identified.

RESPONSE

The acceptance criteria are stated in Section 4.6 of the Results Report. t QUESTION: 7.- Section 4.2 procedures are not identified per program plan attachment 3 ISAP format.

RESPONSE

The procedures in effect are CP-CPM-9.18 Rev. O, dated 6/8/83 and QI-QAP-11.1-26 Rev. 18, dated 12/19/85. QUESTION: 8. Section 4.6 appears to apply to only diaphragm valves - what was the basis acceptance of other types of valves with interchangeable top works and trim.

RESPONSE

The criteria of Section 4.6 applied to all valves inspected under this action plan. QUESTION: 9. Section 5.1 second paragraph states that the review installation procedures, revisions and dates should be identified. 15 NRC

VII.b.2 Valve Disassembly (Cont'd)

RESPONSE

This information can be found in action plan working file 6A, items 6A-1 and 6A-2 (copies attached). QUESTION: 10. Section 5.0 page 11 first paragraph states that a lost bonnet and a damaged bonnet were not deviations because they were properly identified on NCRs and PETS. The valve type, size, gag numbers, date of installation, the NCR and PET numbers should also state if the NPV-1 form was revised, or annotated.

RESPONSE

The NCR and/or PET associated with these valves, or any similar conditions, serve as the key to initiating any required code documentation relative to the repair or replacement. When NPV-1 certified parts of a component are replaced or repaired, an ASME Section XI NIS-2 form is executed to maintain component certification acceptability; this form is completed prior to N-3 certification of the Unit, and is utilized in lieu of annotating or revising an ASME Section III NPV-1 Data Report, which is not permitted by the Code. QUESTION: 11. Section 5.0 page 11 fourth paragraph states that two types of ITT Grinnell valves were supplied. This paragraph should also provide complete identification of the valve types (manufacturer's drawing or identification numbers), valve sizes, rating and applicable code class.

RESPONSE

This information can be found in action plan working file 6.0, item 6B-5 (copy attached). (Note that the Generic Safety Consequence Analysis attached to this item is superceded by Revision I which is provided in response to item no. 5 above.) QUESTION: 12. Section 5.0 fifth paragraph states in part: For some application...the applications should be identified (page 11).

RESPONSE

The applications of the valves rated 300 psi at 150* F. were those within the scope of the NSSS Vendor supply. Westinghouse always specifies this type valve regardless of the application, system or plant for which their NSSS is supplied, for reason of standardization. 16 NRC

VII.b.2 Valve Disassembly (Cont'd) The ITT Grinnell standard valve discussed in paragraph four of page 11 of the Results Report is used in all non-NSSS applications. QUESTION: 13. Section 5.0 page 12 first paragraph is not clear in its description of valve modifications. 1-were the modifications made specifically for CPSES valves at the specified 300 PSIG, or 2-are these valves just different configurations furnished by the supplier when the user specifies service conditions, pressure / temperature, that are higher than design.

RESPONSE

See response to question no. 12. I QUESTION: 14. Section 5.0 page 13 second paragra~ph, identifies two valves by tag numbers. This paragraph should further identify the manufacturer's drawing or identification number, size, rating, code class and date of installation. Additionally this paragraph should identify the documents (e.g., NCR, IR, PET) that substantiated acceptance of the installed valve body and bonnet.

RESPONSE

i The information requested is: Valve Tag No. 2-8422 Mfg. Dwg. No. - SD-C-100552 Rating 300 psig at 150*F. Class 2 Size - 3" Install. Traveler No. MW81-1105-4900 Reinspection Pkg. No. - I-M-VALV-44 dated 10/16/81 i Valvd Tag No. 2-7131B Mfg. Dwg. No. - SD-C-100551 Rating 300 psig at 150*F. Class 3 Size - 3" Install. Traveler No. MW7980361-4100 l Reinspection Pkg. No. - I-M-VALV-56 dated 10/23/79 This information is in the reinspection packages found in action plan working file Section 5.0. The acceptance of the installed valves is documented on the installation traveler. No NCR or PET was in effect documenting the deviation at the time of the CPRT inspection of the valve which is the reason the deviation was declared. 17 NRC

e n VII.b.2 Valve Disassembly (Cont'd) QUESTION: 15. Section 5.0 page 13 the second and third paragraphs, identify two valves by tag number. These paragraphs should also identify the manufacturer's drawing or identification number, size, rating and code class and date of installation.

RESPONSE

This information is: Valve Tag No. 1-7046 Mfg. Dwg. No. - SD-C-101609-Rating 300 psig at 150*F. Class 3 Size - 3" Install. Traveler No. MW80-1020-4900 Reinspection Pkg. No. - I-M-VALV-9 dated 11/11/81 Valve Tag No. XSF-179 Mfg. Dwg. No. - SD-C-105686 Rating 255 psig at 150*F. Class 3 Size - 3" Install. Traveler No. MW79-081-4700 Reinspection Pkg. No. - I-M-VALV-67 dated 12/19/79 This information is in the reinspection packages found in action plan working file section 5.0. QUESTION: 16. Section 5.0 page 14 first paragraph states that because the installed valves (with deviations) match the numbers recorded on the operations travelers, this means'that the bonnets were interchanged prior to issue for installation. The staff finds that this deduction may not be valid if the valve was disassembled, installed and reassembled on the same day. If the traveler records these operations as performed on the same date (same shif t), there is no assurance that the required information was recorded prior to disassembly. Another potential is the switching of valve tags.

RESPONSE

The installation of these valves, as documented on the installation traveler in the reinspection packages, showed that the valve bonnets were removed and stored for a period of months, and then reassembled when all welding was complete and the line was installed in the field. Switching of valve tags would not cause the noted deviations as numbers stamped on the valve body and bonnet were used for the reinspection. 18 NRC

1 ) VII.b.2 Valve Disassembly (Cont'd) QUESTION: 17. Section 5.0 page 14 second paragraph relates to travelers for the other two valves that were written prior to the practice of recording bonnet markings... s This paragraph should identify the two valves in question, the date installed, the procedure and applicable revision at the time of installation.

RESPONSE

The valves in question are valve tag nos. 2-7131B and XSF-179 i discussed on page 13. They were installed under procedure no. CP-CPM-6.9 Rev. O, dated 10/6/78 on 10/23/79 and 12/19/79 respectively, based on traveler completion dates. QUESTION: 18. Section 5.0 page 15 second paragraph refers to early procedures. The specific procedures, revisions and dates should be identified.

RESPONSE

CP-CPM-6.9 Rev. O was the project source procedure which contained integrated Construction / QC direction for the disassembly / reassembly of valves on CPSES. CP-CPM-6.9 was divided into subsections shortly after its issuance, and the requirements for valve disassembly / reassembly were then encompassed in CP-CPM-6.9E. CP-CPM-6.9/CP-CPM-6.9E Rev. 0 (2/6/80) set forth the following requirements with respect to valve disassembly / reassembly: Detailed instructions, including the general requirements of CP-CPM-6.9/CP-CPM-6.9E, would be provided to Construction /QC via an Operational Traveler (OT), prepared and apptoved in accordance with CP-CPM-6.3; and, Section 3.14 of CP-CPM-6.9/CP-CPM-6.9E requires, in part "All

• parts removed from the valve shall be stored in a heavy duty plastic bag, or in the case of a large valve a wooden or cardboard box. The MS [M111 wright Superintendent] shall mark the box / bag with the valve number.

"Any valve that will remain dismantled for an extended period of time will have the bag / box of parts stored in a secure place in the i Mi11 wright Shop or Warehouse. If the MS estimates that the valve will remain disassembled for only a short period or that it is too large to be easily removed from the work area, then the bag / box may remain in the. field." 19 NRC

1 9 VII.b.2 Valve Disassembly (Cont'd) The above requirements remained as written through DCN #5 to CP-CPM-6.9E Rev. 6 (8/1/83), at which time they were deleted and CP-CPM-9.18 (Rev. O, 6/8/83) was referenced. Additionally, Quality Instruction QI-QAP-11.1-39A Rev. O was issued on 6/8/83 to prescribe specific QC inspection and documentation requirements for valve disassembly / reassembly. Additional details can be found in action plan working file 7.0, items 7.0-1 and 7.0-2 (copies attached). QUESTION: 19. Section 5.0 page 15 third paragraph last sentence states; sufficient information for evaluating valve storage prior to this time is not available. The issue of concern was the storage of disassembled valve components. The TRT found that the storage at installation locations was poorly controlled. The paragraph should address the storage of disassembled valve components. Addit'ionally, this paragraph refers to an effective program implemented by M111 wrights. This " Effective Program" should be addressed in the aspect of the implementation of an identified procedure and the verification of training of millwright personnel in the applicable procedure.

RESPONSE

The Results Report does refer to the storage of valve parts. It was intended to relate that the M111 wrights had effectively implemented the existing program. See response to Question 18. Records for the training of Mi11 wright personnel are on file in the Construction Department Training Records. QUESTION: 20. Section 5.0 page 15 the fourth paragraph states that the issue related to documentation of the interchange of valve bonnets was recognized by TUGCO... This paragraph should state the basis (NCR's, irs, etc.) for TUGCO's recognition and address this subject by including the identification of the procedures, revisions and dates. 20 NRC

- ~. _ _ ~ 1 1 VII.b.2 Valve Disassembly (Cont'd)

RESPONSE

The RTL did not identify a specific event or discrete occurrence. The recognition was manifest by the recording of body and bonnet numbers on travelers which began in late 1980. This was a general practice within existing procedures. It was formally proceduralized by TUGC0 with the issuance of CP-CPM-9.18 Rev. O in June 1983. QUESTION: 21. Section 5.0 page 16 the second paragraph states that the QC checklist requires recording of the bonnet identification number. L For the installation of valves, since valve tags can also be interchanged. the staff finds that the procedure should require that the checklist should j record both the body and bonnet identification. - RESPONSE: i i As stated in the first paragraph on page 16 of the Results Report, the i checklist does require recording of both body and bonnet } identification numbers stamped on the valve parts. l I QUESTION: 22. Section 5.0 page 16 third paragraph states the administrative action was taken. (by TUGCO) in the startup test program. The administrative action should be identified in terms of identification of any applicable procedures, revisions and the CPRT verification of the training of personnel.

RESPONSE

0 The administrative action taken by TUGC0 in 1985 was to require control of all work processes during the construction phase of CPSES, through implementation of the work package conce;t defined in the CP-CFM-7.1 series of procedures. Verification of program impigmentation and the awareness of project personnel with the program was evident from the process in which CPRT was required to obtain project documentation, prepare inspection packages and initiate work i l processes. The only question of applicability during implementation of the CP-CPM-7.1 (series) involved the Start-up Organization, which, as documented in the action plan working file 7.0 item 7.0-4, was resolved by letter CPPA #45,538. 4 4 21 NRC

a. VII.b.2 Valve Disassembly (Cont'd) QUESTION:

23. Section 5.0 page 16 the fourth paragraph cites an example identified by the TRT as evidence of procedure implementation and effectiveness.

The TRT also identified (in SSER-11) numerous PETS that documented the interchange as replacements for lost and/or damaged valve components. The staff wishes to emphasize that the issue essentially was procedural inadequacy to control the interchange, loss and damage of disassembled valve components. The staff disagrees with the CPRTs reasoning that this is an example of procedure effectiveness. The TRT stated that although the deficiency was reported on the NCR, and procedures were in place, the loss and damage continued to occur.

RESPONSE

See the response to question 4. QUESTION: 24. Section 5.6 page 18 identification and discussion of corrective Action first paragraph is vague. The paragraph should identify the level of responsibility of the changed personnel and identify the procedure, revisions and dates as they apply to the subject of this paragraph.

RESPONSE

As addressed in response to questions 18 and 19, the corrective action was to implement effectively the existing program rather than developing a program to implement. Implementation was effected at the craftsman level and procedural compliance was and is stressed at the supervisory levels. QUESTION: 25. Section 5.7 page 19 Out-of-Scope observations. The paragraph refers in part tot acceptable TUCCO Procedures... The procedures should be identified.

RESPONSE

CP-QAP-12.4 Rev. 1, dated 12/28/83. 22 NRC

t VII.b.2 Valve Disassembly (Cont'd) QUESTION: 26. Section 6.0 page 20 the second paragraph states that procedures were reviewed and found to be adequate except for.. and further, the last sentence states that improvements to the control process since 1983... The procedures, revisions and dates should be identified, and the improvements to the control process should be specifically detailed in this paragraph.

RESPONSE

See response to question no. 18. QUESTION: 27. Section 7.0 page 20 does not clearly identify any of the results of the implementation of this plan (e.g., procedure inadequacy, lack of control, etc.) that must be addressed by TUGCO, and then evaluated under ISAP VII.a.2.

RESPONSE

TUGC0 must disposition the 4 identified via the Project NCR process. No programmatic concerns were identified during the conduct of this ISAP (See response to question 20). ISAP VII.a.2 will assess handling of any programmatic corrective actions by TUGCO. One of the specific allegations being investigated in ISAP VII.a.2 is the portion of the TRT issue on valve dis / reassembly (as stated in AQ-52 of SSER-11) that concerns " effective programmatic corrective action was not implemented... 23 NRC

4 4 Mlnat10n Search 0FFICE MEMORANDLH QA/QC-RT-076 TO: J. Hansel FROM: M. Solon DATE: April 8, 1985

SUBJECT:

Valve Dis.tssembly, Issue VII.b.2 Generic Valve Evaluation Summary Documentation (i.e. specifications, vendor instruction manuals and drawings) were reviewed to determine which generic valve types required disaasembly prior to welded installation into the piping systems. It is concluded that diaphragm valves, manufactured by ITT-Grinnell, are the only valves which required disassembly prior to weldup. Purchase orders CP-0020A, 00205, 0604 and 0001 (S.O.0220) contain nuclear safety related (Q) diaphragm valves with the potential for mismatching valve bodies and internals when the valves were reassembled. The number of valves in these purchase orders is apprcximately 600 total for Units 1, 2 and Common. Non-Q diaphragm valves contained in purchase orders CP-0021B.1, 0021D and 0604 are identical in form and fit to the Q valves, and will be considered as a source for rismatching internals and valve bodies. Discussion In accordance with the Action Plan, para. 4.1.1, an evaluation was made to determine the generic valve types that require disassembly and removal of internals prior to welding. Project specifications, drawings and vendor instruction manuals were reviewed. The latest specification index pages containing valves were marked up, and Table 1 was prepared to summarize the results of the documentation review. All Q valve types were reviewed first. For those valve types that ward found to require disassembly, similar non-Q valve types were evaluated as a possible source for mismatching non-Q internals with Q valve bodies. Valves supplied with vendor packaged equipment were not reviewed. Specific discussion of all valve types, by specification, follows. Referring to Table 1, Page 1: (1) The vendor instruction manuals for the diaphragm valves (MS-20A, 208) require that the bonnet assembly be removed to protect the diaphragm during weldup into the piping system. (2) The vendor instruction manuals for the bulk valve orders (MS-20A.1, 20.B.1, 20.8.2) do not require valve disassembly for welded installation into the piping system. C/ ~m' ' ' 4 - 800 Oak Ridgelhrnpike Suite 501 OakRidge. Tennessee 37830 (615)482 79 3 i-

i s j QA/QC-RT-076 Page 2 April 8, 1985 (3) The specification for the rubber lined check valves (MS-208.3) has only four 24 inch valves. These valves, which are in the service water system, are all valve type 24CC302WA, Notes 3, 39 and are identical. Therefore, there is no need for further evaluation of potential mismatch. (4) Butterfly / wafer disc valves use bolted installation exclusively. Referrina to Table 1. Pane 2: (1) The non-Q diaphragm (MS-213.1, 21D) require valve disassembly for veldup. They are identical in form and fit to the Q diaphragm valves, and will therefore, be considered a potential discrepancy source for the Q valves. i (2) The remaining non-Q valves (MS-21A, 215, 21C, 21D.2, 21E) have no Q valve counterpart that requires disassembly; and therefore, they were not reviewed. Referrina to Table 1 Pane 3 (1) No review is required for tha non-Q circulating water valves (MS-75). Referring to Table 1. Pase 4 (1) The main steam valves (MS-76, 77, 78, 79) are special valves and therefore were not reviewed. (2) Review of the specifications and vendor instruction manuals for the butterfly deluge valves and the HVAC containment isolation valves (MS-82.1, 86) showed the valve. installations to be bolted. Referrina to Table 1 Pane 58 (1) Review of the Q and non-Q control valves (MS-600, 601) shows that where soft seats are used, the internals must be removed prior to welded installation. Specification MS-600 (Q valves) has only four valves with soft seats. These valves (HV-4710, 4711 Data Sheets A0-19) are identical 4 inch 150 psi carbon steel globe valves. Specification MS-601 (non-Q) does not contain non-Q valves of similar configuration. Therefore, mismatch of valve internals and bodies need not be considered. (2) The vendor instruction manual for the process solenoid valves (MS-603) does not require valve disassembly. 800 Oak Ridgelhnipike Suite 501 Oak Ridge Tennessee 37830 (615) 482 79~3

e s QA/QC-RT-076 Page 3 April 8, 1985 (3) The instruction manual for the power operated diaphrage valves (MS-604) requires valve disassembly before welded installation. The specification contains four Q valves. These are identical 150 psi 4 inch stainless steel valves (Tag No. HV-5157, 5158, Data Sheets A2-12, 13). Specification MS-604 contains 1.-2 and 3 inch air operated non-Q valves. These valves Q and non-Q, are similar dimensionally to the air operated Q diaphragm valves in the NSSS purchase order, CP-0001 (S.O.0220). Therefore, there is a potential for mismatching parcs. Referrina to Table 1. Pase 6: (1) The non-Q automatic pump recirculation valves (MS-627) need not be reviewed. l (2) Per the specification for the pilot solenoid valves (MS-632), the valve ends are threaded. (3) The NSSS purchase order CP-0001 (Shop order 0220) contains valves supplied by Rockwell, Fisher, Velan, Copes Vulcan, Crosby, Westinghouse and ITT-Grinnell. Vendor drawings and instruction manuals were reviewed to reach the following conclusions: (a) The Crosby valves are safety and relief valves, and are not considered. (b) The Rockwell, Fisher, Velan valves have metal seats and do not require disassembly before weldup. (c) Some Copes Vulcan valves have non-metallic seats. However, the i instruction manual does not require valve disassembly before weldup. (d)' The ITT-Crinnell valves include 3 and 4 inch manual Q valves, similar dimensionally to those in MS-20B; and air operated Q valves from 3/4 inch to 4 inch, of which the 1, 2, 3, 4 inch valves are similar dimensionally to those in MS-604. Therefore, these valves, with the possible exception of the 3/4 inch valves, i will be added to the population of valves with the potential for having mismatched parts. L Further review and evaluation is required to better define the sub populations, taking into consideration the characteristics of the valve topworks. This effort will be limited to the ITT-Grinnell valves in purchase orders CP-020A, 0208, 0604 and 0001 (S.0. 0220). i i 800 04 Ridge hrnpike Suite 301 OA Ridge. Tennessee 3~830 (615) 482 ~973

i QA/QC-RT-076 Page 4 April 8, 1985 YMC / M. Solon / cc: D. Alexander V. Hoffman P. E. Ortstadt File VII.b.2.45 .MS/s1 3 800 04 Ridge ihrnpike Suite 501 Oak Ridge, Tennessee 37830 (615) 482 7973

s' e ' s t ' I I S N e ' *) e q \\ C Ng = 4 l E 5's. 7 D J, d= Mb2 4 t t s x T 2 L, - R S g. L v y S 4 s.

• C S

b*o 5 4 3 6 s ('( .i 41 s\\!. ,d D 5

• Q' 3

4 4 Q = N gs a S qw: s i Dg!

' i 's. s.p*S I4 h

} l s {'g;- i ? I N /Q ? W o 4[ 21 Yh

• [

y .; y

.i 4

.3 l .5 4. f; AT T. - $,IAo4su

== p i, s s e 7 *s t 5 % l to -{ { % 3 ",", 3 e g gi N 6 :% 4C 4 0, .j e ; n c .t %., k N \\.O C S): N4 s t u* gr. 3* h' 0 61 iN g "w" A (qg. gg g Fo rd ,j g= % : %q g uh= 5 .N T ) N D k-D e 4 '. S 7b h o - =- s c si c - - eg %G R y$ by L: \\ d* s M 7 8'= 4, , 't c ghn -f +4 %) t = [* 'a b $b 4 N CC g w' 'N S 4Nkk (Y h f I .) i S* Gf l h b '. 1 h h b I 2.. u C,, $1 ~~- . ~.. ... ~.. _. f h h N t ) ,1 i I-8 i .x. I 6. 1 g6 = E WG 9 6 d 4m -- c. s u s .5 > .1 d a y.31 W.,* { O' 4 e a e 4x =, aw r e. es u es we g s a :..s e = s. .s. > w - sw u.: a : ' s. =t w (g W 4 =

*

= d *

• u ra M4 w 14 s-6 g

6 3. l 3== = sa .to 3 gd 4 = .=d =s b =

1. = t, 3

w we 6-d com w e A ge g = s: <g c=g

c. - - u.n e

a u =. a v 1 /3 eNo e c-u e 344

4. N g

ta-e w h = 4 h 1 m 5 /v e l' w.g n: 63 6.

'c a

= n w 3 @ 'J. 9 .8 ) ed 99 9 ). .4 h W D -d

== b&= sta -d g g d e / 3

• et W3.* W.ag ed WD e 'J M.a. d 1

D m- = se b. 3 ed 'C.*.> . i b.=* g . f... a -6 = 3 .t, 4 e.s.

• s s

.J im 4g 4 e4 ETN'

i. } s:

y" a >6 = a .-<y J k - = 6a .s -4 c -e .d = {- -.=. . s =3 s as t 1 as r; i 3 e er = = ce is e e 6e

• - ia ===. -

ej t

a : =,

a. e o.s }== 4> e

• a u

5

=, =, = = =.

s. w.%.

nJ .s. :-. = y / :===- 24 ,s -. u E r. E> Z<> E.l N E44 z> D. = /. 3 x 'd fi = = -e ,/ 9 g%. i = / \\ 8 N N 3 f I I W *J. l t I* N e-w q 3

a. '- N 6

2 9 =t ,a . 4 =.' u 'q <2 I =3~ 3 e e C =. 4 = e = c c 2 = -e ? ?., es \\' f. i N2 r,w t,w. N, A .t .a. .o. .c a

== e .E .sD * :#l e e e e en % P u.,. b e 'a .e N an ir. Q ,.J ' W (( e e e

• E
• E Eg s c

= ': (S f* W 9 A

• 9
• e C

a .e a e n (. a g A E '4 E ' e e e j

Spec t ricat ines Hewis.luis/ Ailaines.lise He a.ese t " Jisly ??, 1983

r. egis 1 I)* ' u el, h vee w

.l M 'E drt_c, reu. Il Il L ..:'#{.__ H3 9~ t, At_. C-c> lvi en t u T S ~ ~ - ~ - - ' - ~ - f. ~ ' ~ ' - ~ '

• sts'-?utt ~%

v.e i vis us.c ra t wes ~~'~~'~ .. _13 terns t nes Sl.eios t.is. F i -- e I Its-/lA St.asio.e * & Se l f-At:lise l t?al Nt,n-Q VoWeC pe onre 4 4.a s t issus Velves t P I'C V'ho Od * $ g r.as esiclees er Emelles -flans-sseeclose r 185 - 7 111 fl.senea l 6 5e l l'- Ac t isa t ssel S t etse l - - - V.stwes 2.H leicipe s a: sma l lu r .,Non-Q VMv8C (dLp f 0x US 9 vlVS). s tt;..? t H. I II. sine.e l 4 Se l f - Actsen te:I %te?ni I deniet d in.f m 4 f.~f-Jo tv} g _ w 4 Ir r.,;.,,,,,,, /g V '.I ves 2.te l ese:Ise s & Sm.n l l e r (5iseaselers Illaples mete lylan) .} .co o g,g a : _ C o o v-c e b.,_de,s c y g p_m c 4 _ ns-plc H.essisa l anet Sati r-As:Llea t eil Mrsus7s* No rs - Q ValveC & l'.e 6: Ia esen vaIvee 2-1/? 4see:Ises p not A C $'d eut ._... h I ** 8 '8 " I _ _.

u... _ _.

__ 4e3-q Vatvec ( appe, szs vle) j ns-?lp n......a t a se l t-A : pa t e I st ee l 7 g g g,., ,, 4 g ,,, g v.e s.es a sann.Iers lilaglie av p7,,,, t y. I 7.s.ncl.es a t.rea te r y j Cons": lev et. t coucer 4, d i ce p u e sg _ ___ _essen-seuclear i 1 *.*.- ? l p. I Masnea l reswet r Ope ra t e:I & Se l f-Ac tina t est Stee t Valves 2.5 issclies Non. Q q alyc 0 i & lasejer SE Ve h n of IW tib-? l p. 2 rower E3peratent I =t ract inse Strame Elecck valvers toon-nuclea r ) tts-fit Itog e nrrly Water Blisc. Valves lien t' rest ec i Pasm/ f 'N na.. c....iii.. [t usi 11:n-// l n..e3 h f:.,. / t ecitwater llea ters. m.. Ma.'es 414- ?'.," I ravellsig Water Se:e stesisx rse.i:. ras s-d.e f e?e y Itn l a teil a Gn// Styss hvemm rarme A1S! 2 er (. ---a

+ e 6 i N '2s n 9 Y b N ? d 5 Q / N

• n I

r a Q S 2 k b 4 9 s' s a g d b g% ( >s 4 g M 5# o p Q eu. =h e{ kNP e '4 De C E n. ? 5"4 q' te a

• s 0's N

\\ t b 'N ) - g O' r D cm = e k -s, 6 g w 4 d w W G els 3 4 ed F w 3 .g = d d =. -= 4 C E e og d U d s = = = 3'. d .n.,s,' w a = r 2 n m 9

== L e e s 3m b 6 i-e 48 e 7

• 8 W

-.tA. T ee d .a 3 a

: 9 e 's :

e. .t. "3 3 .s =.* 3-8B "3 D C 3 e w - e = =d e 3 3 m e 6 %.d dd b d ? d C 2 d .C. se tb cwa 3 e u= e ee q y w -6 2 ) e-* A > #4.7 1 ab = = T cA .= wt es = . = = 3 3 s* .A = = a a 3 a C4 e f4 - E, E-O n= m 34 - a u .s..=. w ~* s of. :w y g w. a 6 -

=-

e >. = 8 = a

e a e z.

<

a.: o c .: = e- = u. 3 ..= e. s = .g u

== + o: :. .a .s. =: es 4 2 - -= a. 'a e 3 = t mg 6 .f => LZ --

L b C C h n \\ w I s o = =. s I w. s a= m tS N '- s 6e:A edZ s 2 4 s 3-e = s< ur

4=

=. =

== e' c/ e

• 4

=

== - N a ". g i

• C.a =

g; c c c .e ~

== s

• ** *=

e = "J [ W-e o e Ut e M fA e e e e L e f. 69 ag M A g A A A A 1 = =,a T C E E g .Ys .2 e a ,A. g g 4 g g

== s'

e I M e s 3 N I t 4 k1 .i w ~ l k N i \\,% -c w i q-f. i \\ s. ~ t. = t t' h Li x N N

=

l =d ~u e e o c F:t s i o ~ u q 5 s o e s

== s N -e b. _2 \\ s ~~ a 3 1% b! e u E t q 'C U>= t '4

r..: :

-C D t

. - :n s

s L: s e w t v c N g lv 4 4 M l !Q ' H l.a -. w -)

.r x N

jg wl c - !,s L;H = .n._. t -c s. u C h \\ sI c w I i O T' 1 i i t 4 I / .' s u g I,

= E, a. -a c = = i _-

s s 3

o a F i. ) e -:/.= c h im A im b g g > i 4 I s - = =.C 3 u g= z. l =.= o = s t _ =y E a s : s 3 3 z e - = = q g a as s s e I. t

s. ;=

/w- ,o- = e

c a

c : s. u t.. 3 = s fJ O-

• = **

T = Cd .s.,. f,. g g

• .I

- = e e s* T T f- ~5 Y I EC g; w 'F d g

• J e

U, =. bE C 'J 4 e 4" >v 3 < E ,h e 3 s -l .9 g 8

3 = =

= 43 e3 g N ** 3 Ma = A. / 3 c !i e oe.- , - e i < /T o = o. 6. 3 2. ? c e e s 'd f h

• ?

n A e

• .e C

7 er = g y to

• ~

6.= - ef d st* e* E3 g .s.n. E g ,C A 4* A e. g e a a a

• g

m 2.cc 4 r ic t ien. Itev i s Ioss/ A.Idess. lum Cca <> - 3 Jes t y ;';', 1981) ragu 18e " Doc _ U rss R C.J.L. D Y 'o 'l ( C ##'7/*18 C T J SIi ty{lk Illi4 $ Pf C r.' s. g A s_ fl c'<f uor* :.:.o /* /.ss., f so n t -.... s / 1,,o st w o.s e d e e ns.... u s

r...i..

sle..e.e s e.1 con t a a l m.,e e,. y x /u,,fre n / v " r/..:- ,ff /, 3,ff 3m /,,o /)", y,,g v.. e aum.Inne Ica e / / _____..3..,..,.,.,..,,_,_1,...,,,..gy,_.____. g,_,,,,

n..e,,;l e.....s e.. I va l vo s reuss.teuc s ca r go y yjy,

,( ga,,,,,p p.,, p3_ gen u,;,j, / v' s.rf = ~ rc r,Sj, e, .,c. .. i

r.... e s s :. l on... o ve :v.is y,, (w

/ / Di.r a s:*9 AM.ac9,d ,, i s.,.....

c.,...

u.....t.nl u. rier.9m m -,,,,-. r ,c.. <a 7 o f m m,, e. 4 m.,,..,,. q c r r- <. J. p g, s = i,, D' -1 f .c.=. e " e-- =* =., -6 e c g 3 es a= m M sa- = c. i m.. i -u Y f; 22 0 s. e , fe :E 22 e' c= I :s : se J.2 k # - jj ~i .:2 m = }

48

=

W t> D w m ": j AC .*j 9 D = 6,5 1 8 c. =, e, p c z.

g. 2

'lr 1 ? e e e vi N u .a. m,. e.: A .s. o y a n. e. .a e 7 [Ty dj 3 W E l E E s' e,

f Eva nation Research @Ip0 rad 0n OrrtCE MEMoRinnuM QA/QC-RT-090 TO: J. L. Hansel rROM: M. Solon DATE: May 2, 1985

SUBJECT:

Valve Disassembly, Issue VII.b.2 Additional Generic Valve Evaluation

References:

(1) Office Memorandum, M. Solon to J. Hansel, " Generic Valve Evaluation", dated 04/08/85 (2) SDAR CP-83-01, Corrective Action for Borg-Warner Check Valves (3) Telecon, M. Solon to P. Milinazzo, " Disassembly and Reassembly of Borg-Warner Check Valves", dated 04/22/85 Summary Reference 1 evaluated the generic valve types which required disassembly prior to welded installation into the piping system. The objective of this further evaluation is to determine if there are generic valve types which required disassembly and subsequent reassemblysafter the valves were delivered to the site. It was determined that although many types of valves were disassembled and reassembled for purge, flush, test and repair, there was only one generic valve type (in addition to those in Ref.1) which required disassembly. These were check valves supplied by the Borg-Warner Nuclear Valve Division (B-W), under P.O. No. CP-0020B.1. There are approximately 160 valves, total for Units 1, 2 and Common, which fall into this generic type valve category. It was concluded that of this total only some of the low pressure (150 and 300 psi) valves could be reassembled with an incorrect body / bonnet generic configuration. All valves in question are ASME III, Code Class 2, and therefore, code classification violations could not have occurred. Discussion B-W check valves were found to have possible design and manufacturing deficiencies (Reference 2), which required that the valves already en site be disassembled for inspection and repair if required. Review of the B-W check valve drawings, with confirmation by the vendor ~ (Reference 3) resulted in the conclusion that valve bodies and bonnets of the same size and pressure rating could have been reassembled, regardless of ASME III Code Class (Class 2, 3) or material (carbon, stainless steel). However, per the specification (MS-20-B.1, paragraph 3.3.3) the valves were all supplied as Class 2. M b. 2- $h ') J 800 Oak Ridge'Ihrnpike Suite 501 Oak Ridge, Tennessee 37830 (615) 482 7973

t QA/QC-RT-088 Page 2 May 2, 1985 A matrix of B-W check valve types is given in Table 1. All valves are ASME III, Code Class 2. Valve types which have the same body / bonnet fit-up are circled. The valves which could be reassembled with incorrect bonnet and internals are as follows: 3 inch /150 psi (carbon and stainless steel) 4 inch /150 psi (carbon and stainless steel) 10 inch /150 psi (CS and SS) and 300 psi (CS) There are approximately 70 valves falling into these categories. Valves which were disassembled, other than those defined herein and in Reference 1, will be identified by reviewing operations travelers. &n M. Solon / cc: D. Alexander V. Hoffman P. Ortstadt ERC File File VII.b.2-4B File VII.b.2-9 MS/s1 Attachments I I 1 l 1 2 800OakRidge'lbrnpike Suite 501 OakRidge, Tennessee 37830 (615)482-7973

T46LZ f Boe 4 - 41sgen ex Cn c x Vaz.es T.,ps r AMs1 Peess s oncn

• / in cu 6 inca 8 incu

/0 in cu 16 inca R n rin g iso eci e s /g g c s/,, c g jg g cg es SS ....O'S, / SS s.. 360 pst 900 pss CS CS CS CS $o76.5 : G) Att 494 re-c sea ACNt.?27, 6s e Cm x.c 2. (z) s s = srsine ss.c sra auey j aanner CS= Caeson Sri Ba by $ Bonadr b) so~e voc.vs 2oay/s.naer i=>rros = % A sN/as

l< 1 %'ah13tlGR s On 0FFICE MEMORANDUM QA/QC-RT-103 TO: J. L. Hansel FROM: M. Solon DATE: May 20, 1985

SUBJECT:

Valve Disassembly, Issue VII.b.2, Generic Safety Consequences Analysis ]

REFERENCES:

1. Memorandum QA/QC-RT-076, " Valve Disassembly, Issue VII.b.2 Generic Valve Evaluation," April 8, 1985

2. Memorandum QA/QC-RT-090, " Valve Disasse.nbly, Issue VII.b.2 Additional Generic Valve Evalu' tion," May 2, 1985 a

J. Telecon, M. Solon and B. Borst (ITT-Grinnell), April 9, 1985

4. Telecon, M. Solon and B. Borst (ITT-Grinnell, May 15, 1985 5.Telecon, M. Solon and P. Milinazzo (Borg-Warner), April 22, 1985

SUMMARY

The generic valve types that required disassembly and reassembly were identified in References 1 and 2. The safety implications resulting from reassembly of incorrect valve components were evaluated, and are summarized as follows: 1. Manual and air operated ITT-Grinnell diaphragm valves (except the 3/4 inch, stainless steel, Class 3, air operated valves), if reassembled l with incorrect bonnet assemblies, could result in significant safety j implications ranging from violation of the ASME III code

• to failure i

of the valve. 2. The following Borg-Warner swing check valves, if reassembled with inecerect bonnet assemblies, could result in corrosion problems, potential failure of the bonnet and/or loss of function of the valve: a. ' Three and four inch /150 psi valves b. Ten inch /150 psi and 300 psi valves The combination of valve bodies and bonnet assemblies which can be bolted up are shown in Table 1 (manual diaphragm valves), Table 2 (air-operated diaphragm valves) and Table 3 (Borg-Warner check valves). The potential generic safety consequences of incorrectly reassembled valves are summarized in Table 4.

• Code violation herein loosely defined as an ASME valve reassembled with a bonnet assembly from a lower ASME class valve.

._va.;u-( 4B I _ - 800 Oak Ridge 1hrnpike S-ite 501 Oak Ridge, Tennessee 37830 (615) 482-7973.. _ - _..

i 4 QA/QC-RT-103 Page 2 May 20, 1985 Valves which do not fall into the generic categories defined in References 1 and 2 will be treated on a case by case basis. Since there are many different valve types which were disassembled for test, repair, flush, etc., generic evaluations prior to defining the population are not practical. A recommended approach is given in Section 3 of Discussion. Discussion In accordance with the Action Plan, paragraph 4.1.3, an evaluation was made to determine the consequences of reassembling incorrect bonnet assemblies on valves which required disassembly. The two generic types of valves identified in References 1 and 2 were evaluated and are discussed below. 1. ITT-Crinnell Diaphragm Valves The ITT-Grinnell diaphragm valves were supplied under the following purchase orders: Purchase Order, CP-Description 0020A ASME III, Manual, 2 Inch and Smaller 0020B ASME III, Manual, 3 and 4 Inches 0604 ASME III and Non-ASME, Power Operated 0001 ASME III, Manual and Power Operated 0021B.1 Non-ASME, Manual, 2 Inch and Smaller 0021D Non-ASME, Manual, 3 and 4 Inches Based on References 1, 3 and 4, the following conclusions were drawn regarding possible reassembly configuration errors and resulting differences in valve construction: a. Valves of the ssue size have the same body / bonnet fit-up, regardless of ASME III Class (including non-ASME), material and pressure rating. b. Bonnet material is stainless steel regardless of body material - (Stainless Stael or Carbon Steel). c. Bonnet wall thickness depends on valve size only, and is the same for 150 psi and 300 psi ratings. d. Diaphragm thickness depends on valve size only, and is the same for 150 psi and 300 psi ratings. However 300 psi, 2 inch, 3 inch and 4 inch valves have a diaphragm support cushion. e. Two, three and four inch, 300 psi manual valves use a brass spindle; whereas the 150 psi valves and the 300 psi air operated valves use a stainless steel spindle. All other internals are of the same materials.

QA/QC-RT-103 Page 3 May 20, 1985 ~ The following additional information was obtained from the valve drawings. f. Operator action (air to open or close) was determined and is summarized in Table 2. Except for the 4 inch valves, all the valve operators with the same action were the same size for a given valve size. g. The 4 inch Class 2 valves have a larger actuator than the 4 inch Class 3 valves. It is presumed that reassembly of a manual valve with a bonnet assembly having an air operator, or vice versa, is not credible. Such an error would be obvious, both visually and during preop testing. The evaluation was performed for the highest level of valve (be it ASME Class, pressure rating or material), assuming reassembly with a bonnet from a valve of lower level. In addition, valve operator action and size was considered. The possible reassembly errors were obtained from Table 1 (manual valves) and Table 2 (air operated valves) wherein the number of ASME valves, broken down by class, pressure rating and material, are shown for each valve size. The various types of non-ASME valves are also shown in the tables. Except for the 3/4 inch and 4 inch air operated valves, non-ASME valve bonnets could be installed on the ASME valves. A summary of the evaluation is given in Table 4, Items 1 through 10B. Except for the 3/4 inch, Class 3, 300 psi, stainless steel valves (Item 7B), reassembly with an incorrect bonnet assembly could result in a code violation and/or potential valve failure or loss of function. 2. Borg-Warner Swing Check Valves The Borg-Warner swing check valves were supplied as part of purchase order CP-00208.1. Based on References 2 and 5, the following conclusions were drawn regarding possible reassembly configuration errors and resulting differences in valve construction: a. Except for the 10 inch valves, only valves of the same size and pressure rating have the same valve body / bonnet fit-up, b. Ten inch valves have the same body / bonnet fit-up for 150 psi and 300

psi, c.

Carbon steel valves have carbon steel bodies, seats and bonnets. Stainless steel valves have stainless steel bodies, seats and bonnets. d. Except for the 10 inch carbon steel valves, all valves have stainless steel disks. The 10 inch carbon steel valves have carbon steel disks. 1 i 800 Oak Ridgeihrnpike Suite 501 Oak Ridge Tennessee 37830 (615) 482-7973

c' s QA/QC-RT-103 Page 4 May 20, 1985 e. All. valves were provided as Class 2, regardless of class specified. The possible reassembly errors were determined from Table 3, wherein the number of valves in each assembleable category is given. Only the 3 and 4 inch 150 psi valves and the 10 inch valves could be reassembled with body / bonnet errors with potential safety significance. A summary of the analysis is given in Table 4, Items 11 through 13. In each of these cases, reassembly errors could result in valve failure or loss of function. 3. Other Valve Types Dis / Reassembled Analysis of the generic valves for safety consequences is practical only for the ITT-Grinnell diaphragm valves and Borg-Warner check valves. These valves were known to have required dis / reassembly of all the valves. This type of analysis for the remaining valves that were dis / reassembled for repair, test, flush, etc. should be done on a case by case basis. The recommended approach would be to include all the other valves

• in the population. When a valve is selected as a sample, the documentation should be reviewed to determine if adve se effects could result from errors in reassembly.

If no adverse effects are identified, the valve should be discarded from the sample, and another selected. If the evaluation is not j conclusive, the valve should remain in the sample, and the evaluation would take place after the valve is inspected, if discrepancies are found. $hYr1+1h ?b Tb as M. Solon / cc: D. Alexander V. Hoffman P. E. Ortstadt File VII.b.2-9 File VII.b.2-49 ERC File MS/s1

• Other screening criteria, e.g. short time span between disassembly and reassembly, may be considered to eliminate valves from the population, i

2 800 Oak Ibdge hrnpike Suite 501 Oak Ridge. Tennessee 37830 (615) 482 7973

CS SS cs ss

1. 74 0 o 4 4.

/)ys.<< tit 3lY " vd/ N.' op /" y 9 y v m e ss& a e 1 /fC N l7 y ll /S0 # 30 6 19 +/o re t (7YP) 3 r9 t /2 3ros I ce ce cs ss /)A-}AJt.odt. l f] !)nt..?L 6,n /. 'y 2" n a3 3" .,i .m 'v m / d b-b 6 /50 # y6 // yy /Sc" / 7 27 l I d'

• 0
• y2 3c^=

25 /DB i C5 SS Manunc. an-4:ME M/ /cua t. 'b l / ,/ / ,/ / A .L Q /30 # to 26 9 .o, .) y =% q k 300 # L 5 e E, R 4SQ M ml s s s s i doa : 1. A/wese or has s 7ku / /2 b w u, :. hikpmAc.v1 Vaieis

3/4 - 3 0 0.fsi /"- 3 e o pri /7itz Opsp.; reb - A 70 /-Me dree,4rsn - A 7 o c S -et 2 5 5-CL 5 C 5 -ct 3 ss y // ore i 'E CL 2 CL? fryP) / // 3 z* - E O o psi 3 " -. 5.,.. gri. Att Cpeakreb /?st Opcsnreb S r. C ri.. Ct 2 CL 3 CL z. dL3 A ro Aro ATC ATO ATO ATC. /SO 51 Y / 3cro psi 3 /6 '/ " - Sr. Src - N':nu-A SMS t Adz Off9ATEb })t]2 O 64, W 2p kl~O f S r. 3rt. ^ dlz CL 3 -a ii / 2,, l 0 ATO ATO ATc /60 ISO /SO 3m psi Psi psi Ps i /SOfSI 4/ 300fd 2-l 7~984E 2 /QE2 Oft #4710 bihP///A4n VLVS hlo16 : /. llu N ele of VA wtS E. 8 70 = 41/; 70 epr+) OYC ~ Atti 70 Clos 6

b 3" - 15c Pti ~2,' ' - 'f 0 o p s i. d5 55 C5 55 dorsi. A/m we /9

• W l lb oc viv$

(??'/) Y"- /SOPU '/"- 90 0fst C3 55 Cs SS /O 3 Z/ .o e w.m.. o Pem

• 5 9

M6e & "- /30 psi & "- Too rsi I cs SS cS ~S'S ~ 5 ~ ~ z'8 L ~ 8 8 " - > [ O P.: i 8 - 9c/0 fs/ f dC CC CE 5E ggg g i $0/E&

• hkU$f b

E C#6c.t VnwtS /0 " /6 "-300 Psi f f5 CS c5 l 55 /5:0. :i L d .- i, d O, [ f:/

e ? TABLE 4 GENERIC SAFETY CONSEQUENCES ANALYSIS Page I r ITEM DESCRIPTION SAFETV PRESSURE POTENTIAL POTENTIAL i CLASS RATINC. REASSEMBLY ERROR FAILURE'8 EFFECTS I ITT-Grinnell 1 Diaphragm Valve 3 300 psi 1. Bonnet assembly from C.S. 1. No failure. All bonnets are St. St. with internals Manual valve of same materials. 3/4 inch Stainless Steel 2. Bonnet assembly from 150 psi 2. No failure. The bonnet and diaphragm thicknesses are the valve same for 150 psi and 300 psi valves. 3. Bonnet assembly from non-ASME 3. a. Potenial failure during a seismic event. Loss of valve function leakage. b. Code violation. i ITT-Grinnell l 2 Diaphragm Valve 2 150 psi 1. Bonnet assembly from non-ASME 1. a. Potenttal failure during a seismic event. Loss of Manual i valve function, leakage. 3/4 inch i 2. Bonnet assembly from ASME III. ,b. Code violation. Carbon Steel I Class 3 valve 2. Code violation. ITT-Grinnell

• Diaphragm Valve 2

' 150 1. Bonnet assembly from C.S. 1. No failure. All bonnets are St. St. with internals of 3 f Manual velve same materials. I 1 inch 2. Bonnet assembly from non-ASME 2. a. Potential failure during a seismic event. Loss of l Stainless Steel valve function leakage. 3. Bonnet assembly ASME III, Class 3 3. Code violation. valve

e e ti ? t TABLE 4 (Cont'd) r.ENERIC SAFETY CONSEOUENCFS ANALYSIS (Cont'd) Page 2 SAFETY PRESSURE POTENTIAL POTENTIAL ITEM DESCRIPTION CLASS RATING REASSEMBLY ERROR FAILURE 4 ITT-Grinnell Diaphragn 3 300

1. Bonnet assembly from C.S.

1. No failure. All bonnets are St. St. with in-Stainless Steel.

' Valve. ternals of sane materials. valve Manual 2 inch

2. Bonnet assembly from 150 2.a. Galling of St. St. spindle (300 pst valve psi valve, spindle is brass). Januning of 6alve.

l

b. No support cushion. Failure of diaphraqm

& leakage. g f.

3. Bonnet assembly from non-3.a. Potential fatture during a seismic event.

ASME Valve. Loss of function. leakane.

b. Code violation.

j i 5 ITT-Grinnell Diaphragm 2 150

1. Bonnet assembly from C.S.

1. No failure. All bonnets are St. St. with in-l Valve Manual 2 inch Valve.

ternals of same materials. Stainless Steel.

2. Bonnet assembly from non-2.a. Potential failure during a seismic event.

4 ASNE Valve. Loss of function & leakage. l

b. Code violation.

3. Bonnet assembly from ASME

3. Code violation.

j III. Class 3 valve. ITT-Grine11 Diaphragm 2 300

1. Bonnet assembly from C.S.

1. No failure. All bonnets are St. St..with in-6 Valve Manual 3 inch &

Valve. ternals of same materials. 4 inch Stainless Steel.

2. Bonnet assembly from 150 2.a. Galling of St. St. spindle (300 pst valve i

psi valve. spindle is brass). Jacining of valve. a

b. No support cushion. Failure of diaphraom

& leakage.

3. Bonnet assembly from non-3.a. Potential failure during a seismic event.

ASitE Valve. Loss of function & leakage.

b. Code violation.

l

4. Bonnet assembly from ASHE

4. Code violation.

III, Class 3 valve. i.

e ? TABLE 4 (CONT'D) GENERIC SAFETY CONSEQUENCES ANALYSIS Page 3 i ITEM DESCRIPTION SAFETY .'RESSURE POTENTIAL POTENTIAL CLASS RATING REASSEMBLY ERROR FAILURE i 7A ITT-Grinnell 2 300 1. Bonnet and actuator 1.a. Code violation. Olaphrage Valve assembly.from Class b. No failure. Actuator action and size the same. Atr operated (ATO) 3 valve 3/4 inch Carbon Steel 78 Stainless Steel 3 300 1. Bonnet and actuator 1.a. No failure. All bonnets are St. St. with internals assembly from C.Stl.. of the same materials. Actuator action and size the Class 2 valve same. t n 8A ITT-Grinnell 2 300 1. Bonnet and actuator 1.a. Code violation Diaphragm Valve assembly from C. Stl., b. No failure. All bonnets are St. St. with internals of Air operated (ATO) Class 3 valve the same materials. Actuator action and size the same. 1 inch Stainless Steel 2. Bonnet and actuator 2.a. Code violation. assembly from non-ASE. b. Potential failure during a seismic event. Loss of 150 psi valve function & leakage. 88 3 300 1. Bonnet and actuator 1. Same as 2 above. assembly from non-ASE. 150 pst valve 9A ITT-Grinnell 2 300 1. Bonnet and actuator 1.a. Code violation. Diaphragm Valve assembly from Class 3 b. Incorrect actuator action and system operation. Air operated (ATO) valve with ATC actuator 2 inch and 3 inch Stainless Steel 2. Bonnet and actuator 2.a. Code violation. assembly from non-ASE. b. Potential failure during a seismec event. Loss of 150 pst valve function & leakage. 98 Air operated (ATC) 3 300 1. Bonnet and ATO actuator 1. Same as 2 above. assembly from non-ASME, 150 pst valve 2. Incorrect actuator action and system operation. i ( 4

o ? TABLE 4 (Cont'd) i GENERIC SAFETY CONSEfM1ENCES ANALYSIS (Cont'd) Page 4 l SAFETf PRESSURE POTENTIAL POTENTIAL l ITEM DESCRIPT!0'i CL ASS RATING REASSEPSLY ERROR FAltuRE i 10 A ITT-Grinnell Diaphragm ? 150

1. Bonnet and actuator assembly 1.a. Code violatten valve air operated (ATO) from class 3. ATO valve,

b. Smaller actuator; slower valve I

4 inch Stainless Steel opening & clestiin times

2. Bonnet & actuator assembly 2.a. Code violation from class 3 ATC valve.

b. Incorrect attuator action and system

/ operation. 10 B Air operated (ATC) 3 300

1. Bonnet and actuator assembly 1.a. Fallure of bonnet seal and/or bonnet from class 2, 150 psi, ATO cover. External leakage, valve.

b. Incorrect actuator action and system operation.

11 Borg-Harner swing check 2 150 psi

1. Bonnet assembly from C.S.

1.a. Corrosion and potential failure of valve M Stainless'

valve, bonnet. Contanimation of system from corrosion products.

3*M f MCII3 M* #

b. Corrosion of C.S. seat. Loss of leak tightness & check valw function.

12 Borg-Warner swing check 2 300 pst

1. Bonnet assembly from 150

1. Failure of bonnet seal and/or boneet l

valve 10 inch Stainless psi valve, cover. External leatage. Steel. l

2. Bonnet assembly from C.S.

2.a. Corrosion of bonnet. Potential valve. failure of bonnet. Contamination of system frce corrosion products. j

b. Corrosion of C.S. seat. Loss of leak l

tightness and check valve function.

c. Corrosion and failure of 'C.S. disk.

d 13 Borg-Warner swing check 2 150 psi

1. Bonnet assembly from C.S.

1.a. Corrosion and potential failure of valve 10 inch Stainless

valve, bonnet. Contamination of system from corrosion products.

b. Corrosion of C.S. seat. Loss cf lea 6 tightness and check valve function.

c. Corrosion and failure of C.S. disk.

Rev. I s/23/gg

1 a' 'AL4JATION g J wdCEA"CM COTPOWATION QA/QC-RT-149 TO: J. L. Hansel FROM: J. N. Barger DATE: June 19, 1985

SUBJECT:

Valve Disassembly, Issue VII,b.2, Dis / Reassembly Procedural Control

REFERENCE:

Memorandum QA/QC-RT-106 Review of the construction and QA procedures"have been completed. Based on the review it was found that construction procedure CP-CPM-9.18 Revision 0 and QA procedure QI-QAP 11.1-26 require positive identification of parts for valves listed in supplements of CP-CPM-9.18. This function is controlled by QI-QAP 11.1-26 which requires the use of an approved, standard form, QC Checklist (QCV). The QCV lists inspection points for positive identification of valve parts which includes body, bonnet and disc heat numbers and, where prescribed the application of match marks for alignment purposes. Valves not addressed in supplements of CP-CPM-9.18 are dis / reassembled in accordance with construction operation travelers. These travelers are prepared in accordance with construction procedure CP-CPM-6.3 and further covered in QA procedure QI-QAP 11.1-26. CP-CPM-6.3 requires that the valve part, serial or tag number be recorded on the traveler prior to the start of valve disassembly. Additional positive information such as body, bonnet and disc heat numbers were included in some cases by personnel initiating the traveler, but was not required. Subsequent to issuing CP-CPM-9.18 and QI-QAP 11.1-26, positive identificatien of most valves were recorded prior to the start of valve disassembly. A number of valves have been dis / reassembled more than one time. Therefore, it is conceivable that a valve may have been dis / reassembled using the early procedures and again using the current procedures, i l Based on the forgoing it is concluded that valves reassembled using early l procedures had more potential for reassembly errors than using the procedures now in e f f ect'. The significant difference being that the earlier procedures did not require recording the bonnet, body and disc heat number before disassembling the valve. The potential for reassembly error is considerably reduced for valves disassembled for the first time after the establishment of the QCV. sta;6.1 - 7, o -l

i l 8 0E WATION ) WWWEATCH COWPOWATION The assessment made in the reference memorandum has changed due to the large percentage of valves dis / reassembled using early procedures and some valves currently not covered by QCV. Therefore, the subpopulation for Issue VII.b.2 vill not be made up of valves dis / reassembled using early procedures. The basis for the subpopulation will be finalized and reported in the near future, J 4 arLM sa.n N.B/rger cc: D. J. Alexander M. Obert & V. Hoffman FILE VII.b.2-4 File VII.b.2-9 / ERC File JNB/sp l l )

a VALUATION E (3CEAWCM COTPOWATION QA/QC-RT-688 TO: File FROM: M. Obert DATE: October 2, 1985

SUBJECT:

Review of Procedures Pertinent to Valve Disassembly The following procedures were reviewed including a review of the historical file of previous revision: Procedure No. Title CP-CPM-6.9 General Piping Procedure CP-CPM-6.3 Preparation, Approval, and Control of Operation Travelers CP-CPM-9.18 Valve Disassembly / Reassembly QI-QAP-11.1-39A Valve Disassembly / Reassembly QI-QAP-11.1-26 ASME Pipe Fabrication and Installation Inspections The results of these reviews are reported in Memorandum QA/QC-RT-149 dated 6/19/85 and in the ISAP VII.b.2 Results Report. / -. X X 7//W-h^ M. P. Obert MP0/my i l Mir. 6.2-7. 0- %

w-A:h; R3vicion't 1 11/25/s5 Page 1 of.6 ITEM NUMBER VII.b.2 GENERIC SAFETY CONSEQUENCES ANALYSIS TTEM DESCRIPTION SAFETY PRESSURE POTENTIAL POTENTIAL CLASS' RATINC REASSEMBLY ERROR FAILURE & EFFECTS 1 ITT-Crinnell 3 300 psi

1. Bonnet assembly from C.S.

1. No failure. All bonnets are Stainless Diaphragm Valve valve.

Steel with internals of same materials. Manual 3/4 inch Stainless Steel

2. Bonnet assembly from 150 psi

2. No failure. The bonnet and diaphragm

valve, thicknesses are the same for 150 psi and 300 psi valves.

3. Bonnet assembly from non-ASME

3. Code violation.

valve. 2 ITT-Crinnell 2 150 psi

1. Boanet assembly from non-ASME

1. Code violation.

Diaphragm Valve valve. Manual 3/4 inch Carbon Steel

2. Bonnet assembly from ASME III,

2. Code violation.

Class 3 valve. 3 ITT-Crinnell 2 150

1. Bonnet assembly from C.S.

1. No failure. All bonnets are Stainless Diaphragm Valve valve.

Steel with internals of same materials. Manual 1 inch Stainless Steel

2. Bonnet assembly from non-ASME

2. Code violation.

valve.

3. Bonnet assembly ASME III,

3. Code violation.

Class 3 valve. j 1D.b.2 4 i

r O. Revicion: I 11/25/85 Page 2 of 6 ITEM NUMBER VII.b.2 GENERIC SAFETY CONSEQUENCES ANALYSIS REM DESCRIPTION SAFETY PRESSURE POTENTIAL POTENTIAL CLASS RATING REASSEMBLY ERROR FAILURE & EFFECTS L ITT-Crinnell 3 300

1. Bonnet assembly from C.S.

1. No failure. All bonnets are Stainless Dicphragm valve valve.

Steel with internals of same m'aterials. Manual 2 inch Stcinless Steel

2. Bonnet assembly from 150 psi

2. a. Possible galling of Stainless Steel valve.

spindle (300 psi valve spindle is brass).

b. No support cushion. Reduced diaphragm life-increased maintenance.

3. Bonnet assembly from non-ASME

3. Code violation.

valve. i ITT-Crinnell 2 150

1. Bonnet assembly from C.S.

I'. No failure. All bonnets are Stainless Dicphragm Valve valve. Steel with internals of same materials. Manual 2 inch Stcinless Steel

2. Bonnet assembly from non-ASME

2. Code violation.

valve.

3. Bonnet assembly from ASME III,

3. Code violation.

Class 3 valve.

e Revicinn: 1 11/25/85 Page 3 of 6 ITEM NUMBER VII.b.2 CENERIC SAFETY CONSEQUENCES ANALYSIS TEM DESCRIPTION SAFETY PRESSURE POTENTIAL POTENTIAL CT. ASS RATINC PEASSEMBLY ERROR FAILURE & EFFECTS 6 ITT-Grinnell 2 300

1. Bonnet assembly from C.S.

1. No failure. All bonnets are Stainless Dirphragm Valve valve.

Steel with internals of same materials. Manual 3 inch & 4 inch Stainless

2. Bonnet assembly from 150 psi

2. a. Possible galling of Stainless Steel Stsel

valve, spindle (300 psi valve spindle is brass).

b. No support cushion. Decreased diaphragm life-increased maintenance.

3. Bonnet assembly from non-ASME

3. Code violation, valve.

4. Bonnet assembly from ASME III,

4. Code violation.

Class 3 valve. A ITT-Grinnell 2 300

1. Bonnet and actuator assembly

1. a. Code violation.

Dirphragm Valve from Class 3 valve. Air Operated (ATO)

b. No failure. Actuator action and size 3/4 inch Carbon the same.

Stcel B Stainless Steel 3 300

1. Bonnet and actuator assembly

1. No failure. All bonnets are Stainless from C. Sci., Class 2 valve.

Steel with internals of the same materials. Actuator action and size the same.

r-R vision: 1 11/25/85 Page 4 of 6 ITEM NUMBER VII.b.2 CENERIC SAFETY CONSEQUENCES ANALYSIS TEM DESCRIPTION SAFETY PRESSURE POTENTIAL POTENTIAL CLASS ' RATING REASSEMBLY ERROR FAILURE & EFFECTS 8A ITT-Crinnell 2 300

1. Bonnet and actuator assembly

1. a. Code violatian.

Diaphragm Valve from C. Scl., Class 3 valve. Air Operated (ATO)

b. No failure. All bonnets are Stainless 1 inch Stainless Steel with internals of the same Steel materials. Actuator action and size the same.

2. Bonnet and actuator assembly

2. Code violation, from non-ASME, 150 psi valve.

88 3 300

1. Bonnet and actuator assembly

1. Same as 2 above.

from non-ASME,150 psi valve. 9A ITT-Crinnell 2 300

1. Bonnet and actuator assembly

'I.

a. Code violation.

Diaphragm Valve from Class 3 valve with ATC Air Operated (ATO)

actuator,

b. Incorrect actuator action which would 2 inch & 3 inch be discovered during testing.

Stainless Steel

2. Bonnet and actuator assembly

2. Code violation.

from non-ASME, 150 psi valve. 95 r.ir Operated (ATC) 3 300

1. Bonnet and ATO actuator

1. Same as 2 above.

assembly from non-ASME, 150 psi valve.

2. Incorrect actuator action which would be discovered during testing.

I

r v R2vicion: 1 11/25/85 Page 5 of 6 ITEM NUMBER VII.b.2 GENERIC SAFETY CONSEQUENCES ANALYSIS 'EM DESCRIPTION SAFETY PRESSURE POTENTIAL POTENTIAL CLASS RATING REASSEMBLY ERROR FAILURE & EFFECTS A ITT-Grinnell 2 150

1. Bonnet and actuator assembly

1. a. Code violation.

Dirphragm Valve from Class 3, ATO valve. Air Operated (ATO)

b. Smaller actuator. Incorrect actuator 4 inch Stainless action which would be discovered Stsol during testing.

2. Bonnet and actuator assembly

2. a. Code violation.

from Class 3 ATC valve.

b. Incorrect actuator action which would be discovered during testing.

B Air Operated (ATC) 3 300

1. Bonnet and actuator assembly

1. a. Incorrect actuator action which would from Class 2, 150 psi, ATO be discovered during testing.

valve. Borg-Warner 2 150 psi

1. Bonnet assembly from C.S.

1. a. Corrosion and potential failure of Swing Check

valve, bonnet. Contamination of system from Valve 3 inch &

corrosion products. 4 inch Stainless Stool Rev. I

b. Corrosion of C.S. seat.

Loss of leak tightness and check valve function. l l

Ravicion: 1 11/25/85 Page 6 of 6 ITEM NUMBER VII.b.2 CENERIC SAFETY CONSEQUENCES ANALYSIS EM DESCRIPTION SAFETY PRESSURE POTENTIAL POTENTIAL CLASS RATING REASSEMBLY ERROR FAILURE & EFFECTS Borg-Warner 2 300 psi

1. Bonnet assembly from 150 psi

1. Failure of bonnet seal and/or bonnet Swing Check valve.

cover. External leakage. Vcive 10 inch Stainless Steel

2. Bonnet assembly from C.S.

2. a. Corrosion of bonnet. Potential-failure valve.

of bonnet. Contamination of system from corrosion products,

b. Corrosion of C.S. seat.

Loss of leak tightness and check valve function.

c. Corrosion and failure of C.S. disk.

B:rg-Warner 2 150 psi

1. Bonnet assembly from C.S.

1. a. Corrosion and potential failure of Swing Check valve.

bonnet. Contamination of system from Vcive 10 inch corrosion products. Stcinless Steel

b. Corrosion of C.S. neat.

Loss of leak tightness and check valve function.

c. Corrosion and failure of C.S. disk.

v. 1 05/23/85

L I 4 ERCI systems integration and Management Corporation = QA/QC-RT-1638 March 13, 1986 Mr. Frank Milliken ITT-Grinnell Valve Co., Inc. P. O. Box 6164 Lancaster, PA. 17603-2064

Dear Frank:

Enclosed please find a Record of Telephone Conversation for our telecon of March 13, 1986. Please review it for correctness and completeness. Please advise me of any coments at (817) 897-8962. If you have no comments, please note your concurrence (inical and date) and return a copy in the enclosed addressed envelope. s ss x ' fife Obert ERC c/o Texas Utilities Generating Co. Comanche Peak Steam Electric Station P. O. Box 1002 Glen Rose, Texas ~6043 y e p

RECORD OF TELEPHO?iE C0f4VERSATIO.i PAGE 1 0F 1 INCOMING OUTGOING X TIME 10:30 A.M. M. DATE. March 13. 1986 Person called: Frnnk Millike

Title:

0A Division Manager Representing: ITT-Grinnel Tel. ( 712) 291-1901 Person Celling: Mike Obert @ -

Title:

ISAP VII.b.2 Issue Coordinator Representing: ERc Tel. (871) 897-8962 Other Parties Involved: None TOPICS REF. ITEM 1. I discussed with Mr. Milliken the differences between the bonnet assemblies of an ASME diaphragm valve and a non-ASME diaphragm valve. He stated the differences are as follows: The castings used for making the bonnets are purchased from the foundry by ITT-G under differe.t specifications. For ASME valves, an ASME material spec is used and for non-ASME valves an ASTM spec. is used. The sama pattern is used for the castings of both ASME and non-ASME bonnets. The only difference in the castings is the paperwork that accompanies them. The chemical and physical properties of the metal required by the ASME material spec e,re the same properties specified in the ASRi material spec. The machining of the bonnets for both ASME and non-ASME bonnets is essentially the same. Again the only differences are in paperwork. There is more QA involvement in the repair of any defects found in ASME bonnets. The post manufccturing Non Destructive Examination program is the same for both ASME and non-ASME bonnets so it is not any more likely that a non-ASME valve bonnet with an undetected defect be shipped than an ASME valve bonnet. 1 2. It is a correct conclusion that there is no functional difference ,betweeh an ASME and non-ASME bonnet. They are physically the same with a different " pedigree" or paperwork package. .}}