ML20245C923

Forwards Response to Each Unresolved Item Identified During Integrated Design Insp of ERCW Sys Which Must Be Resolved Before Restart of Unit 2,per 871009 Request.Plant Design Sound & Addl Independent Reviews Not Warranted

ML20245C923
Person / Time
Site:	Sequoyah
Issue date:	10/29/1987
From:	White S TENNESSEE VALLEY AUTHORITY
To:	James Keppler NRC OFFICE OF SPECIAL PROJECTS
References
NUDOCS 8711040210
Download: ML20245C923 (136)

Text

,.- Dec1}

ps~

TENNESSEE VALLEY AUTHORITY ,

CH ATT ANOOG A, TENN ESSEE 37401 4

6N 38A Lookout Place

}

00T 291987 Mr. James G. Keppler, Director Office of Special Projects U.S. Nuclear Regulatory Commission 4350 East-West Highway EWW 322 Bethesda, Maryland 20814

Dear Mr. Keppler:

In the Matter of ) Docket Nos. 50-327 Tennessee Valley Authority ) 50-328 Your letter dated October 9, 1987, provided TVA with descriptions of preliminary deficiencies and unresolved items identified during the Integrated Design Inspection (IDI) of the Essential Raw Cooling Water (ERCW) System which 4 in your judgement must be resolved before restart of Sequoyah unit 2.

In response to your October 9, 1987 letter, I am enclosing TVA's response to each item from that letter. Commitments made in these responses and/or any revisions which may develop during the followup NRC inspection the week of November 2,1987 in Knoxville, Tennessee, will be formally documented in TVA's response to the IDI inspection report. As discussed with the NRC IDI team on October 23, 1987, TVA does not agree with the staff's categorization of some of these issues as prerequisites to unit 2 restart. After the inspection any remaining areas of disagreement with the staff or the scope of IDI restart issues will be pursued by TVA before our final response to the IDI inspection report.

The TVA responses provided herein continue to support my position that the Sequoyah design is fundamentally sound and that additional independent reviews recommended by your staff in the civil / structural area are not warranted. TVA looks forward to the NRC followup inspection to resolve these matters and the remainder of the IDI issues.

Very truly yours, TENNESSEE VALLEY AUTHORITY CA (S.A. White Manager of Nuclear Power Enclosure cc: See page 2 8711040210 871029 PDR  %

8 G

ADOCK 05000327 \

PDR An Equal opportunity Employer

f[. ; '

.i y Mr. James G. Keppler- QCI[2hlh87 1:

cc (Enclosure):

-Mr. G.-G. Zech, Assistant Director-for Inspection Programs Office of Special Projects U.S. Nuclear Regulatory Commission Region II 101 Marietta Street, NW, Suite 2900 Atlanta, Georgia 30323 Mr. J. A. Zwolinski, Assistant Director for Projects Division of TVA Projects Office of Special Projects U.S. Nuclear Regulatory Commission 4350 East-West Highway EHW 322 Bethesda, Maryland 20814 Sequoyah Resident Inspector.

1Sequoyah Nuclear Plant 2600 Igou Ferry Road Soddy Daisy, Tennessee 37379 I

j i

-s .

I ITEM NO.: D2.2-1 l TITLE: Design Pressure of ERCH System i

SUMMARY

OF ITEM:

1. Calculation B44 870326 006 is nonconservative'in that it does not address

-the effect of components which may be isolated for maintenance.

2. The calculation is also nonconservative in that it does not use maximum flood levels in evaluating static' head or discuss protection against overpressurization when river level rises above normal maximum elevation 683 feet.

.3. The 150 lb/in*g design pressure recommended in calculation 844 870326 006 is inconsistent with the 160 lb/in g2 value in Design Criteria SQN-DC-V-7.4 R2.

4. The governing code, ANSI B31.1-1967, does not allow throttling to reduce pressure without installing an overpressure relief valve.

CLASSIFICATION: Documentation

RESPONSE

This finding will be addressed by first making some general coments regarding code requirements and then responding to each of the four parts defined above.

General Comments: The-applicable design code is ANSI B31.1-1967. Paragraph  !

101.2.2 of B31.1 states the following requirements:

The internal design pressure shall not be less than the maximum i sustained fluid operating pressure within the piping, and shall include allowance for pressure surges, except as these conditions may be modified by the provisions of paragraphs- 102.2.4, 102.2.5,

'102.3.2, and 102.3.3.

TVA interprets the above statement as requiring that design pressure equal or l exceed the maximum anticipated system working pressure during all modes of l normal plant operation, which would cover the period from plant startup to l plant shutdown.

]

There has been a natural progression over time in code requirements related to  :

pressure' components. For piping systems, this progression started with ANSI B31.1-1967, followed by ANSI B31.7, ending with ASME Section III. In general, each new code did not negate the design efforts of a previous code, although more detailed and stringent requirements related to some components were added. Therefore, although SQN is not an ASME Code plant, TVA believes it enlightening to review the ASME requirements for design pressure.

l ZTEM NO.: D2.2-1 (Continued)

TITLE: Design Pressure of ERCH System (Continued)

RESPONSE: (Continued)

Paragraph ND-3112.1 of the 1974 ASME Code states:

Components shall be designed for at least the most severe condition of coincident pressure and temperature expected in normal operation.

This reinforces TVA's B31.1 interpretation that only normal operating .

conditions should be considered in selecting design pressure.

Specifically, TVA considers isolation of safety-related equipment for maintenance or operation during floods as abnormal conditions which are not appropriate considerations for determining design pressure. NRC stated that.

the time restraint for abnormal operating conditions was clarified in the 1976 l Addendum to ANSI B31.1 to be based on a 24-hour operating period. However, 1 ANSI B31.1-1967, which is the code of record (COR) for SQN, does not specify a quantitative time restraint but defines it qualitatively as "occassional operation for short periods." Furthermore, Section ND-3600 of the ASME B&PV 1974 Code also does not specify quantitative time restraints for abnormal operation. In fact, the use of B31.1-1967 and ASME B&PV Code Section III for  !

the ERCH system would result in the same design and abnormal condition l pressures. The only difference would be the separation of abnormal conditions into more precise categories (i.e., upset, emergency, and faulted) using Section III versus the general abnormal condition category using B31.1.

TVA believes that an addendum to B31.1 issued in 1976, some nine years later than the COR, cannot reasonably be considered an interpretation, but rather new requirements. Furthermore, it is our understanding that the 1976 Code addenda has never been adopted in 10 CFR 50.55a for use in nuclear power plant design.

Therefore, TVA finds no basis for applying rules from a design code never ,

invoked or applied to SQN and apparently never used for nuclear plant design i when the Code of Record which preceded it and the ASME Code which followed it would support the present design. Consequently, TVA and its Code consultants agree that the appropriate operating period for abnormal operation is from startup to shutdown.

Specific findings: Items 1 and 2 are attributed to differing interpretations of normal versus abnormal operating conditions by TVA and NRC. As discussed in the General Comments, TVA considers isolation of safety-related equipment for maintenance or operation during floods as abnormal conditions which are not' appropriate considerations for determining design pressure. TVA does agree that an analysis of abnormal ERCH operating conditions, including flood and maintenance isolation, demonstrating compliance with the associated ANSI B31.1 requirements is needed. Although not available during the initial IDI review, calculation 844 has now been issued and does

_ ._ _ _ _ _ _ _ _ _ - _ \

l l

)

l ITEM NO.: D2.2-1 (Continued)

TITLE: Design Pressure of ERCW System (Continued) l l

RESPONSE: (Continued) j demonstrate compliance with the B31.1, Paragraph 102.2.4 requirements for abnormal operation. Furthermore, TVA commits to investigate upgrading the ERCH design to ND-3600 of ASME Section III to allow a clearer definition of j what TVA considers to be abnormal, upset, emergency, or faulted plant {

conditions. TVA will complete this investigation by July 1988. )

l The explanation of item 3 requires an understanding of the chronology of activities associated with ERCH design pressure calculation B44 870326 006.

TVA had identified a mismatch between the ERCH system design pressure and the design pressure of certain components before the IDI. This was documented in SCR No. SQNMEB8623 R2. To determine if working pressures actually exceeded <

component design pressure, TVA prepared calculation 844 870326 006 which found l a potential overpressurization problem in the following ERCH loads.

Auxiliary Control Air Compressors  !

Station Air Compressors Electric Board Room A/C Centrifugal Charging Pump 011 Cooler Safety Injection Pump 011 Cooler Corrective action was identified in SCR SQNMEB8623 R2. Action associated with the compressors is discussed in D2.2-3 and D2.2-4. For the latter three items, the working pressure was calculated to exceed the component design pressure by 2.1 lb/in 2g if two pumps should be operating during periods of low system flow demand. (This is now considered an abnormal condition, as discussed in D2.2-2.) To preclude this slight overpressurization, part of the corrective action recommended in SCR SQNMEB8623 R2 was to incorporate a precaution on ERCH system operating instruction SOI-67.1 to limit ERCH pump discharge pressure to 124 lb/in g. 2 The procedure had been drafted but not issued at the time of the IDI.

The final part of the corrective action recommended in the SCR was to lower the design pressure in the 160 lb/in*g parts of the system to 150 lb/in'g. Because this part of the corrective action was a documentation ,

change only, this was to be accomplished after restart under ECN-L7125.

Therefore, during the IDI, this was an apparent discrepancy. The documentation changes are currently scheduled to be completed by April 1988.

Also refer to TVA's response to findings D2.2-2, D2.2-3, and 02.2-4.

l XTEM NO.: D2.2-1 (Continued)

L TITLE: Design Pressure of ERCW System (Continued) l RESPONSE: (Continued)

Lastly, item 4 was caused by a misunderstanding of TVA's intent regarding operator actions to maintain pump discharge pressure below 124 lb/in*g. As noted above, the revised operating instruction had not been issued at the time the IDI review was performed. However, the draft procedure did show that ERCH pumps would be turned off or the flowrate through the CCS heat exchangers increased should pressure exceed 124 lb/in*g. It was not TVA's latent to throttle the pump discharge valves; therefore, the overpressure protection requirements in paragraph 102.2.5(b) of ANSI B31.1 did not apply except to the air compressor situations discussed in D2.2-3 and D2.2-4 where throttling was employed for pressure control.

However, questions raised during the IDI prompted TVA to scrutinize further the assumptions and methods used in the ERCH design pressure calculation. TVA has concluded that the " unusual condition" which created the apparent overpressure condition in the latter three coolers listed above is not a normal operating condition and is therefore inappropriate as a design pressure consideration. Calculation B44 supersedes the document originallyreviewedbytheIDIteamandshowsthatworkingpressuresarebelow 150 lb/in g. Therefore, the operating instruction will not be required to assure Code compliance, although TVA will issue it to clarify 124 lb/in 2g as the upper bound for normal pump operation. Refer also to D2.2-2.

Generic Review: As noted earlier, SCR SQNMEB8623 identified a discrepancy  !

between the ERCH system design pressure and the design pressure of certain components. TVA then initiated a generic review of this problem. Results of the generic review for SQN are documented in memorandum 825 861023 022 which references PIRs SQNMEB8657 and SQNNEB8611. These PIRs identify similar problems for the raw service water system and the chemical and volume control system. Tnus, the generic review for the system / components design pressure mismatch has been completed.

Programmatic Issues: To-identify programmatic weaknesses, it is first necessary to determine the root cause of the perceived problems. Although several issues were raised under this topic, TVA believes the following statements summarize NRC's specific findings:

A. Some SQN ERCH users have design pressures less than system design pressure.

B. THe SQN ERCH design pressure calculation is nonconservative .

C. NRC disagrees with TVA's corrective action for SCR SQNMEB8623 R2 4 regarding the need for relief valves and the use of administrative procedures.

l l

'1

_4_

\

l ITEM NO.: D2.2-1 (Continued) .

TITLE: Design Pressure of ERCH System (Continued)

RESPONSE: (Continued) I Most equipment served by ERCW was procured before 1973, whereas the initial t

issue of procedures dealing with design review (EP-4.04) and design verification (EP-3.10) were in September 1973 and August 1974, respectively.

- Therefore, differences similar to item A would not occur under TVA's current.

QA program. Nonetheless, since a number of deficiencies have been identified in this area, TVA identified an adverse trend under SCR GENNEB8603 (B44 860623

. 003), and because of this, has taken further corrective action beyond reliance on current programs. TVA has committed to incorporate the'following statement in mechanical standard procurement specifications:

Design pressure and temperature must be specified in all system and component interfaces. Refer to the appropriate system design criteria for pressure and temperature requirements.

TVA believes this action, along with the design control programs currently in l place, will avoid this type problem in the future. 2 Under item B, TVA disagrees that the consideration of low probability events such as abnormally high water levels and maintenance isolation of equipment for extended periods are appropriate considerations for design pressure.

However, we do agree that such effects should be considered and appropriately categorized to verify code compliance. Therefore, TVA commits to include

" abnormal operation" in the list of' required calculations in MEB EP-23.2.

Item C was caused by NRC's perception that TVA intended to throttle the pump discharge valves to control system design pressure, thereby requiring relief valves for overpressure protection. No throttling of pump discharge valves for this purpose will be performed. Therefore, no programmatic issues exist for this item.

REFERENCES:

1. Calculation B44 870326 006, ERCW Design Pressure, Revision 1 2 .' Calculation B44 . ERCH Design Pressure, Revision 2

3. Calculation B44 , ERCW Abnormal Pressure Conditions, Revision 1

4. SCR SQNMEB8623 R2 (B25 87040 075)

5. System Operating Instruction 501-67.1, ERCW System

6. ERCW System Design Criteria Document, SQN-DC-V-7.4 R2

7. ANSI B31.1-1967

8. ASME Boller and Pressure Vessel Code, Subsection ND, 1974

9. ECN-L7125

10. PIR SQNMEB8657

11. PIR SQNMEB8611

12. Memorandum B25 861023 022

13. EP-3.10 14 EP-4.04

15. SCR GENNEB8603 (B44 860623 003) 16 MEB EP-23.2

l l

ITEM NO. D2.2-2 TITLE: Procedure Not Available to Limit ERCH Pump Discharge Pressure

SUMMARY

OF ITEM:

Design pressure calculation B44 870326 006 identified the potential overpressurization of the following ERCH loads:

Electric Board Room Air-Conditioner Centrifugal Charging Pump 011 Cooler-18-B Turbine Driver Auxiliary Feedwater Pump Safety Injection Pump 011 Cooler The calculation recommended administrative 1y limiting ERCH pump operation to a max,imum discharge pressure of 124 lb/in*g to prevent overpressurizing these components. However, as of March 1987, no such procedure had been approved.

A DNE memorandum dated July 31, 1987, had been sent from DNE to Operations requesting a procedure revision. This demonstrates lack of timeliness in issuing procedures associated with identified design deficiencies.

CLASSIFICATION: Documentation

RESPONSE

TVA had identified the mismatch between ERCH system design pressure and component design pressure before the IDI. This was documented in SCR SQNMEB8623 R2. To determine if working pressures actually exceeded component design pressure, TVA prepared calculation B44 870326 006 which found j a potential overpressurization problem in the following ERCH loads.

Auxiliary Control Air Compressors Station Air Compressors 1 Electric Board Room A/C '

Centrifugal Charging Pump Oil Cooler Safety Injection Pump 011 Cooler (Note: The pressure at the normally closed AFH pump interface valve did exceed 150 lb/in8 g slightly, but the pressure at the AFW pump would be less than 150 lb/in* when ERCH is actually supplying flow.)

The compressors are discussed in D2.2-3 and D2.2-4. For the latter three items, the working pressure was calculated to exceed the component design pressure by up to 2.1 lb/in gz if two pumps should be operating during periods of low system flow demand. This was considered an unusual operating condition, but was analyzed in the calculation nonetheless. To preclude this slight overpressurization, part of the corrective action recommended in SCR SQNMEB8623 R2 was to incorporate a precaution in ERCW System Operating Instruction 501-67.1 to limit ERCW pump discharge pressure to 124 lb/in g.

TTEM NO. D2.2-2 (Continued) I d

TITLE: Procedure Not Available to Limit ERCH Pump Discharge Pressure l (Continued)

RESPONSE: (Continued) 1 The procedure had been drafted but not issued at the time of the IDI. An )

expedited procedure revision was not considered necessary due to the small overpressure magnitudes involved and the unusual nature of the initiating condition.

Questions raised during the IDI have prompted TVA to scrutinize further the assumptions and methods used in the ERCW design pressure calculation. TVA has concluded that the " unusual condition" which created the apparent overpressure condition in the latter three coolers listed above is not normal operating condition and is therefore inappropriate as a design pressure consideration.

Calculation supersedes the document originally reviewed by the IDI team. The current revision shows that only the compressors will be  ;

overpressurized.

Nevertheless, TVA will proceed with the SOI-67.1 revision to clearly establish 124 lb/in g2 as the upper bound for normal pump operation. As stated in the draft procedure, the operator will either turn off operating ERCH pumps or increase the flowrate through the CCS heat exchangers to control pressure; he will not throttle the pump discharge valves.

ANSI-831.1 recognizes the need to occasionally operate above system design pressure for short time periods and defines rules for such abnormal operation in paragraph 102.2.4. Although not available at the time the IDI was performed, calculation B44 has now been issued and does demonstrate compliance with B31.1 requirements for abnormal operation. Also refer to D2.2-1.

In summary, the 501-67.1 procedure had not been issued before the IDI review because it was not considered critical during this preiod since the plant was already in shutdown more and the magnitude and probability of overpressurization were small. Further review and analysis by TVA have concluded that no overpressure condition exists (except for the compressors) during normal operation and the design complies with the B31.1 rules for abnormal condition operation. Nonetheless, TVA will issue the procedure to establish the upper bound for normal pump operation.

REFERENCES:

1. Calculation 844 870326 006, ERCW Design Pressure, Revision 1.

2. Calculation 844 , ERCH Design Pressure, Revision 2.

3. SCR SQNMEB8623 R2 (B25 870409 075).

4. System Operating Instruction 501-67.1, ERCW System.

5. DNE memorandum July 31, 1987 (B44 870731 014)

6. Calculation B44 , ERCW Abnormal Pressure Conditions, Revision 1 j

7. ANSI B31.1-1967 4

l ITEM NO.: D2.2-3 TITLE: Overpressurization of the Auxiliary Control Air Compressor Coolers

}

SUMMARY

OF ITEM:

Design pressure calculation B44 870326 006 states that the safety z related auxiliary control air compressor's design pressure is 67 lb/in g in the l cylinder cooling jacket and 75 lb/in g in the aftercooler, whereas the 2 recommended ERCH system design pressure is 160 lb/in g. No approved l procedure of methodology has been developed to prevent overpressurization of j the cylinder jacket or aftercooler. A memorandum from DNE to Operations had requested that throttling be employed as a means of overpressure protection.

However, throttling is not in compliance with the governing code (ANSI B31.1) or ASME Section III as an acceptable means of overpressure protection.

CLASSIFICATION: Design Deficiency

RESPONSE

Calculation 844 870326 006 found that the component design pressure of several ERCH users, including the auxiliary control air compressors, was less than anticipated normal working pressures, thereby violating ANSI-831.1, paragraph 101.2.2. TVA had identified and documented this condition in SCR SQNMEB8623 R2 (B25 870409 075) before the IDI.

Corrective action for the ACA compressor overpressure condition will be to use the compressor inlet valve to throttle flow and reduce pressure. Calculation B44 870304 003 determined the design requirements for the valve adjustment, which has already been completed in the field. Although valve positions are verified periodically by surveillance tests, relief valves will also be added to prevent overpressurization, as required in ANSI-B31.1 paragraph 102.2.5(b),

if the inlet or outlet valves should be inadvertently opened or closed (Reference ECN-L7297).

The July 31, 1987, DNE memorandum referenced in the description of the finding was not related to the ACA compressors. For a discussion of this issue, see TVA's response to 02.2-2.

Corrective action for the ACA overpressure condition, as well as other safety-related ERCH components, will be tracked under SCR SQNMEB8623 R2 and completed before restart. Corrective action for the station air compressors will be tracked under PIR SQNMEB87107. Also see responses to IDI findings D2.2-2 and D2.2-4.

Refer to D2.2-1 for a discussion of generic and programmatic issues associated with this finding.

REFERENCES:

1. ERCH System Design Pressure Calculation (B44 870326 006)

2. SCR SQNMEB8623 R2 (B25 870409 075)

3. ECN-L7297

(<

i ITEM NO.: D2.2 H

/

TITLE: Overpressurization of Station Air Compressor Coolers

SUMMARY

OF ITEM:

Design pressure calculation 844 870326 006 states that the nonsafety related station air compressor's design pressure is 67 lb/in*g in the cylinder a cooling jacket, 50 lb/in*g in the intercooler and 150 lb/in*g in the n aftercooler whereas the ERCH system design pressure is 160 lb/in g. No i procedure or methodology has been developed to prevent overpressurization.

CLASSIFICATION: Design Deficiency

RESPONSE

Calculation B44 870326 006 found that the component design pressure of several ERCW users, including the station air compressors, was less than anticipated normal working pressures, thereby violating ANSI-831.1, paragraph 101.2.2.

TVA had identified and documented this condition in SCR SQNMEB8623R2 (B25 870409 075) before the IDI. PIR SQNMEB87107 (B25 870812 103) was written later to focus the appropriately. lower level of attention on the nonsafety-related station air compressors.

Because the station air compressors were not safety-related, PIR SQNMEB87107 did not meet the SQN restart criteria and corrective action had not been .

determined at the time of the IDI. Furthermore, Calculation B44 070824 006 has shown that a critical crack in the nonseismic station air compressor piping could be detected and isolated without affecting performance of safety-related areas. Nonetheless, TVA has now decided to correct the overpressurization problem before restart. Corrective actions will be to- '.

control the inlet pressure to compressor components to acceptable-levels.and provide relief valves to protect against failure or inadvertent operation of g the pressure control features (Reference ECN-L7294). -  !

Required corrective action for other ERCW users will be tracked under SCR SQNMEB8623 R2. Also see responses to IDI findings 02.2-1, D2.2-2, and D2.2-3.

Generic and programmatic issues associated with this finding are discussed in D2.2-1.

REFERENCES:

1. ERCH System Design Pressure Calculation, Revision 1 (844 870325 006) l

2. SCR SQNMEB8623 R2 (B25 870409 075) l

3. PIR SQNMEB87107 (B24 870812 103) 1

4. ANSI B31.1-1967

5. ERCH Critical Crack - Station Air Compressors, Revision 0 (B44 870824 006)

6. ECN-L7294 j

l

l ITEM N0.: 02:2-5 5 TITLE: Evaluation of Failure of ERCH Nonseismic Piping

SUMMARY

LOFITEN:

No analysis had been performed to determine the effects of the 10 minute interval before operator action to isolate the station air compressor break.

l. CLASSIFICATION: Documentation l' RESPONSE:  ;

Calculati'on.B44 870824 006 evaluates a critical crack in the nonseismic portion of the ERCH piping'to the station air compressors. Flow to the safety related components on that loop is not reduced during the initial 10 minute interval before manual isolation.

The appropriate pipt break in the station air compressor loop is a critical crack. In order-to confirm that catastrophic collapse of the turbine building i during a seismic event may be neglected, the ser dt.es of EQE were obtained to assess the SQN turbine' building (Reference NUREG-1030). Their conclusion (B44 I 870916502) is that "no conditions have been identified which can lead to a  !

credible failure that would result in a guillotine break of the piping."

Procedure SOI-55-0F.278-XA-55-278-D instructs the' operator to isolate the ERCW supply to the station air compressors if a high ficw altrm occurs during .,

normal plant conditions. In addition, befort restart, TVA will add provis h, ns in the AOI-9 procedure to isolate the air congressor cooling water on detectioncfhighflowauringearthquakesbasedonseismicallyqualified indicatir.g lights. For long term system improvement, TVA will consider automatic isolation of the ERCW supply to.the. station air compressors by  ;

closing valves 0-FCV-67-205 and -208, on datection of high flow.  !

The root cause of the failure to evaluate the 10 minute interval was the undocumented presumption that calculation MEB 840928008, which addressed manual realign'nent of the CCS heat exchangers, bounded the station air compressor break. A supplement (B05'861222501) to NEP 3.1, Calculations, requires documentation of technicLi judgments made in calculations so that a written basis. for the preparer's thought process will be established.

TVA originally designed ERCW and other systems at SQN in the 1968 to 1972 timeframe wnu little. documentation was available for class break boundary requirements'and TVA had relied on the experience of the design engineers and the limited data that was available at the time in the design for SQN class break boundaries.

t .

t ,

i ms,*  ;

A '- r 1

j

' ITEM NO.:

~

D2.2-5 (Continued)

• i I

&" ' v/ s TITLE:

.', f Evaluation of Failure 4.f YRC'W Nonseismic Piping (Continued) 1 in y ,.* a RESPONSE: (Continued) l TVA'sdesignofclassbreakboundarieshas'notresulted.inanysituations[

whichcouldhaveasignificantimpactonthesafeoperationoftheplantas[t [

4

. presently designed. p ('

TVA will commit to perform a sundy of tte other A/Es to determine the methods {

they used to specify class break' boundary requirements on the nuclear plants ]

of SQN's vintage and what'is currently behlq used. TVA will follow the l current industry practices star.G1rds for.fidure design. . TVA will also review other class break boundary.' interfaces between safety and nonsafety-related systems and examine the appi! cation of single failure criteria to these '

interfaces. The above actipns will be completed by April 1,1988. If the

.above A/E survey identifiednadequacies in TVA practice in designing class O breaks a generic review wil'fbe performed to identify deficiencies at all TVA nuclear plants.

REFERENCES:

i ,

,4

1. Evaluation *of:a Critical Crack in ERCW Piping to the Station Air'

^,

( ' '-

Compressors (B44 870824 006) t n

2. EQE Seismic P.eview qf ERCW Piping Inside the Turbine Building (844 870316

, , : c': H /

502) .

5' '

3. Procedure 501-55-0M276-XA-55-278-D F l

'4. Procedure A0I 9.y n

,- ]

w.j

5. 'NUREG-1030 's ,

\

q s  ? \,

) k

-f i

/

(( g.' i

\ '

j

't , ,.

37 \

,t / f ') [

3 Y '

s  !

-e f~

\

't y T' $

t

, . f,3-)

s 4 3

l ITEM NO.: D2.2-7 i TITLE: Determination of ERCH Pump House Ambient Temperature for Environmental Qualification (Mild) l

SUMMARY

OF ITEM:

l This deficiency addresses a problem associated with establishing the maximum and minimum temperature limits for the ERCH pump station. The calculation (TI-ECS-53) which established the mild environment conditions for SQN was reviewed for the portion that applies to the ERCH pump station.

Review of this calculation identified two problems. First, the maximum temperature limit was not based on technically justified heat gains and  ;

losses, but rather on unjustified assumptions. Secondly, the lower temperature limit.did not assume the worst case failure of non-IE equipment, I namely, room heater off and ventilation fan on. Also, heat losses through the uninsulated concrete walls were incorrectly neglected.

As a result of these concerns, TVA needs to ensure that this entire calculation provides a technically justified basis for establishing the mild environment temperature limits throughout the rest of the plant.

In response to these concerns, TVA provided a more rigorous analysis of the ERCH pump station mild environment temperature limits. However, this newer calculation did not adequately establish the lower temperature limit associated with the LOCA condition due to a number of nonconservative assumptions.

CLASSIFICATION: Minor Calculation Error

RESPONSE

TVA agrees, and as a result will revise calculation TI-ECS-53 to provide a technically justified basis for establishing the mild environment temperature limits in the ERCH pump station and throughout the rest of the plant, where required. Currently, this calculation does not address the effects on mild environment areas caused'by a LOCA/HELB inside primary containment.

Therefore, the environmental drawings (47E235 series) delineate only normal and abnormal, minimum and maximum temperature limits for mild environment areas. As a resuli of this deficiency, TVA will revise calculation TI-ECS-53 ,

(and the 47E235 series drawings) to delineate the LOCA/HELB ( in primary containment effects) conditions for those mild environment areas that are affected by such an occurrence.

Maximum LOCA temperatures for the affected ERCH pump station will be determined in a revision to calculation TI-505 for input to calculation i TI-ECS-53. Minimum LOCA temperatures for the affected ERCH pump station will be determined in a revision to MEB Calculation FSG-HVC-082087 for input to calculation TI-ECS-53. The revisions to TI-ECS-53 TI-505, and FSG-HVC-082087 will (1) delete unverified and nonconservative assumptions, (2) document ,

proper reference material used to establish design temperatures, and (3) include technically justified heat gains and losses. However, the minimum

ITEM NO.: D2.2-7 (Continued)

TITLE: Determination of ERCH Pump House Ambient Temperature for' Environmental

-Qualification (Mild) (Continued)  ;

)

RESPONSE: (Continued) temperature calculation will include credit for.the non-IE heaters. When the current heaters were installed, fully qualified electric heaters were not available; however, the non-IE heaters were supplied with trained IE power.

Therefore, in accordance with the requirements of NRC Generic Letter 8209 (and NUREG-0588 for Category II plants), the existing heaters are adequate for mild environment.IE applications. In adoitton administrative instructions have been established to shut off the fan seasonally when the temperature in the ERCW pump station drops below 65'F (Reference B25 870826 019).

Calculation TI-ECS-53 supports the minimum normal and abnormal temperatures for the mild environment areas in the rest of the plant (and the revision will include the LOCA condition temperatures for affected areas). However, in these areas, design temperature limits were established early in the plant design (as design input documents) based on accepted industry standards and practices, in lieu of being established using conservatively justified, detailed calculation of existing plant configuration. Although calculation TI-ECS-53 is adequate for its intended use and for establishing mild environment conditions, TVA will review the temperature' limits in the mild environment areas of the remainder of the plant and generate calculations to support justifiable minimum temperature limits for only those areas that have minimum temperature operational restraints necessary for safe shutdown, operation, and Chapter 15 accident mitigation (i.e., freeze protection, etc.). There are no mild areas with maximum temperature operational  !

restraints as they have been included in the environmental drawings as harsh environment areas. The TI-ECS-53 revision will be completed postrestart.

In the event that the revised FSG-WVC-082087 calculation results in a minimum temperature less than 32*F, which could adversely impact the safety-related operation of the ERCW pumps, TVA will issue a CAQR to resolve the lack of proper freeze protection. These ERCH pump station calculation revisions are scheduled for completion by November 1,1987.

REFERENCES:

1. DNE Calculation TI-ECS-53, Summary of Mild Environmental Conditions for Sequoyah and Watts Bar Nuclear Plants

2. Environmental Drawings, 47E235-Series

3. DNE Calculation TI-505, ERCH Pumping Station Ventilation '

4. DNE Calculations FSG-HVC-082087, ERCH Pumping Station HVAC Heat Transmission and Heating (B44 870825 012)

5. NUREG-0588 l 6. NRC Generic Letter 82-09

7. TVA Memorandum, Hosmer to Nobles dated August 26, 1987 (B25 870826 019)

I

JTEM NO.: D2.3-3 TITLE: -ERCH Screenwash Pump Not Included in TVA ASME XI Pump Inservice Test Plan l'

SUMMARY

OF ITEM:

This item addresses the omission of.the essential raw cooling water (ERCW) traveling screen wash pumps from the ASME Section XI pump test program. These pumps are listed in the SQN Final Safety Analysis Report (FSAR) as TVA Class C.

CLASSIFICATION: Valid Deficiency i

RESPONSE

In.the design basis analysis of the ERCH system, the screen wash pumps were assumed to be operable in all modes of system operation to prevent a drop in pump suction pressure or more than 8 inches of water due to a flow restriction across the traveling screens. This assumption is not specifically stated in the SQN FSAR or the system design criteria. The basis for determining which '

pumps are to be included in the Section XI Inservice Testing (IST) Program are the plant FSAR and the system design criteria. Based on the Inservice Inspection Program Section being unaware of this design assumption, and informal discussions with other nuclear Utilities, the screen wash pumps were '

omitted from the IST Program. i To address this item, the assumption of operability of the screen wash pumps during all modes of operation will be added to the SQN FSAR and the ERCW traveling screen wash pumps will be added to the SQN IST Program.

THE IST Program has been reviewed by both TVA and contractor personnel to ,

determine which pumps are required for inclusion. All pumps in TVA Class A, B, C, or D, as well as selected lower class pumps, were reviewed. Those pumps which are within the scope of ASME Section XI, and are required to perform a specific function in shutting down the reactor or in mitigating the consequences of an accido , and are provided with an emergency power source are included in the pump test program. The FSAR and system design criteria l documents provided information for this determination.

REFERENCES:

1. ASME Boiler and Pressure Vessel Code,Section XI  ;

2. 10 CFR S0.55a

XTEM NO.: D2.5-2 TITLE: Kerotest Packless Y-Pattern Valves Used for Throttling l

SUMMARY

OF ITEM:

Procedural control is needed to prevent Kerotest packless metal diaphragm l design globe valves from being used for throttling flow in ranges outside of manufacturer's area of recommended use. The full extent of the misapplication i of these valves needs to be established.

CLASSIFICATION: Deficient Design

RESPONSE

A survey was made on unit 2 ERCH system for these type globe valves in a throttling application. Four of these valves were found and identified to '

NRC. All.four valves were within the manufacturer's acceptable throttling region.  !

It was determined that TVA personnel used these type valves outside the specified application. The designer selected packless metal diaphragm Kerotest globe valves for throttling application because:

1. He was unaware of the unique limitation on throttling of these valves.

2. He failed to review the specification and depended on the bill of materials which did not identify these valves to be used only for shutoff or drains.

3. The vendor supplied these valves before issue of 10 CFR Part 21 so no notification was supplied by the vendor for the limitation.

In endeavoring to establish the full extent of any possible misapplication of these manual 2-inch and under Kerotest packless metal diaphragm globe valves used in a throttling application, TVA personnel will perform a review of all other safety-related systems at SQN. The review intent will be.to locate all manual Kerotest packless metal diaphragm valves used in a throttling application. A second review for flow requirements will be performed on the located valves (Reference 1). Any valves found to be in the acceptable flow application will have their specifications reviewed, and if required, revised.

A notice-has been sent to all other TVA plants to notify them of the specific limitations of the Kerotest packless metal diaphragm globe valves and to initiate a survey. Misapplication will be documented by a CAQR (Reference 3). Control of existing valve applications will be provided by the addition of a note to the applicable surveillance instructions to caution operators not to throttle Kerotest valves into unacceptable regions (Reference 4).

i

-ITEM NO : D2.5-2 (Continued)

TITLE: Kerotest Packless Y-Pattern Valves Used for Throttling RESPONSE: (Continued) i Future applications of this type valve will be limited by a notification to be sent to all mechanical design groups to' caution designers about the unusual l limitations of this type.of valve and not to use components outside of their specified application (Reference 5).

The survey and identification of any CAQRs at SQN will be accomplished before restart (October 15, 1987). The addition of notes to the Surveillance Instructions, notification to designers, and the reports of findings from the other plants will be postrestart (April 1, 1988). i

REFERENCES:

l

1. Memorandum to SQN System Engineers from R. Daniels dated  !

September 16, 1987, " Corrective Action - NRC AI No. D2.18 - Kerotest  !

Valves"  !

2. CAQR SQP871490 (S13'871002 812) - Kerotest Valves  !

3. C. A. Chandley's memorandum to Those listed dated September.18, 1987 (B44 870918 004), "NRC D2-19 (Deficiency) - Kerotest Valve Misapplication -

Corrective Action - All Plants"

4. Memorandum to Operations (Later)

5. C. A. Chandley's memorandum to Those listed (All Lead Mechanical l Engineers) dated September 18, 1987 (B44 870918 004), "NRC Deficiency D2.19 - Kerotest Valve - Corrective Action " l l

0

ZTEM NO.: D2.5-3 1 l

TITLE: Environmental Qualification (Mild) ERCH Pump House Components

SUMMARY

OF ITEM:

This item addresses that a discrepancy exists between the Environmental Requirements in the Specifications for the ERCW Pumps Strainers, Traveling Screens and Screenwash Pumps and the Environmental Data Drawings (47E235-35 &

47E235-38). The maximum temperatures listed on the Environmental Data Drawings exceed the Specification Requirements. The TVA Design Documents '

(Environmental Data Drawings, Equipment Specifications and Equipment Qualification Documentation) should be consistent and the temperature values should be justified.

CLASSIFICATION: No Deficiency

RESPONSE

TVA disagrees. The attached table should serve to clarify that TVA drawings and documents are consistent and the temperatures justified. Although the screen wash pump motors were specified for a maximum temperature of 104*F, motors were provided by the vendor qualified for 122*F, which exceeds the '

current maximum design temperature of 120*F. This insignificant inconsistency (Design: 120*F, Spec: 104*F) exists because the design temperature was revised in 1981 when the Environmental (47E235-series) Drawings were issued as a -

NUREG-0588 (later 10 CFR 50.49) product. The specification was issued in 1976. TVA verified the adequacy of the motor, but elected not to change the specification because the contract was complete.

Additionally, it should be noted that this equipment is located in a mild plant area and that the subject equipment was procured in the mid 1970's time frame. NRC Generic Letter 82-09 dated April 20,1982 states that for existing equipment located in a mild environment, Environmental Qualification (EQ) can be adequately demonstrated and maintained by periodic maintenance, testing and surveillance programs. Thus, there is no requirement to backfit a formal EQ documentation program for equipment like that cited in this deficiency, and no further action is deemed necessary.

The motors are quallfled for the maximum normal temperature in lieu of the maximum abnormal temperature which is conservatively postulated to occur less than one percent of plant life. However, TVA Calculation TI-505' Revision i establishes a LOCA maximum temperature of 120 F for the affected areas, separate from the maximum abnormal temperature (previously listed together).

Drawing 47E235-38 will be revised to delineate the new temperature. See also the response to D2.2-7 for additional details involving environmental parameters (Normal, Abnormal and LOCA Temperatures).

TVA has reviewed the motor data and found that all these affected motors are qualified for an effective ambient temperature rating of 50*C (122*F) whicn is greater than the LOCA and normal maximum. (Motors are Class B insulated with ambient rating of 50*C and Class F & H insulated with Class B insulation temperature rise giving an effective ambient rating greater than 50*C.) Sin:e the motors provide their own inherent ambient temperature, minimum qualified temperature is meaningless and therefore not quantified or required.

ITEM NO.: D2.5-3 (Continued)  :

TITLE: Discrepancy in Environmental Temperatures Between TVA Documents for ERCH Pumping Station (Continued) ,

This item has no impact on safety and no physical changes are required. l l

REFERENCES:

1. NRC Generic Letter 82-09 dated April 20, 1982 (A27 820427 001).

2. DNE Calculation TI-505 R1 (B45 870904 235).

3. Drawing 47E235-34, -35, and -38 l

l

1 l

COMPARISON OF EQUfPMENT AND ENVIRONMENTAL TEMPERATURES Area EL 690 EL 704 Mechanical Electrical Pump Deck Equip Rm Equip Rm

. Environmental Data Drawing 47E235-35 47E235-34 47E235-38 Temperature Maximum 104*F- 104*F 120*F Temperature Minimum 65*F 65'F 10*F Physical Drawings 37W206-3 37H206-2 37W206-1 ERCH PUMP MOTOR Spec Temperature Maximum 123*F Spec Temperature Minimum -20*F

'Oualified Maximum 122*F +

Qualified Minimum None required TRAVELING SCREEN MOTOR Spec Temperature Maximum 120*F Spec Temperature Minimum 10*F Qualified Maximum 122*F +

Qualified Minimum None required ERCH SCREEN WASH PUMP MOTOR Spec Temperature Maximum 104*F Spec Temperature Minimum -20*F i Qualified Maximum 122*F Qualified Minimum None required l ERCW STRAINERS Spec Temperature Maximum 104*F Spec Temperature Minimum 60*F Qualified Maximum 122*F +

Qualified Minimum None required NOTES: 1. Motors provide own ambient minimum temperature; therefore minimum

' qualified temperature is the specified insulation class ambient.

2. 122*F + represent Class H insulation with specified Class B temperature rise.

l l

l t

l

. ! TEM NO.: D2.5-4

~

TITLE: Inadequate Substantiation of-Procedure for ERCW Screenwash Pump Manual Operation

SUMMARY

OF ITEM:

The ERCW screenwash pumps are designed to be operated automatically when traveling water screen differential pressure is high. The duration of the screenwash cycle is controlled by an automatic timer. This logic provides assurance for cleanliness of traveling water screens and thus assurance of operability of ERCH pumps through provision of adequate suction pressure. The automatic operation of the screenwash pumps was disabled pending replacement of the traveling screen differential pressure instrumentation with a bubbler type.

The modification of traveling water screen differential pressure instrumentation to the " bubbler type" has not been incorporated. A temporary I change, TACF 1-82-258-67 removed the existing wiring and logic for the differential pressure and timer, with a recommendation for procedurally controlled, manual operation of the screenwash pumps. TVA does not have an approved procedure for manual operation of the screenwash pumps since the temporary change was made on October 7, 1982.

FSAR Section 17.lA.5, " Instructions, Procedures and Drawings" requires that activities affecting quality are prescribed by documented instructions in the form of drawings, specifications and procedures. Contrary to this requirement, TVA had no approved procedure in place for manual operation of the screenwash pumps as required by an implemented temporary change notice 5 years ago.

Without screenwash, the. traveling screens could become blocked. This could affect ERCH pump operation by decreasing the available suction head to the pumps. Also, if the difference between the water level across the screens exceeds.several feet, the screens may collapse.

As a result of operating procedures not being issued in a timely manner,  !

" temporary" changes utilized in lieu of design changes may disable system j safety functions without providing compensatory action.

CLASSIFICATION: Documentation ,

1 1

I

__m_______.-____ ._ ..m._ l

ITEM NO.: 02.5-4 TITLE: Inadequate Substantiation of Procedure for ERCW Screenwash Pump Manual Operation (Continued)

RESPONSE

TVA agrees that the appropriate level of approved instructions was not being utilized to perform manual operation of the ERCH traveling screens. System operating instruction 501-67.1 has been revised to incorporate manual operation instructions as required per TACF 1-82-258-67 and subsequent safety evaluations.

This finding has no impact on the safety of the plant, and no hardware or l design changes are required. {

REFERENCE:

1. System Operating Instruction 50I-67.1, Revision 30 l

i i

I l

1 l

l ITEM NO.: 'D3.2-2 ,

TITLE: ERCH Cold Thermal Mode

SUMMARY

OF ITEM:

Background - Section 3.1.1, Normal Functions, of the ERCH design criteria

'(Reference 1) states in part that "ERCH will be supplied to the various heat exchangers with the inlet temperature between 35*F and 83'F."

Description - The TVA operating modes table (Reference 2) does not specify the'

-35'F cold' thermal mode for the ERCW pipe from header 2B to containment spray i heat exchanger 28. Therefore,,the TVA piping analysis (Reference 3) does not ,

analyze the associated piping and supports for the cold thermal mode or thermal range. However, the analysis did consider the containment spray' heat exchanger nozzle thermal displacement due to the cold thermal' mode and range.

Basis - Section 3.1.1 of the FSAR specifies that the range of temperatures for the ERCH supply to the various~ heat exchangers will be between 35*F and 83*F.- Criterion 4 of 10 CFR 50, Appendix A, requires that systems important q to safety shall be' designed to accommodate the effects of and to be compatible with the environmental conditions associated with normal operation. The basis for this deficiency-is.TVA's lack of a cold thermal mode and stress range calculation for the ERCH pipe (a safety system) from header 2B to the containment spray heat exchanger, as required by the ERCH design criteria.

Impact on Design - TVA should review the ERCH operating modes tables to confirm inclusion of the 35"F cold thermal mode.. TVA should review the ERCW piping analyses to confirm that the: cold thermal mode is considered. TVA should revise piping analysis N2-67-2A to include cold thermal mode and a thermal range effects for the ERCH pipe from header 2B to containment spray heat exchanger 28.

REFERENCES:

1. TVA Design Criteria No. SQN-DC-V-7.4, Essential Raw Cooling System (67), Revision 2, July 11, 1986 (B05 860721 505).

2. TVA Drawing No. 47B466-67-22, Insulation and Operation Mode ,

Analysis Data, ERCH Supply Header 2B, Revision 1,  !

November 22, 1985.

3. TVA Piping Analysis N2-67-2A, Revision 7, January 17, 1987 (B25 870123 814).

CLASSIFICATION: Minor Calculation Error  !

i l

\

ITEM NO.: D3.2-2 (Continued)

TITLE: ERCW Cold Thermal Mode (Continued)

RESPONSE

TVA agrees with this deficiency in that the issued OP Mode table did not adequately address thermal conditions required for piping analysis. CAQR .

SQP871496IDI (reference 1) has been written to adcress this situation.

Previous analysis (RI) of n2-67-2A included a. temperature of 30 F for all piping segments and pipe stress and support loads met design allowables (reference 2). Reanalysis performed on R3 (record analysis) incorporated operating modes as shown on the operating mode drawings 478466-67, sheets 20, 21, and 22, which did not include a 35*F temperature for some segments.

However, the procedure for. revision of the support load tables involved revision to loads only when they increase. Since there were no major piping geometry or support modifications involved in the subject piping, no significant change in pipe stress or support design loadings would result. ,

For. piping such as the above, the resulting thermal gradient from ambient (delta-T - 35"F) results in loads and stresses of low magnitude and will not produce a significant increase in pipe stress or support loads.

Since previous analysis ensured that proper thermal stress cmi loadings were evaluated, and the fact that the small delta-T (35"F) produce insignificant results, no immediate corrective action for this analysis is required. This analysis will be reanalyzed to correct this effort as a postrestart item.

Design Guide DG-MS.I.11 offers guidance for generation of analysis operating modes. Paragraphs 4.2.2 and 4.2.3 indicate that all modes which are of a sufficient duration for the piping to reach thermal equilibrium be indicated in the operating modes. This guidance was not followed for the generation of the operating modes in question.

Piping segments which see temperatures.below 70*F and which have issued Op Mode drawings may be affected. NCR SQNCEB 8501 (reference 4) addressed the need for issued Op Mode information. Only reanalysis or new analysis performed after

,lanuary 1, 1985, have issued Op Mode drawings, and only those problems have the potential for being affected by this condition.

Review of all issued 0p Mode drawings and associated piping analyses for piping segments which are subjected to thermal conditions below 70 F will be accomplished as a restart activity.

Mechanical Design Guide M-DG5.1.1 will be upgraded to a design standard (making it a requirement) and emphasis added to ensure that temperatures above and below 70*F be included when applicable, before Mode 2 operation.

REFERENCES:

1. CAQR SQP8714961DI (S13 871002 819)

2. N2-67-2A R1 (CEB 801 007 015) l 3. Mechanical Design Guide M-DG5.1.1

4. NCR SQNCEB 8501 (B41 850312 011)

i ITEM N0,: D3.2-7 TITLE: ERCH Piping Spool Pieces

SUMMARY

OF ITEM:

BACKGROUND A spool piece has to be inserted into the ERCH line from header 2B to the component cooling water surge tank to supply emergency ERCW water to the tank during maximum flood mode. There are also several other spool pieces in the ERCH system which are to be available for rapid installation under certain flood conditions.

DESCRIPTION On September 17, 1986, TVA prepared Drawing Deviation 86DD952 to document discrepancies between the 3-inch diameter 1-foot long spool pieces fabricated ]

for the ERCH piping running from header 2B to the CCS surge tank for units 1 1 and 2 and the as-built dimensions of the installed piping. The 3-inch j diameter blind flanged lines for unit I had horizontal and vertical i misalignments of 2 inches and 2-7/16 inches, respectively, and a j flange-to-flange distance of 11-1/8 inches. The corresponding unit 2 piping had horizontal and vertical misalignments of approximately 3/4 inch and 1/8 inch, and a flange-to-flange distance of 11-3/4 inches. TVA has initiated Work Request No. B129163 to fabricate temporary (non-CSSC) spool pieces until permanent spool pieces are fabricated. It appears that some of the spool pieces were fabricated to the nominal dimensions specified on the piping physical drawings rather than to the as-built dimensions of the installed piping. Their use may induce substantial loading into adjacent supports when installed since the piping needs to be cold sprung to get the flanges to mate.

BASIS Criterion III of 10 CFR 50, Appendix B, " Design Control," states in part that measures shall be established for the control of design interfaces and for verifying or checking the adequacy of design. Contrary to this commitment, TVA did not verify that the spool pieces installed in the ERCH system were fabricated to the as-built dimensions of the installed piping.

IMPACT ON DESIGN System accident-mode function could be impaired due to mismatch between spool pieces and piping.

EXTENT The extent of this condition is not known at this time.

CLASSIFICATION: Design Deficiency

RESPONSE

Twenty-two spool pieces are required to be installed during the SQN units 1 and 2 postulated flooding condition. Fifteen spool pieces are required for unit 2. The spool piece identified in the " Background" section of NRC deficiency D3.2-7 is not required for the maximum flood mode. A Condition Adverse to Quality Report (CAQR) (reference 2) has been written to correct the design flow diagram drawing error.

l l

l l

l l

ITEM NO.: D3.2-7 (Continued) l TITLE: ERCH Piping Spool Pieces (Continued)

Ten of the fifteen unit 2 spool pieces have been checked for fit in place. To check for fit the spool pieces were first bolted on one end. An experienced piping analyst observed the fit and if the analyst determined that no misalignment existed (such that an insignificant additional loading would be induced into the piping system), the second side of the spool piece was bolted in place and hand-tightened with a wrench. If the spool piece was determined not to fit, the misalignment was measured and the spool piece removed.

Five of the ten measured spool pieces were found not to fit. A Condition Adverse to Quality Report (CAQR) (reference 4) and a Potential Reportable Occurrence (PRO) (reference 3) have been initiated to evaluate and document any of the 22 spool pieces found not to fit. Each of the five spool pieces found not to fit was evaluated by the plant maintenance staff responsible for their installation. The action that would have been taken to install the spool pieces should an actual flooding condition exist has been determined and ~

will be evaluated as part of the CAQR and the PRO before the unit 2 mode 4 shakedown.

In addition, before the unit 2 restart:

The remaining 12 spool pieces will be inspected for fit. l

1. ,

2. All spool pieces found not to fit will be modified.

3. If in the remaining 12 spool pieces, any misalignment exists that will induce significant additional loading into the piping and supports, thc piping stresses and support loads will be evaluated.

REFERENCES:

1. QIRCEBSQP870545 (B25 871015 022)

2. CAQR SQP871118IDI (S13 870623 866)

3. PRO 1-87-382

4. CAQR SQP871497IDI (S13 871002 820)

5. Miscellaneous Calculation N2-IDI-SP00L-MISC (Installation of Flood Mode Spool Pieces - Analytical Inspection).

ITEM NO.: D3.2-8 TITLE: Valve Operator Fundamental Frequencies

SUMMARY

OF ITEM:

This item identifies a concern that nonrigid valves were not modeled properly '

in piping analyses problems resulting in the possibility that adjacent pipe support stresses could be under predicted. Specifically, valves 1- and l 2-LCV-70-63 have nonrigid extended structures, but they are modeled as rigid i in piping analysis problem N2-70-39A.

CLASSIFICATION: Documentation

RESPONSE

This is a valid observation. TVA's procurement specification for the LCV-70-63 Masonellan diaphragm control valves permitted acceptance of nonrigid valve extended structures if the valves were qualified by dynamic test. The root cause of this item is that there was no formalized procedure km ,

for communicating which valves were qualified as nonrigid to the piping analyst. The valves themselves were adequately qualified.

CAQR SQF8701691DI has been filed to identify this issue. Based on preliminary results which will be fully documented in reference 3, the area of concern has been bounded to Masonellan valves and diaphragm control valves located in SQN Category I piping systems. This bounding has been achieved through a review of valve frequencies for all Masonellan valves within the DBVP Phase 1 restart scope, an ongoing effort associated with the Hatts Bar Hanger and Analysis Update Program (HAAUP), and a review of natural frequencies for other diaphragm control valves.

The effects of the nonrigid Masonellan valves on attached piping and supports have been preliminarily evaluated with no indication of any adverse effects on the piping or supports. This evaluation included a review of approximately 104 piping analysis problems with Masonellan valves. Twenty-eight of the valves were determined to be flexible and involved 15 piping analysis problems. This evaluation will be documented in reference 4. It should be noted that flexible Grinnell diaphragm control valves were previously evaluated as a result of reference 5. No other flexible valves have been identified.

A nonrigid valve list (reference 6) will be prepared to ensure that the piping analysts are informed of the current list of nonrigid valves. The SQN piping analysis haadbooks will be revised to reference this list.

Fo'r prerestart consideration, the area of concern has been limited to the DBVP Phase i scope boundaries as defined in reference 2.

Postrestart , Masonellan and other diaphragm control valves within the DBVP Phase 2 scope will be addressed in a similar manner as the Phase i scope valves.

\

ITEM NO.: D3.2-8 l

TITLE: Valve Operator Fundamental Frequencies (Continued) l No impact on safety and no hardware changes are anticipated.

REFERENCES:

1. CAQR SQF870169IDI (805 870914 301).  ;

Report CEB-78-3, R3 (B41 870717 003).

2.

3. DNE Calculation later (To be issued by November 6, 1987)

4. DNE Calculation later (To be issued by November 6, 1987)

5. SCR SQNNEB8631 (B45 861110 851)

6. CEB Report 87-13, SQN Nonrigid valves (to be issued by November 6, 1987) l l

1

. - - - _ - n

ITEM NO.: 03.3-1 TITLE: ERCH System Pipe Support Calculations N2-67-2A

SUMMARY

OF ITEM:

Background - The piping stress analyst prepares stress isometrics for each safety-related piping subsystems which detail the types, orientations, and locations of the associated pipe supports. The pipe support calculation for each support design uses the output loads from the piping stress analysis to confirm the adequacy of the support design.

Description - Part B - Pipe supports IERCWH-71 and lERCWH-135 are both supported by a 7-foot 9-inch span of H10X25 structural steel. However, the pipe support calculation for lERCHH-71 which evaluates the W10X25 beam does i not include the support loads due to pipe support IERCHH-134 (References 1, 2).

Part A - Pipe support lERCWH-133 was modeled in the piping analysis as an axial restraint. The pipe support detail sheet shows a lateral clearance of 1/16 inch between the pipe and a vertical piece of 4 inch by 4 inch tubing, which does not appear to be adequate to accommodate the 0.17 inch lateral l motion specified in the pipe support calculation (Reference 3).

Basis - FSAR Table 3.9.2-3 specifies B31.1 as the piping design code of record for TVA safety Class B, C, and D piping and supports. Section 120.1 of ANSI B31.1 notes, in part, that the design of elements for supporting or restraining piping systems, or components thereof, shall be based on all the concurrently acting loads transmitted into the supporting elements.

Impact on Design - Pipe supports lERCHH-71 and -134 are not adequately qualified. Pipe support IERCHH-133 may require modification to provide l adequate lateral clearance. l Extent - The extent of these conditions is not known at this time.

References - 1. TVA Pipe Support Calculation IERCWH-71, Revision 902, January 30, 1986.

2. TVA Pipe Support Calculation lERCWH-134, Revision 902, February 25, 1986.

3. TVA Pipe Support Calculation 1ERCWH-133, Revision 903, February 24, 1986. j 1

CLASSIFICATION: Design Deficiency j l

l

ZTEM NO.: D3.3-1 (Continued) .

TITLE: ERCW System Pipe Support Calculations N2-67-2A (Continued)

RESPONSE

Part A TVA agrees that this is an unanalyzed condition. However, this condition would have been captured by existing TVA programs as described in the following response. This condition is being tracked by CAQR SQP871484 (Reference 6).

The latest rigorous piping analysis of record (Reference 1) has the following movements at node 248, location of hanger IERCWH-133:

+X (in) -X (in)

Deadweight NA -0.0058 TH max 0.0276 NA l SAM sse 0.0252 NA OBE max 0.0829 0.0829 SSE max 0.1174 0.1174 i

For the normal operating condition (DH + Th), the detailed clearances are  ;

adequate: 1

+X - 0.0276" less than 0.1875" (detailed)

-X - 0.0276" less than 0.1875" (detailed)

For the total maximum displacement condition, the gap provided in the -X direction is potentially too.small to accommodate the maximum movement: '

+X - 0.0276 + 0.0252 + 0.1174 - 0.17 less than 0.1875" (detailed)

-X - 0.0058 + 0.1174 - 0.12 greater than 0.0625" (detailed) l Although the movement in the -X direction exceeds the clearance provided by 0.06 inch, this condition does not impair the functionality of the piping system or support. The basis for this conclusion is given in the attached memorandum. Also, the load applied to the support due to the restraint of

. lateral motion is expected to be small compared to the support capacity in X direction. The schedule for ccmpletion of the evaluation of this support to assure compliance to the design basis will be as committed to in the Pipe Support Calculation Regeneration Program.

The root cause of this condition is that at the time this work was done, procedural control or criteria was not adequate.

l

ITEM NO.: D3.3-1 (Continue.d) 1 TITLE: ERCH System Pipe Support Calculations N2-67-2A (Continued)

This condition potentially applies to " boxed-in" type, uni-directional, rigid pipe supports on safety-related piping.

To prevent recurrence, Design Criteria SQN-DC-V-24.2 has been issued, requiring supports on rigorously analyzed piping to be evaluated or designed to accommodate piping movements. SQN-DC-V-24.2 will be amended on the next revision to include supports on-alternately analyzed piping. Pipe support designers have been trained to the requirements of this criteria (Reference 2). i This condition is being generically addressed for supports on. rigorously analyzed piping by the Piping Support Calculation Regeneration Program (Reference 3). Developed as a result of findings by various TVA quality and technical review programs, the Pipe Support Calculation Regeneration Program is evaluating the design basis documentation and technical adequacy of these supports. The program. details have been presented to the NRC (Reference 4).

As required by this program, supports will be evaluated per the requirements of SQN-DC-V-24.2, Support of Rigorously Analyzed Category I Piping. Section 6.14 of this criteria states ". . . supports shall be evaluated to accommodate I all piping movements." Corrective action for pipe supports will be completed as committed in the Pipe Support Regeneration Program. ,

Implementation of the Alternate Analysis P.eview Program (

Reference:

Nuclear '

Performance Plan, Volume 2,Section III, 5.0) will correct this condition generically f or Category I alternately analyzed supports. This program was initiated in early 1986 to resolve potential deficiencies in alternately analyzed piping and supports. These potential deficiencies were previously documented in nonconformance reports SQNSQP8215 and SQNSHP8222. TVA has received a draft SER on this program. As required by this program, each support will be walked down per Special Maintenance Instruction SMI-0-317-55.

Gaps and clearances between the pipe and support structure will be recorded.

The piping system and support will then be evaluated for the as-built condition.' Corrective action will be completed as committed in the Alternate Analysis Review Program.

Part B TVA agrees that the design evaluation of pipe support IERCHH-71 did not ,

address the load applied to hanger IERCHH-71 by hanger IERCHH-134. This condition would have, however, been captured by the ongoing programs for pipe supports. This item is tracked by Reference 6.

. i a

I

I' t

fTEM NO.: D3.'3-1 (Continued)

TITLE: ERCH System Pipe Support Calculations N2-67-2A (Continued)

Pipe hangers lERCWH-71 and ERCHH-134 have been reevaluated considering the combined effects of both hangers. Calculations indicate these supports are in .

compliance with design basis requirements (Reference 7). l This condition is caused by the existing hanger documentation (drawing or calculation) not properly referencing the attachment of additional hangers, resulting in the failure to properly document ECN L6534, L6491, and L5009.

I This concern is generically applicable to " gang" type pipe supports at SQN.

]

l As a result of the Rigorous Pipe Support Regeneration Program and the

! Alternate Analysis Review Program, the calculation documentation for each Category I support will be revised / regenerated to reflect the-results of walkdown information. Future changes will be evaluated considering this l l

issued documentation thereby preventing the recurrence of this condition.

l Also, a Civil Engineering Branch Instruction, CEB-CI.21.90, has been prepared to address the handling of " gang" hangers.

Implementation of the Alternate Analysis Review Program (

Reference:

Nuclear i

i Performance Plan, Volume II,Section III, 5.0) will correct this condition generically for pipe supports on Category I alternately analyzed piping. Each support will be walked down per SMI-0-317-55 and attachments to each " gang" ,

hanger identified. The documentation package and drawings for each hanger j will be revised / regenerated and attachments evaluated. Completion of this program is scheduled as committed in the Alternate Analysis Review Program.

The Pipe Support Calculation Regeneration Program will generically address this condition for supports on rigorously analyzed piping. Each pipe support on rigorously analyzed piping will be walked down to physically verify attachments to " gang" hangers (Reference 5). A new/ revised calculation package will be generated, accounting for and documenting attachments to

" gang" hangers. Corrective action for these pipe supports will be completed as committed in the Pipe Support Calculation Regeneration Program.

REFERENCES:

1. DNE Calculations N2-67-2A (B25 870123 814).

2. Training Documentation, SQN-DC-V-24.2 (841 870902 005 and B25 870910 812).

3. Pipe Support Calculation Regeneration Program Plan (841 870716 253).

4. Letter from S. A. White to NRC, dated August 21, 1987 (L44 870821 804).

5. Functional Verification of Supports for Rigorously Analyzed Category I Piping: Sequoyah 2; CEB-DI 21.83 (B41 870716 252). l

6. CAQR SQP871484IDI (S13 871001 847).  !

7. DNE Calculations IDI . Review (later).

fTEM NO.: 03.3-2 <

TITLE: Pipe Support Discrepancies

SUMMARY

OF ITEM:

Background - Ten pipe support analyses for piping problems N2-67-10R and N2-67-11R were reviewed for compliance to CEB design criteria and FSAR commitments.

Description - Three pipe support analyses associated with piping problems N2-67-10R and N2-67-11R were reviewed and found to have errors relating to assumptions and dimensional data. One analysis was found to have used an unconservative assumption during the development of the structural model. Two other supports were modeled with incorrect dimensions.

Part A: Pipe support analysis 47A450-25-344, . Revision 2, dated July 28, 1981, j used a dimension of 17 inches between support points at nodes 1 and 4 that attach to a surface-mounted baseplate. Two. separate dimensions on the support i drawing indicated that this dimension was 11 inches, while the difference I between the elevation listings on the tube steel indicated that the dimension was 17 inches. It appears that the analyst used the larger number based on the assumption that use of the 17-inch dimension was conservative. In reality, the 11-inch dimension between supports to the braced cantilever structure will result in higher bolt loads and structural loads. CEB concurred with this conclusion.

Part B: Pipe support _ analyses 47A450-25-348 and 349, Revision 2, dated ,

July 28, 1981, used incorrect dimensions for the vertical location of each  !

support on a common 6 inch by 6 inch by 3/16 inch structural tube. Pipe .

support 47A450-25-349 was modeled as connecting to the common beam at a j location of 26 inches above the baseplate, while the correct dimension was determined to be 29-3/4 inches. Similarly, pipe support 47A450-25-348 was modeled as connecting to the common bcam at a location of 44 inches above the baseplate, while the correct dimension was determined to be 46-1/2 inches.

Since this common beam is essentially a cantilever beam, the location of the loads at a greater distance above the baseplate will result in higher anchor bolt loads and structural loads. CEB concurred with the larger dimensions. l Basis - Criterion III of 10 CFR 50, Appendix B, " Design Control," states in part that measures shall be established for the control of design interfaces and verifying or checking the adequacy of design. Criterion VI of 10 CFR 50, Appendix B, Document Control," states in part that' adequate control shall be maintained for safety-related drawings. This requirement is also reiterated in the SQN FSAR paragraphs 17.2.5 and 17.2.6 (Amendment No. 20). Contrary to these commitments, a piping support detail drawing specified inconsistent dimensions., and two pipe support calculations used incorrect dimensions to qualify the respective pipe supports.

ITEM NO.: 03.3-2 (Continued)

TITLE: Pipe Support Discrepancies (Continued)

Impact on Design - The unconservative assumption used in pipe support calculation 47A450-25-344 and the incorrect dimensions used in pipe support calculations 47A450-25-348 and 349 will result in higher anchor bolt loads and structural loads. These effects must be evaluated in order to determine their safety significance. The team notes that CEB had scheduled all three pipe supports for regeneration to CEB Design Criteria SQN-0C-V-24.2 before unit 2 restart.

Extent - Since errors were detected in three out of ten support analyses evaluated, this deficiency suggests generic implications. ,

References - 1. Pipe Support Calculations 47A450-25-340, 348, and 349.

2. FSAR Section 3.9.2, Safety Class B, C, and D Fluid Components.

3. CEB Design Criteria WB-DC-40-31, Revision 0, August 29, 1975. '

4. CEB Design Criteria SQN-DC-V-24.1.

5. ANSI B31.1-1967.

CLASSIFICATION: Minor Calculation Error.

RESPONSE

l It is recognized that minor discrepancies do exist between the drawing and-calculations as discussed below. It is noted that each of these conditions will be corrected by the ongoing programmatic evaluations on Category I pipe support documentation. These conditions are being tracked by CAQR SQP871498IDI (Reference 1).

Part A: Pipe support 47A450-25-344 was initially issued for ECN 2548. The available calculation documentation for this support is in agreement with the original issued configuration. The calculation documentation was not, however, updated to record the acceptability of subsequent construction configuration changes made by the Support Modification Request (SMR) process.

Engineering did, however, signoff on the SMR granting approval. Support 47A450-25-344 has been evaluated considering the as-detailed condition. These evaluations confirm that the supports meet design requirements (Reference 2).

Part 8: Supports 47A450-348 and 349 have been evaluated considering the as-detailed configuration. These evaluations confirm that the supports are within the design requirements (Reference 2).

The root cause of the part A condition is inadequate change control to ensure drawing changes were incorporated into the calculation documentation.

ITEM NO.: D3.3-2 (Continued)

~ TITLE: Pipe Support Discrepancies (Continued)

The root'cause of part B condition is human error.  !

l Potential extent applies to. pipe supports within the scope of the Rigorous Pipe Support Calculation Regeneration and the Alternate Analysis Review Program.

To prevent recurrence of the part A condition, SMRs are no longer used at SQN.

These conditions are being addressed generically by the Pipe Support Regeneration Program for supports on rigorously analyzed piping, as described in D-3.3-1. As required by this program, existing support calculations ". . .

shall be reviewed for . . . specific configuration attributes." This

. programmatic review will generically correct these conditions for supports on rigorously analyzed piping.

l Implementation of the Alternate Analysis Review Program as described in D3.3-1 i will correct these conditions generically on Category I alternately analyzed l piping. As required by this program, each support on Category I alternately I analyzed piping will be walked down per Special Maintenance Instruction i SMI-0-317-55. The as-built configuration will be recorded and a calculation documentation package generated to verify the conformance of the support to  ;

design requirements. This verification process will correct documentation / drawing discrepancies for supports on Category I alternately i analyzed piping.

REFERENCES:

1. CAQR SQP871498IDI (S13 871002 821). l

2. DNE Calculations IDI-Review (later).  !

l l .

l

ITEM No.: D3.3-3 TITLE:- Incorrect Pipe Support Allowable Stress

SUMMARY

OF ITEM:

Background - Pipe supports for safety-related systems'are analyzed using loads from the TPIPE piping analysis computer code. Support loads are developed for various loading conditions and load combinations and are associated with '

specific system service conditions (i.e., normal, upset, faulted). TPIPE lists the pipe support loads as normalized loads, which are then used by the pipe support designer to analyze the pipe supports.

Description - Pipe support normalized loads derived from the TPIPE computer i analysis are based upon the largest value arrived at when each specified load or load combination is divided by the appropriate factor associated with the allowable stress for each service condition. This normalized load is then .

used in the pipe support analysis and the resulting stress is then compared to l the normal condition allowable stresses. For example, in a faulted condition- i the primary plus secondary support load would be divided by a factor of 1.6 i

'(SQN Pipe Support Design Criteria SQN-DC-V-24.1. This resultant normalized load would then be applied in the pipe support analysis and the weak-axis '

bending stress.for any support members. It would be compared with 0.75 Fy (Fy

- material yield stress). The faulted factor of 1.6, therefore, allows an allowable stress for the faulted condition of (0.75) Fy (1.6) - 1.2 Fy. Both the pipe support design criteria (reiterated by the design manual in note (a) of Table 7.17.4 A-1) and the FSAR require a faulted allowable stress limit of 0.9 Fy.

Basis - SQN FSAR Section 3.8.4.5.2 and Table 3.8.4-2 limit the allowable '

stress in steel structures for faulted load cases (SSE) to 0.9 Fy. This commitment is also reiterated in the SQN Pipe Support Design Manual, Volume 3, Section 7.17, Table A-1, Contrary to this, pipe support allowable stresses can exceed 0.9 Fy for some types of stresses in linear supports (i.e.,

weak-axis bending).

I ITEM'NO.: 03.3-3 (Continued)

TITLE: Incorrect Pipe Support Allowable Stress (Continued) i References - 1. SQN FSAR Section 3.8.4.5.2, " Structural Steel" and Table l 3.8.4-2, " Auxiliary Control Building Structural Steel Loads, Loading Conditions, and Allowable Stresses."

2. SQN Pipe Support Design Manual, Volume 3, Section 7.17, Table A-1, Revision 0, April 22, 1983,

3. US NRC Standard Review Plan, Section 3.8.3. " Concrete and Steel Internal Structures of. Steel or Concrete Containments." l

4. Design Criteria SQN-DC-V-24.1, " Location and. Design of Piping I Supports and Supplemental Steel in Category I Structures," l Revision 0, June 23, 1986. l

5. Design Criteria WB,-DC-40-31.9, " Location and Design of i Piping-Supports and Supplemental Steel in Category I l

Structures," Revision 6, February 10, 1986. l CLASSIFICATION: Design Deficiency.

RESPONSE

TVA agrees that Design Criteria SQN-DC-V-24.1 permitted faulted allowable stresses to exceed FSAR limits. This condition has been previously identified '

and corrective action instituted during the programmatic evaluation of pipe supports on rigorously analyzed piping. This condition is being tracked by CAQR SQP8714951DI (Reference 2).

This condition is being addressed by the Pipe Support Calculation Regeneration Program for supports on rigorously analyzed piping. This condition is applicable to all pipe supports on rigorously analyzed piping as described in 03.3-1. As required by this program, a new design criteria for Category I )

supports on rigorously analyzed piping, SQN-DC-V-24.2, has been issued, restricting the maximum allowable stress for the faulted condition to 0.9 Fy.

Implementation of the Alternate Analysis Review Program (AARP), as discussed  !

in D3.3-1, will correct ~this condition on Category I alternately analyzec

, piping. However, the 1.6 factor increase of normal AISC allowances is part of the interim acceptance criteria for Phase 1 of the AARP. As required by the Phase 2 portion of this program, all supports will be evaluated to determine '

compliance with design basis requirements. Design Criteria, SQN-DC-V-24.2, which limits stress for the faulted condition to 0.9 Fy will be revised before the start of the Phase 2 program to include Category I supports on alternately analyzed piping.

fTEM NO.: 03.3-3 (Continued) ,

l' TITLE: Incorrect Pipe Support Allowable Stress (Continued)

'The cause of this deficiency is that the faulted factor is a commonly accepted /

i criteria. TVA choose, however, to invoke a more stringent criteria in the ,

FSAR. The design criteria for pipe supports (SQN-DC-V-24.1) was not ,

I -

reconciled with the FSAR commitment.

This condition potentially applies to other CEB criteria and has been previously identified in Condition Adverse to Quality (CAQR) SQP871388 (Reference 1) to apply to evaluations and modifications made since 1983 which are in accordance with general civil structural Design Criteria , - r SQN-DC-V-1.3.3.1. The corrective action and recurrence prevention for this condition in the civil / structural area are as defined in the aforementioned CAQR.

To prevent recurrence in the CEB Engineering Hechanics Group Design Criteria SQN-DC-V-24.2 for supports on rigorously analyzed piping has been issued.

re:onciling FSAR and other licensing commitments with the design criteria requirements. SQN-DC-V-24.2 will be amended on the next revision to include '

supports on alternately analyzed piping.

REFERENCES:

1. CAQR SQP871388 (S13 870903 810).

2. CAQR SQP8714951DI (S13 871002 818).

i l

f i

i-1 F

r ITEM NO.: D3.3d TITLE: Pullou'. Loadings for Baseplate and Anchor Bolts y

{:

SUMMARY

Or 11EM:

Background - TVA Uses prequalified " typical" pipe supports where possible as .

part of their alternate arcalysis process. These typicals have been qualified l to loading 3 specified.as part of the alternate analysis process.

Description - Upon r9 viewing the calculations (47A053-101, 102, 114) performed i to verify these ' typ! cal'l pipe support designs, it was observed that when calculating the was used with no,boic loacs and

' consideration baseplate given thicknesses, to the pullout onlySince loadings. the moment the loading pullout loading wi'l increase both the bahMate stresser and volt pullout loads, it should have been considered.

Basis - iSAR Seption 17.lA.3, " Design Control" (Amendment No. 20), provides for the proper control and use of loadings. Contrary to these requirements, the applied pullout loads for the anchor polts and baseplates were not I considered.

Impact on design - The omission of a pertinent loading condition could result in undersized baseplates and anchor bolts.

Extent - This error may only be limited to the 47A053 series of alternately analyzed sup;, orts. TVA should verify the extent of the error.

References - 1. Pipe Support Calculations: 47A053-101, 47A053-102, i 47A053-114, Revision 0, January 7, 1980, t

CLASSIFICATION: Design Deficiency.

RESPONSE

TVA agrees that the baseplate attachments for typical pipe supports 47A053-lC>), -102, and -114 were incorrectly evaluated. It is noted that this condition'would have been corredted for each Category I support based on these ty;iical supports by the ongoing prc#emnratic evaluation of Category I pipe 36p p t documentation. This condition is being tracked by CAQR SQP871494IDI ,

y (Reference 1).

The calculations for typical pipe supports on drawings 47A053-101, -102, and

-114 did not consider in combination, the effects of the applied moment and tension in determining the anchor bolt loads and plate stress. However, the analytical method used to determine bolt loads due to applied moment was conservative. As a result, an evaluation of es.ch of the baseplate' attachments ]

of the aforementioned typical supports (Reference 2) has shown that under  ;

combined moment and tension the anchor bolt loads and plate stresses predicted i by the original calculation are acceptable. l 1

l ,

,c

__-_____a

.p +-

g

/

n -

ff

/ >

.)' p0y ITEM NO.: D3.3-4 (Continued) , g,j f TITLE:

Pullout Loadings for Baseplate,and , ,i , Abchor) Bolth( Continued) <

P, ThecauseofthisconditionisthatinWalculation'prepa'rErWnd* checker (vendor: EDS fleid engineering group) apparently considered thC effects of the direct applied tensile loading to be insignificant compared to tht! applied- i moment and, therefore, did not consider the combination.

h Theextentofthiscondition~ispotentiallyapplichbletot$et/pical' pipe support calculations prepared by EDS field groual. ,

To prevent recurrence, the applicable TVA design standards and criteria ,

explicitly require the co,nsideration of both inoment and' tension (if both exist) in the determinWon' of

• anchor bolt loads and' piate stresses. Since  %

the EDS field engifsering" group--no longer exists aad since.this condition will - 1 be covered by the generic programs discussed belog no further preventive action is required. P This condition is being generically addressed by the Pipe Sud ort Calculation Regeneration Program for support?.on rigorously analyzed piping af )iscussed in D3.3-1. As required by thiuprogram, supports will be evaluhted per che requirements of SQN-DC-V-24.2, "k pport for Rigorously Analyzed Category I Piping." Appendix III.,Section 4.1, states "The tensile load in anchors shall be calculated using a gethod that accounts for the applied tensile or compressive load pardlel to the anchors and the applied bendk.g moments about axes perpendicular to the anchors."

a , .

Implerantaticn afsthe Alternate ' Analysis Review Program as discussed-in D3.3-1 will h.'o rrect this condition ' generically for supports on Category I alternately anabrtdgiping. As required by this program, each support will be evaluated

.to det9rgine compliance with Civil Design Standard DS-C1.7.1 which states "The

';tensfie load in anchors shall be calculated using a method that accounts'for Jpe applied tensile or compressive load parallel to the' anchors and the' Wpplled bending moments about axes perpendicular to the anchors." )

' /

A/aresultofthePipeSupportCalculationRequirementProgramaththe Alternate Analysis Review Program, the calculation documentation fur @ tegory I pipe supports is being individually revised and updated or regead.aud to ,

fJaport'designcriteriarequirements. Supporttypicalswhichare' pot} updated

^? lkiiloeceactivatedandnolongerusedinsupportofpiping. /) !

j i '! ,

iEFERENCES: 1. CAQR S09871494. (513 871002 817).

s

2. DNE Calculations IDI-Review (later).

x iM r f fy 4

l

(

e f

,/ / '

t)

• 4.

( '

)

l

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ . _ -.t.

1 fTEM NO.: 03.3-5 ,

l

\

TITLE: Incorrect NCR Corrective Action 1

I

SUMMARY

OF ITEM-Background - Unistrut clamps have commonly been used at SQN on small piping )

(under 6-inch diameter) for multi-directional loading. Unistrut clamp test {

data were originally based on one direction of loading. Designers have used the allowable load based upon the one directional test in the appropriate direction regardless of the number of loading directions. Ga June 30, 1982, an NCR was written to address this issue.

Description - During a review of design change documentation, the IDI team determined thr.t nonconformance report SQNSHP8213, dated June 30, 1982, was written to address the use of Unistrut clamps qualified for one directional loading but used in situations with multiple loading directions acting simultaneously. The NRC concluded that an interaction equation should be used when these clamps were subjected to multi-directional loading. The TVA corrective action included an investigation and review of the clamp loadings consisting of testing of the clamps for multi-directional loading. The testing accounted for separately applied clamp loadings in each of two orthogonal in-plane clamp loading directions, as well as loading parallel to the pipe axis. In order to address the interaction effects of the three loading directions, the resultant test individual maximum allowable loads were compared with the previously established maximum load applied for each loading direction. These resultant ratios were then evaluated by one of two methods as outlined in NCR reevaluation calculation HBN SQP 8237, dated January 25, 1983, with Revision 1 dated July 6, 1984. A first attempt was made to interact the ratios using a straight line formulation that algebraically adds the ratios and accounts for contributions from two bending stresses and axial stress on the Unistrut clamp. If the sum of these three ratios is less than or equal to 1.0, then the particular clamp (based on pipe diameter) is considered quclified to the multi-directional test loads. Compliance with this interaction equation would indicate that previous use of the particular clamp size was acceptable with no further evaluation required. This straight line interaction equation demonstrates the clamp body capacity for the interaction of tension and bending on the clamp. For those pipe clamp sizes, ,

however, where the straight line interaction equation failed to demonstrate the clamp's qualification for prior usage, CEB used an elliptical interaction equation. This type of interaction has historically been associated with the combination of tension and shear stresses (i.e., bolting), and is not readily applicable or justifiable for the interaction of tension and bending. CEB I

e

XTEM NO.: 03.3-5 (Continued) i i '

TITLE: Incorrect NCR Corrective Action (Continued)

' states that.tne basis for using the elliptical interaction equation when the clamp failed the straight line interaction equation was that the test data indicated bolt failure as the predominant test load failure. The IDI team agreed that the use of the elliptical interaction equation was suitable for bolt shear and tension interaction, but questioned CEB's method of combining ,

tension and shear. The sum of the squares of the ratios associated with each l' l loading direction was added algebraically and compared with unity. This method does not account for the fact that two of the loading directions  !

provide tension in the bolts and their contribution (i.e., ratios) should be summed before squaring, and then added to the square of the ratio associated l with the bolt shear loading direction. In addition, this interaction only qualifies the clamp assembly for bolt tension and shear and does not address i the clamp body. Currently, the calculations show that the straight line interaction equation for the evaluation of clamp bending and tension does not provide adequate qualification for pipe clamp sizes 1-1/2-inch, 2-inch, 2-1/2-inch, 3-inch, and 4-inch diameters. The clamp tension and bending f interaction needs to be addressed independently of bolt tension and shear  !

interaction for these pipe clamp diameters in order to qualify previous Unistrut pipe clamp usage. In addition, the use of the straight line interaction equation may not be suitable to properly evaluate the contribution due to friction loading on the clamp. This is discussed in further detail in  !

related deficiency D3.5-3. -

Basis - FSAR Section 17.2 refers to TVA Topical Report .TVA-TR75-1A, Section 17.2. Section 17.2.16, " Adverse Conditions and Corrective. Actions" of that report stipulates that procedures shall provide for the identification and correction of adverse conditions. Contrary to this, the corrective action for NCR SQNSHP8213, dated June 30, 1982, did not properly consider the interaction .)

of tension and shear on the clamp bolts and tension and bending on the body clamp.

References - 1. NCR SQNSWP8213, June 30, 1982. .

2. EN DES Calculation, NCR Evaluation HBNSWP8237, January 25, j 1983, with Revision 1, July 6, 1984.

3. Salmon and Johnson, " Steel Structures: Design and Behavior," 2nd Edition, 1980.

4. EN DES Calculation, "Unistrut Pipe Clamp Load Ratings,"

July 27, 1982. -

I i

i 4

l

TTEM NO.: D3.3-5 (Continued)

TITLE: Incorrect NCR Corrective Action (Continued)

CLASSIFICATION: Design Deficiency

RESPONSE

TVA agrees that the method of interacting the combinea effects of loads in l three orthogonal directions.is incorrect. TVA does not consider it i i appropriate to evaluate the bolt and clamp strap separately since the j l evaluation methodology is by load rating the total clamp assembly rather-than 1 l by an analytical evaluation of each of the clamp components. To resolve this j condition, the following actions have or will be taken in the areas which '

utilize the Unistrut P-2558 clamp (CAQR SQT871487IDI, Reference 1, has been written to track resolution of this condition. (Note: The acceptability of a straight-line interaction is being addressed under IDI. deficiency D3.5-3.) j l

Supports on Rigorously Analyzed Piping - The rigorously analyzed Pipe Support Calculation Regeneration Program, as discussed in D3.3-1, is utilizing a 4 straight-line method to combine the effects of applied loads. This is an acceptable interaction method and will resolve this condition for pipe l supports on rigorously analyzed piping. Corrective action will be completed as committed in the Pipe Support Calculation Regeneration Program.

/

Supports on Alternately Analyzed Piping - Implementation of the Alternate Analysis Review Program, as discussed in D3.3-1, will address this condition l for supports on. Category I alternately analyzed piping. Unistrut clamp attachments on Category I alternately analyzed piping supports verified during the unit 2 prerestart phase were evaluated using a straight-line interaction. The Phase 2 program will be per the requirements for Design '

Criteria SQN-DC-V-24.2 (Refsrence 4) which states that a straight-line combination should be used in Unistrut P2558 clamp evaluations. Correction action will be complete.

Conduit Supports - The design basis for Unistrut P2558 clamps used on conduit ,

supports is a straight-line interaction. No further action is required. l l

Instrumentation Sensing Line Supports - The use of the Unistrut P2558 clamps on instrumentation lines is limited to 1/2-inch diameter, Schedules 80 and 160 instrument sensing lines. Instrument sensing lines have an ambient operating temperature and are supported by multiple 3-way supports at frequent intervals. An assessment of the clamp using a straight-line interaction provides adequate margins compared to the allowable capacity. No additional work is required for instrument lines (Reference 2).

l ZTEH NO.: D3.3-5 (Continued)

TITLE: Incorrect NCR Corrective Action (Continued)

This condition is considered to be an isolated design error with unknown cause. 'A Quality Information Release (Reference 3) has been issued to Civil Engineering Branch Lead Engineers requiring that a straight-line interaction be used for the evaluation of Unistrut P2558 clamp attachments. No further preventive action is required.

REFERENCES:

1. CAQR SQT871487IDI (S13 871001 8020).

2. DNE Calculations for Instrument Supports (later) ,

3. Quality Information Release, QIR CEB-87-099 '

(B41 870710 250).

4. SQN-DC-V-24.2 (B25 870826 101) l l

l I

l l

1

_____---_ l

ITEM NO.: D3.4-1 TITLE: Motor Operated Valve Design Pressure j

SUMMARY

OF ITEM: l l

The vendor seismic qualification calculation for motor operated valves l 47H427-6 and -7 used a design pressure of 50 lb/in* although the system j design pressure is 150 lb/in . The design pressure used in the vendor  !

calculation dces not conflict with the valve procurement bill of materials, 47BM427 SHT 14 R0, but does conflict with the system flow diagrams, 47H803-2 .

R12 and R13 and the Insulation and Operating Mode Analysis Data Drawings, 1 478466-3-11.

CLASSIFICATION: Documentation

RESPONSE

The root cause of this finding is that before 1977 the flow diagram (47H803-2) did not include a detailed breakdown of the pressure / temperature boundaries.

The bill of material for the above valves, 47BM427 SHT 14 R0 dated 10-28-71, stated that the valves would be 150 lb class valves for feedwater service at 50 lb/in and 120"F. In 1977, a pressure / temperature table was added to the flow diagrams clarifying locations of the pressure / temperature interfaces. This clarification was not present when the equipment was procured. The bill of material also included valves to the motor driven auxiliary feedwater pumps, which are affected by the same statement.

PIRSQNMEB87112IDI (Reference 1) has been written to identify this discrepancy. As corrective action, TVA has evaluated the identified motor operated valves for adequacy using the actual design pressure. These evaluations, documented in calculation package CEB-CQS-311 (reference 2),

concluded that these valves are acceptable. In addition, TVA has identified other potentially affected valves by reference 4. For a portion of the i reference 4 valves, an evaluation of the adequacy of the valves for the correct design pressure is documented in calculations (references 5, 6, and i 7). The review of the remaining reference 4 valves is in process. All affected documents will be revised to indicate the correct design pressure and temperature, postrestart.

FSAR table 3.2.1-2 presently requires that TVA Class C valves installed in the ERCH system be qualified by test. This requirement is being changed by reference 8 to include testing or analysis as an acceptable qualification i method for any valves in the ERCH system.  !

Based upon the review effort to date it is judged that there is no in&act on safety and no hardware changes are required.

I

ITEM NO.: 03.4-1 (continued)

TITLE: Motor Operated Valve Design Pressure (continued)-

REFERENCES:

1. PIRSQNMEB87112IDI

2. DNE Calculation'(841 870831 005)

~3. QIR-87-205 R0 (841 871006 006)

4. QIR-SQP-87-555 (B25 871022 003)

5. DNE Calculation (B41 871013 026)

6. DNE Calculation (B41 871013 027)

-7. DNE Calculation (841 871013 028)

8. Memorandum from H. E. Pennell to J. B. Hosmer dated August 28, 1987 (B41 870828 008) i i

/

l

.)

l

\

l

)

l l

ZTEM NO.: D3.4-2 TITLE: Seismic Qualification of Turbine Driven Auxiliary Feedwater Pump 2A

SUMMARY

OF ITEM:

TVA flow diagram 47K427-57 specifies a design pressure of 1650 lb/in*g at I the' discharge nozzle of turbine-driven auxiliary feedwater pump 2A. The pump j is TVA Safety Class C.

The equipment nozzle loads that TVA supplies to equipment vendors (CEB Report 82-1) to'use in the seismic qualification of equipment requires consideration of nozzle axial load due to internally induced pressure in addition to the forces and moments the piping imposes on the nozzle.

The vendor's seismic qualification calculation for the turbine-driven 1 auxiliary feedwater pump does not consider the axial thrust at the pump I discharge nozzle due to the 1650 lb/in*g design pressure.

CLASSIFICATION: No Deficiency. j

RESPONSE

1 The purpose of the CEB Report 82-1 note was to ensure that the pressure i stresses were properly accounted for in the nozzle stress evaluation. For the AFW pump qualification, the vendor considered both the mechanical loads from the piping and the pressure stresses appropriately in the qualification. The pressure thrust was not included in the anchorage design because it was a self-equilibrating load. A mathematical derivation is attached which shows that a pump anchorage is not affected by steady-state pressure and fluid momentum loads except in the case of a pump connected to an untied bellows. I Untied bellows were not used for the AFW pump. When used, untied bellows I should have been considered part of the analyzed pipe and the unbalanced I pressure thrust loads should have been evaluated for pipe stress and nozzle j load qualifications. As stated in reference 4, two (2) untied expansion joints were found that did not evaluate the thrust loads and require'further investigation. This investigation is under way and will be completed before restart. The intent of the note was to clarify to the vendors and piping analysts, that the allowable load (Px) calculated by using the equation does not include the internally induced longitudinal shell load due to pressure.  !

The intent of the note was interpreted correctly by both vendors and piping  ;

analysts. ]

Each pump vendor is responsible for a design which can withstand normal transient and steady-state operations in addition to loads specified by TVA. ]

4 The pump vendor has developed a standard design through analyses, test, and experience which has an inherent margin of safety to accommodate normal operation conditions. Normal operating loads which can be quantified have L been accounted for in the existing seismic analysis. It is TVA's responsibility to incorporate design features and operational procedures which will be free of significant transients and conduct tests to verify this. TVA l

l

, l l

l ITEM NO.: D3.4-2 (Continued)

TITLE: Seismic Qualification of Turbine Driven Auxiliary Feedwater Pump 2A (Continued)  ;

RESPONSE: (Continued) i has implemented this philosophy throughout the SQN-2 system designs and  !

preoperational test programs conducted (Reference 1). The AFH system was tested for steady-state and transient response by preoperational test TVA-22. j Transients were evaluated by placing observers at strategic locations and )

starting and stopping various pump combinations. j Programmatically, where a postulated transient cannot be tested and where problems have been identified by operation and industry experience, such as those defined by NUREG-0927, analyses have been performed and required modifications implemented. The loads from these analyses are included as applied nozzle loads. Such loads affect the anchorage design and are evaluated accordingly.

In summary, normal transients have been incorporated into the design of pump anchorages. Significant transients are accounted for by piping analysis. The internal pressure at the AFH pump nozzle during steady-state operation does not result in an axial thrust component to be reacted by the pump anchorage.

A letter from the vendor, Ingersol Rand, shows their concurrence with this position in reference 2. TVA's approach to evaluating both steady-state and transient loads is adequate and this item is not a deficiency.

REFERENCE:

1. " Review of Fluid Transients for Sequoyah Nuclear Plant Unit 2," prepared I by D. A. van Duyne and R. E. Fortier, An independent review by Stone and Hebster Corporation under TVA contract TVA-72102A (B41 870924 304), letter dated September 1, 1987.

2. Letter of concurrence from Ingersol Rand transmitted by telecopy dated October 8, 1987 (B41 871020 004).

3. QIRCEBSQN870165 (B25 871009 150)

4. TVA Memo dated 10/8/87, from R. E. Roemer to K. S. Seidle (10/22/87, from K. S. Seidle to H. E. Pennell, (B41 871022 004)) J ATTACHMENT: Steady State Loads on Pump Anchorage (Figures 1, 2, & 3)

~

Ital NO.: D3.4 ATTAQBENT, FIG 1 TITE: Scianic Qualification of 'ntrbine Driven Aux Feedwater Pmp 2A STEADY STATE LOADS ON PUMP ANCHORAGE Y

=x COORDINATE SYSTEM ,

@ 0)

FORCE AT ELOOW OUE TO o b CHANGE IN MOMENTUM TORCE AT ELBOW 00E TO CHANGE IN MOMENTUM

/0V, p0va P, p R = ANCHORAGE .

Q ..' -

$ i

. i

/G ( Va - V,) g

& CHANGE IN MOMENTUM l '

BETWEEN STATION h AND h V, .

V:

.- SIMPLIFIED PUMP CONFIGURATION O

CON 0lTl0NS  : 1. STE,ADY STATE OPERATION ,

2. PlPE Et.00WS ARE UNRESTRAINED IN THE X OIRECTION

3. SINGLE PHASE FLOW ,..

4. CLOSED SYSTEM -

5. MOT 0fi THRUST BEARING INTEGRAL WITH PUMP 1

I H ui NO.: D3.4 ATTAOBENI', FIG 2 TITE: Seisnic Qualification of 'Ibrbine Driven Aux Feedwater Pimp 2A

STEADY STATE LOADS ON PUMP ANCHORAGE r,

PQVi

-* [ _

[ pA Il -

y,

( l -

0 3

LOAD FREE BODY BETWEEN SECTIONS @ AND @

Vi t

EFx = #0(AVx) SUM OF APPLIED FORCES EQUALS CHANGE IN MOMENTUM F piA i- /QV i = 0 i F = p, As, + /QV (l) 3 F2 V2 k=

P22 T #0V2 g

LOAD FREE BODY BETWEEN SECTIONS @ AND @

2Fx e /Q(AVx )

"V2 p, A, - F, + /QV, = 0 ,

F2 = p, A2 + #0V2 II

I'IDd 10: D3.4 ATTAONENT, FIG 3 ,

TI'112: Seisnic Qualification of Turbine Driven Aux Feedwater Ptrrp 2A STEADY STATE LOADS ON PUMP ANCHORAGE o b Fi l

=-- p, As _

Z p A,

=- -

=-

LOAD FREE BODY BETWEEN SECTION @ AND @

2Fx = /0 (6Vx) (3)

, I

- Fi + p, A , + R + F2 ~ P2 A2 = /Q(Vg -V i ) l SUBSTITUTE EON (1) AND (2) FOR F AND i F RESPECTIVELY 2

R=0 e 4

L 8 TEM NO.: 03.4-3 i

I TITLE: CCH Heat Exchanger Calculation l

SUMMARY

OF ITEM:

The drawings p'rovided by the vendor of the component cooling water (CCW) heat-exchanger indicated that it had two saddles, one on each end. The seismic l qualification of the heat exchanger was provided by the vendor based upon this configuration. The NRC observed from a field walkdown that the as-built condition actually has a middle support (i.e., three saddles).

CLASSIFICATION: Documentation

RESPONSE

The documentation deficiency in this case is due to the lack of appropriate interface review of a design change initiated by TVA piping analysis engineers. The existing procedures for review and approval of design changes to components should have ensured appropriate review and approval of the addition of the CCW heat exchanger middle saddle by CEB's component qualification engineers. Instead the saddle was specified as a " pipe support" and installed without component qualification or vendor approval.

A preliminary analysis of the CCW heat exchanger assembly indicates that the heat exchanger and supports are or can be qualified for all loadings. A detailed review of support and anchorage qualification for the CCW and 10 other Category I heat exchangers has been initiated to ensure no safety impact on the plant.

References 1 and 2 document two independent reviews which conclude that this specific deficiency was an isolated occurrence.

CAQR SQF870199 has been written to resolve this issue (reference 3).

REFERENCES:

1. Documentation of no further unauthorized " pipe supports" (B41 871021 006).

2. SWEC report on generic heat exchanger review (B45 871019 002).

3. CAQR SQF870199 (805 871021 300) l I

i f

1 1

l L_ ---------

ITEM No.: 03.4-4 TITLE: CCH and CS Heat Exchanger Nozzle Loading

SUMMARY

OF ITEM:

A vendor, Industrial Process Engineers (IPE), provided TVA with the seismic qualification of the component cooling water (CCH) heat exchanger and the containment spray (CS) heat exchanger. The vendor considered four components of nozzle load allowables, but failed to consider shear loads. TVA specifies formulas for calculating nozzle allowable shear loads in CEB Report 82-1, Revision 1. i CLASSIFICATION: Documentation

RESPONSE

The seismic qualification calculations for the CCW and CS heat exchangers were reviewed, comments were resolved, and the designs accepted in early 1972. The vendor's calculations neglected the effect of nozzle shear loads on the local nozzle /shell interface. The original TVA reviewer did not take issue with the J neglect of the shear load at the local nozzle areas. j j

New calculations verify the original reviewer's judgment that the analysis j provided by the vendor was adequate (references 1 and 2).

The effect of nozzle shear loads on anchorage loads were also neglected in the vendor's calculations. However, the vendor nozzle loads were such that they enveloped the actual piping nozzle load effects on the anchorage. A detailed analysis of anchorage loads for the CCW heat exchangers and 10 other heat .

exchangers has been initiated to ensure no safety impact on the plant. The  !

two (2) IPE heat exchangers will be included in this review.  !

l

REFERENCES:

1. DNE Calculation (B41 870827 002).

2. DNE Calculation (B41 870826 001).

. l 1

l l

l l

l

)  !

? 1 i

l

' ITEM NO.: D3.4-5 '

1 TITLE: Vendor-Supplied Flexible Hose q

SUMMARY

OF ITEM: j

'I Flexhose was procured on. contract 73C53-83546-6. Item.5 of contract form l SQN-QAP-III 2.1-A exempted the requirement for a flexhose seismic qualification report. TVA did not adequately document the exemption to the criteria within its own procurement docurnent which reduced the seismic l

requirements that the flexhose vendor was required to satisfy. j CLASSIFICATION: Design Deficiency 3

RESPONSE

The justification for exception from seismic qualification requirements was  ?

not adequately documented. The flexhose vendor was not exempted from quality control requirements. Quality Information Request, QIRCEBSQN87197 RO has established that the extent (for both units 1 and 2) is 16 Flexonics flexible metal hoses'(Mark No. 17W586-226) procured on contract 73C53-83546-6 for use-  ;

in TVA class C (seismic Category I) piping systems. l Recurrence of this deficiency for new equipment purchases is precluded by CEB-DI-21.03 which requires that engineers, experienced in the field of j component seismic qualification, review the seismic qualification portion of purchase requisitions.

Analysis (reference 4) and inspection results (reference 5) demonstrate that the Flexonics flexible metal hose is adequate for interim operation at SQN

)

(i.e., no excessive bulging, misalignment, or loose braids noted).

There is no impact on safety and no hardware changes are required for interim operation of the flexhoses. A review of contract records for other similar items and past procurement practices will be conducted postrestart to ensure that the general-seismic exemption concern has been fully bounded.

Hardware modification is not required for interim operation. For long-term operation, the existing flexhose will be replaced, at the earliest j opportunity, with appropriately qualified flexhose. CAQR SQP871477IDI has i been written to track corrective action for this item. I l

1

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

ITEM NO.: D3.4-5 (Continued)

TITLE: Vendor-Supplied Flexible Hose

REFERENCES:

1. Contract 73C53-83546-6.

] 2. CAQR SQP8714771DI

3. QIR CEB-SQN--197RO

4. DNE Calculation, " Justification for Interim Operation with

• ' Flexonics Flexible Metal Hose"( later ).

5. Letter from Impell dated 10/14/87, Hosford to Seidle, (B41 871019 003).

6. CAQR SQP-7-556 (B25 871016 104)

I

ITEM NO.: 03.4-6 and D3.6-1 f

TITLE: ERCW Upper Containment Vent Cooler Frequency Calculation 6.nd Design Review for ERCH Equipment

SUMMARY

OF ITEMS:

Some vendor seismic component qualification reports appear to violate quality assurance provisions of the contract in that no evidence of a design review could be found. The preparer and checker were not identified by the qualification report. An arithmetic error was found in one of the reports as evidence that the report was not properly reviewed.

CLASSIFICATION: No Deficiency.

RESPONSE

In TVA's specification to the vendor, the vendor was required by QA requirements to submit design computations which were independently reviewed and certified to assure compliance with all requirements. If adequate independent review could not be furnished by the manufacturer, TVA could perform the independent review. In some cases the vendors opted to have TVA perform the independent review and submitted the component qualification reports for review and approval. The reports did not identify a checker which did not imply that the calculations were not checked for arithmetic errors before submittal. In the specific seismic reports identified by the deficiencies, TVA performed an independer t review. This was documented by correspondence and original vendor submittals.

Programmatically, the seismic evaluation of vendor-generated seismic analysis J and/or test of components was conducted by the component qualification engineers of the Civil Engineering Branch to ensure compliance with applicable  ;

design criteria. This review met the intent of an owners' review per ASME by present standards. In addition, the depth of review included a technical assessment to the extent that was necessary to satisfy the TVA reviewer of the component's design adequacy. The TVA review process also included a review by the section supervisor before final approval of the vendor submittal. An arithmetic check was not typically performed unless the magnitude of numbers presented appeared to be "out of the ball park." Minor corrections were mace directly on the vendor document submitted for review and the document approved with such corrections noted. Major errors were returned to the vendor for correction and then resubmitted for further TVA review until approval was obtained.

For the specific items in question, a representative from TVA's QA organization and an independent quality reviewer have evaluated this process and finds that the TVA reviews, from a practical standpoint, served the purpose of the independent review.

1

______________J

! TEM NO.: 03.4-6 and D3.6-1 (Continued) -

TITLE: ERCH Upper Containment Vent Cooler Frequency Calculation and Design Review for ERCH Equipment (Continued)

RESPONSE: (Continued) t To investigate this potential problem, a correspondence list (reference 1) which filed information by vendor and contract was reviewed. The files were searched for the years 1971, 1972, and 1973 as a frame of reference. The list was searched for vendors who had submitted seismic qualification reports which did not identify the preparer and checker. The search was confined to component assemblies such as air handling units, heat exchangers, tonks, and d dampers. From a review of the correspondence list in reference 1, four vendors were identified which did not have a prepared / checked signoff on their seismic qualification reports. One seismic qualification report from each was selected for review. Some minor errors were found.

The scope and quality of all the analyses performed by the manufacturers was accepted by TVA. The arithmetic errors have been corrected and carried through by TVA to determine the effect on seismic qualification. A review of ,

the upper containment vent cooler seismic qualification shows that the corrected natural frequency of 23.7 Hz does not change the calculated design margin. Seismic qualifications were not affected by the arithmetic errors. j In fact, the change in calculated design margins due to arithmetic errors were i insignificant. These changes have been documented by TVA engineering design calculations (reference 2). (

[

Results of this study reveal that there were no deficiencies in the implementation of the QA provisions by the vendors because: 1

1. The specification to the vendors stated that TVA could perform the independent review if adequate independent review could not be furnished by the manufacturer.

2. TVA's correspondence record (reference 3) documents the TVA reviews which were performed in place of or in addition to the vendor's independent review.

3. Changes in calculated design margins due to arithmetic errors were insignificant; therefore, there was no impact on the seismic qualification of equipment. 1 No deficiency has been identified and there is no impact on safety and no modifications are required, j H

I i

1

]

1 1

1 l

1

l l LITEM NO.: D3.4-6 and D3.6-1 (Continued)

L TITLE: 'ERCH Upper Containment Vent Cooler Frequency Calculation and Design Review for ERCH Equipment (Continued)

REFERENCES:

1. Civil Engineering Branch - Component Qualification Section

Correspondence Database by Vendor and Contract

2. DNE Calculations

.2.1 Arithmetic corrections to the seismic analysis, American Foundary and furnace Company (B41 871014 005) 2.2 Arithmetic corrections to the Seisatic Analysis, R. C. Products, Inc. (B41 871013 006) 2.3 Arithmetic corrections to the seismic analysis, H. K. Porter Company (B41 871019.001). .

2.4 Industrial Process Engineers seismic analysis.

, No arithmetic errors (MEB 820203 951).

. ATTACHMENT 1:

List of Contract Letters For: -

1. TVA Contract 71C33-92691

2. TVA Contract 72C33-92730

3. TVA Contract 72C35-92775

4. TVA Contract 73C35-83571-2 .

O l

l l

~ ~ - - _

1 TEM NO.:- 03.4-6 and D3.6-l'(Continued)

TITLE: ERCW Upper Containment Vent Cooler Frequency Calculation and Design Review for ERCW Equipment (Continued)

Attachment.I

1) TVA Contract 71C33-92691

a. Letter from Industrial Process Engineers dated 1/31/72, to TVA, D. B.' Weaver, Chief, MEB (Ref: TVA Contract 77P33-92691) l

b. Letter from TVA, D. B. Weaver, Chief, MEB, dated 3/23/72, to Industrial Process Engineers, Mr. G. B. Houar

c. Letter from Industrial Process Engineers dated 4/25/72, to TVA, l D. B. Weaver, Chief, MEB, (Ref: TVA Contract 77P33-92691)

d. Letter from'TVA, D. B. Weaver, Chief, HEB, dated 5/12/72, to "

Industrial Process Engineers, Mr. G. V. Houar ,

1

e. TVA Memo dated 5/15/72, to B. S. Montgomery from G. E. Givens 1

2) TVA Contract 72C33-92730 1

a. Letter from H. K. Porter Co. Inc., dated 6/2/72 to TVA, Chief 1 Mechanical Engineer i

b. Letter from H. K. Porter Co., Inc. dated 6/14/72 to TVA, Chief (

Mechanical Engineer l

c. Letter from H. K. Porter Co., Inc. dated 6/30/72 to TVA, Chief I Mechanical Engineer.  !

d. Letter from H. K. Porter Co., Inc. dated 7/17/72 to TVA, Chief I

Mechanical Engineer

e. Letter from H. K. Porter Co., Inc. dated 8/1/72 to TVA, Chief Mechanical Engineer  !

f. Letter from TVA, D. B. Weaver, Chief, MEB dated 8/14/72 to H. K. Porter Company, Mr. R. E. Schremp

g. Letter from TVA, D. B. Weaver, Chief, MEB dated 9/6/72 to H. K. Porter Company, Mr. R. E. Schremp

h. Letter from H. K. Porter Co. Inc., dated 7/11/72 to TVA, D. B. Weaver, l l

Chief-MEB

1. Letter from H. K. Porter Co. Inc., dated 9/28/72 to TVA, D. B. Weaver, Chief-MEB

j. Letter from TVA, D. B. Weaver, Chief, MEB dated 10/18/72 to H. K. Porter Company, Mr. R. E. Schremp

k. Letter from H. K. Porter Co. Inc., dated 10/25/72 to TVA, D. B. Weaver, j Chief-MEB l

1. Letter from TVA, D. B. Weaver, Chief, MEB dated 11/15/72 to H. K. Porter Company, Mr. R. E. Schremp

m. Letter from H. K. Porter Co. Inc., dated 12/18/72 to TVA, D. B. Weaver, Chief-MEB l

l l

LITEH;NO.: D3.4-6 and D3.6-1 (Continued)

TITLE: ERCW Upper Containment Vent Cooler Frequency Calculation and Design  !

l Review for ERCW Equipment (Continued)

n. Letter.from TVA, D. B. Weaver, Chief, MEB dated 2/5/73 to H l' H. K. Porter Company, Mr. R. E. Schremp  ;

o. TVA memo dated 1/9/73 to J. A. Hudson from J. I. Givens (B41 870330 004) ,

p. TVA~ memo dated 11/8/72 to R. E. Lyon from J. I. Givens (841 870330 004) l

q. TVA memo dated 10/10/72 to R. E. Lyon from J. I. Givens (B41 870330 004)

r. TVA memo dated 9/15/72 to R. E. Lyon from J. I. Givens (B41 870330 004)

3) TVA Contract 72C35-92775

a. Letter dated 4/21/72 from R. G. Products, Inc., to TVA, R. E. Lyon, l

-Chief-MEB  !

b. Letter dated 6/5/72 from R. G. Products, Inc., to TVA, S. C. Wyatt, Expediting Services

c. Letter dated 6/21/72 from TVA, D. B. Weaver, Chief, MEB to R. G.

Products, Inc., N. H. Finkle

d. Letter dated 8/1/72 from R. G. Products, Inc., to TVA, R. E. Lyon, Chief-MEB '

e. Letter dated 10/6/72 from TVA, D. B. Weaver, Chief, MEB, to R. G.

Products, Inc., N. H. Finkle

f. Letter dated 10/25/72 from R. G. Products Inc., to TVA, R. E. Lyon,. J Chief, MEB '

g. TVA memo dated 12/1/72 from J. I. Givens to R. E. Lyon

h. Letter. dated 12/11/72 from D. B. Weaver, Chief-MEB'to R. G. Products,  ;

Inc., N. H. Finkle

i. TVA, Materials Branch, Status Report dated 10/27/72, S. C. Wyatt a

j. Letter dated 12/29/72 from R. G. Products, Inc., to TVA, D. B. Weaver, Chief-MEB

k. Letter dated 1/24/73 from TVA D. B. Weaver, Chief-MEB to R. G.

Products, Inc., S. W. Klose

1. TVA memo dated 12/1/72 from J. I. Givens to R. E. Lyon (B41 870330 004)

m. TVA memo dated 9/25/72 from J. I. Givens to R. E. Lyon.(B41 870330 004)

4) TVA Contract 73C35-83571-2 a) TVA memo dated 2/27/73 from J. I. Givens to J. A. Madson l (B41 870330 004) I I

b) TVA memo dated 4/17/73 from J. I. Givens to J. A. Madson (B41 870330 004) c) TVA memo dated 4/27/73 from J. I. Givens to J. A. Madson (B41 870330 004) d) Letter dated 5/9/73 from TVA, D. B. Weaver, Chief-MEB to American Foundry and Furnace Company, R. K Maxwell e) TVA memo dated 7/16/73 from J. C. Key to J. A. Hudson (B41 870330 004)

l 2 TEM NO : D3.4-7 and DS.4-2 TITLE: Chiller Unit Seismic Qualification and Seismic Qualification of Westinghouse Switches

SUMMARY

OF ITEM:

The seismic qualification of a chiller unit (D3.4-7) and an electric panel (05.4-2) involve separate qualification of devices for installation onto the equipment assembly. The qualification of these items of equipment were interpreted to violate a provision for a margin of conservatism specified in the equipment procurement documents (Quality Assurance (QA) Manual, Appendix F

- Seismic Criteria) and discussed in FSAR, Section 3.10.

CLASSIFICATION: No Deficiency.

RESPONSE

TVA has a program in place.to ensure the appropriate seismic qualification of SQN's safety-related equipment. This program is consistent with industry standards and NRC requirements applicable to equipment qualification-for nuclear plants of the same vintage as SQN. The equipment seismic qualification criteria is provided in Appendix F of the SQN Quality Assurance Manual, which in turn reflects the criteria of IEEE 344-1971.

One of the primary objectives in the preparation of Appendix F was that it could be used as an inclusion to equipment procurement specifications to define the vendor's seismic qualification requirements. Because of the broad scope of equipment to be purchased to this criteria and the wide-range of seismic qualification expertise of the equipment suppliers, it was necessary-to include provisions for conservatism in the seismic qualification programs which could be invoked at TVA's option. One such provision of Appendix F is the 3/4 factor discussed in Section 6.1 which addresses the seismic .

qualification of devices (i.e., items of equipment which are not attached directly to the building structure for which seismic response spectra are readily available).  ;

The appropriate seismic qualification of a device, as a separate entity, requires that compatibility be established between the seismic responses of the support systems for the device and the seismic qualification level of the device. Paragraph 6.1.3 provides for a margin of conservatism by requiring

• that the predicted seismic response of the device support structure not exceed 3/4 of the device qualification acceleration level. The 3/4 factor for compatible qualification of devices and associated support structure is placed into appropriate context by footnote 1 to Section 6.1. This. footnote makes it clear that where appropriate engineering has been applied to the qualification of the device to' ensure that it envelopes the response level of its support structure, the 3/4 factor is not required.

ITEM NO.: 03.4-7 and 05.4-2 (Continued)

TITLE: Chiller Unit Seismic Qualification and Seismic Qualificat- 1 or Westinghouse Switches (Continued)

The SQN FSAR reflects the equipment qualification program in its c scussions of requirements and by appropriate references. FSAR Section 3.10 addresses the seismic qua'lification of instrumentation and its support structure, making i numerous references to Appendix F and IEEE 344-1971. The introductory l paragraph of Section 3.10.2 contains summary type excerpts from Appendix F. l This paragraph mentions the 3/4 factor requirement; however, in its brevity, the summary discussion does not include the clarifying information of the Appendix F, Section 6.1 footnote. The FSAR will be appropriately revised to eliminate this erroneous indication of an all inclusive application of the three-fourths factor.

With regard to the findings as related to the chiller unit assembly and the electrical panel, their qualification reports were evaluated to ensure that separate qualification of the devices conservatively enveloped their intenced installation requirements. The seismic qualification of these two items of equipment is in full compliance with the seismic design criteria of Appenalx F and the FSAR.

There is no impact on safety and no hardware changes are required.

REFERENCES:

1. Appendix F Seismic Design Criteria, Revision 2

2. SQN FSAR Section 3.10

ITEM NO.: U3.5-1 TITLE:- Piping' Code of Record

SUMMARY

OF ITEM:

Background - The piping code of record for design as stated in'the FSAR is ANSI B31.1-1967 Edition. Since ANSI B31.1-1967 did not define combinations for the normal, upset, and faulted conditions, TVA used the stress allowable equations from ASME Section III, Subsection NC-3000, Winter 1972 Addendum for these plant conditions.

Description.- In addition to the stress equations, TVA used the stress allowable limits specified in the ASME Code. This usage is documented in CEB's Rigorous Piping Analysis llandbook.

' Potential Basis - CEB's use of the ASME Code stress allowable limits is not consistent with FSAR Table 3.9.2-3, which commits to the use of ANSI B31.1-1967 stress allowable limits.

Impact on' Design - The stress allowable limits documented in ANSI B31.1-1967 and stipulated in FSAR Table 3.9.2-3 are generally more conservative for stainless steel materials than ASME Section III allowable stresses. Piping designs using stainless steel materials may, therefore, not meet design limits specified in the SQN FSAR.

References'- 1. ANSI B31.1-1967 Edition.

2. Section 3.9 of SQN FSAR, " Mechanical Systems and Components."

3. ASME Section III, Subsection NC-3000, Hinter 1972 Addendum.

4. CEB Design Criteria 13.3, Revision 3, August 13, 1984, and 13.3.1, Revision 2, December 27, 1983.

CLASSIFICATION: Documentation i

t 1

I l

, l

ITEM NO.- U3.5-1 (Continued)

TITLE: Piping Code of Record (Continued)

RESPONSE

Equivalence of the allowable stresses between ANSI B31.1 (1967) and ASME Section III, Classes 2 and 3 (1971 Edition, Winter 1972 Addenda) is assured because:

1. The stress criteria are identical, and

2. The allowable stresses were developed by the same committee for botn documents.

B31.1-1967 allowable stresses (Sc, Sh) were based on the lower of 0.25 Su and 0.625 Sy, where Su is the ultimate tensile strength and Sy is the yield strength. These are the same criteria as were used by Section I and VIII of the ASME Boiler and Pressure Vessel Code in 1967. When Section III Class 2 and 3 rules were published, they adopted the stress criteria of Sections I and VIII and were, therefore, identical to B31.1-1967 in this aspect. 831.1 was revised in 1973 to eliminate the citation of specific criteria and to replace them with reference to the criteria of ASME Sections I and VIII. Thus, the stress criteria were, and are, identical for B31.1 and ASME Section III, Classes 2 and 3 in that timeframe.

The allowable stresses for 831.1 and Section III are both developed by the 1 Subcommittee on Properties of Metals of the ASME Boiler and Pressure Vessel )'

Committee, using the same data base for both sets of allowable stresses. Even though the stress criteria are identical, minor differences may exist from time to time between 831.1 and Section III, Class 2 and 3 because of a lag on the part of one document or the other in the adaption of changes to the allowable stresses resulting from the addition of later test results to the data base.

The FSAR will be revised to change the source for allowable piping stresses from ANSI B31.1.0-1967 to ASME Section III, 1974 Edition through Winter 1976  ;

Addenda.

i l

ITEM NO.: D3.5-2 ,

TITLE: Use of Selected B31.1 Code Rules

SUMMARY

OF ITEM:

Background - The FSAR specifies ANSI B31.1-1967 as the piping design code of record for SQN. However, the FSAR also allows the use of ASME Section III, Subsection NC-3000, Winter 1972 Addendum, for the normal, upset, and faulted stress equations. This combination has been used to qualify an overstressed piping member by using selected portions of both codes.

Description - The summary of analysis report N2-67-2A for the ERCW system contains a calculation for an overstressed condition (page 83). This calculation uses the interpretation that for ANSI B31.1-1967 the seismic portion of the additive stresses did not require the use of the stress intensification factor (1) to increase stresses at a tee connection. The IDI team believes that the B31.1 code requires the use of the i factor for both weight and seismic stress calculations. The team notes that the use of a portion of a code just because it helps an analysis to document a tolerable condition is an unacceptable practice. This concern is directly related to unresolved item U3.5-1, which deals with the code of record for piping.

Basis - Paragraph 3.9.2.5.2 of the SQN FSAR states in part that ANSI B31.1-1967 is the piping analysis code of record, but that the operating condition stress equations in ASME Section III, Subsection NC-3000, Winter 1973 Addendum, may be used. However, TVA's use of the difference in the two codes to help qualify an analysis indicates that TVA believes that there is a significant difference in the B31.1 and NC-3000 requirements, which is contrary to the FSAR assertion that the B31.1 and ASME codes in question art equivalent.

Impact on Design - Use of this interpretation regarding the application of the stress intensification factor to qualify piping systems may result in a design that is overstressed and does not meet FSAR commitments.

References - 1. Summary of Analysis Report N2-67-2A, January 17, 1987.

2. ANSI B31.1-1967.

3. Section 3.9 of FSAR, " Mechanical Systems and Components."

4. ASME Section III NC-3000, Winter Addendum 1972.

CLASSIFICATION: Design Deficiency l

l 1

TTEM NO.: 03.5-2 (Continued)

- a; .. X-TITLE: Use of Selected B31.1 Code Rules (Continued)

y l l

RESPONSE

f s.

'N l

F 1

TVA agrees that the exclusion of the 0.751 factor was inappropriate. CAQR 7 l SQP871431IDI (reference 1) has been written to address this situation.

This condition developed because of an interpretation of the Codesusage of the

intensification factor as being applied for fatigue loadings only in ANSI ,

B31.1. Since the loading conditions for the equations where this factor was .f reduced did not contain a sufficient number of cycles to develop fatigue, the '

reasoning behind removal appeared justified.

This condition is bounded by the timeframe in which this type evaluation was performed (reanalysis performed after OL). A review of all problems J" reanalyzed since OL has been performed to identify all cases where the method described to reduce stress levels was used and only four cases were identified.

'Requalification of the four cases have been oerformed using 0.751 and refined analytical techniques. All pipe stress requirements were met. Calculations to support qualification of those specific points were generated and documented in reference 2. Those calculations will be incorporated in the individual calculation packages of record before restart of unit 2.

REFERENCES:

1. CAQR SQP871431IDI (S13 870916 814). ~'

2. DNE Calculations N2-03.09IDI-Misc ( later )

a

)

l

l' 3

_ }i y '"n 4- ITEy90.: 03.5-3 TITLE: Unistrut Clamp Load Testing t

SUMMARY

OF ITEM: ,,

~

hd TVA's use of the load rating rules of ASME Section III, Subsection NF instead

, of B31.1-1967 code of record is under current p censing review by the NRC's

! OSP. TVAcalculationSMG-060establisheddesignloads(ratedloads)baskfon the Unistrut load ratings determined by test. The NRC considers the calculations contain the following-errors ana omissions: .

1. The calculation did not apply the 10-percent reduction in the'loadaratingQv i as' required by NF.when only_one sample is tested.

2. The clamp relies upon friction to resist loading along the pipe axis. L .

However, the A-307 bolting material cannot be used for friction  !

connections ac:ording to AISC and NF.

3. The load ratings do not consider temperature effects. , ..

4. The 3-inch. clamp failed at approximately cne-half the load rating of the 2-1/2-inchand3-1/2-inchclampswithoutanyjustificationorexplanatyn for the failure. This anomaly should have been explained since it may i represent a manufacturing defect which could affect all clamp"$Jzes. , j

5. The load rating test report did not discuss the method of test i installation G This co7 sideration is important since these clamps can be, ,

installed in;c e field either tight (to restrain axial movement) or loose (to allow axial hovement). The resulting test load rating for each configuraticgcould therefore differ. 1 y.

5. In Revision'2igf the test report, TVA uses a square root of the sum of the squares formula >to determine the interaction of loadings. ThisVormula is l

inappropriate for this application (see Deficiency D3.3-5) Furthermore,  !

even a linear interaction equation may not be appropriate. For example,

,e the' combination of_ tension and axial loading will result in a reduction of ,.

the normal force between the Unistrut channel and the piping, thus 3

reducing the resistance to axial loading. n  ;

CLASSIFICATION: Documentation. I

RESPONSE

Item 1 - The original load ratings neglected the 10 percent reduction, as required by NF for a single test sample, since it was believed that at least 3 i tests per clamp size were performed for each direct *3cn of loading. However, the documentation which substantiates this cannot be located. <

l 1

r l

3 gh{

p- , - -

':+ , ,

'! TEM N0): '03. e-3 (Continued) t .  ; l TITLE: Unistrut Clamp Load Testing (Continued) l TVA has reviewed the original test results on the Unistrut P2558 climps (Reference _1) and compared these results with 3ther sources of test date on  !

the subject clamps (Attachment A). ThesesolyesincludebothUnistrut l (Reference 2) and TVA (Reference 3).

The referenced Unistrut Test Report documents test results which reflect three tests fo,"each clamp size and in each of the three primary directions /of l oadi'ng . The referenced TVA. test documents three tests per clamp sire in the axial loading direction for each of several finish combinations, and for J i

4 2-inch-diameter through 4-inch-diameter clamp sizes.

kl.

.r

'ti.

This comparison will be used to screen the original test data for / Malficant {

f 1

-h4pjpeintsf ni the data, when compared.to the average of larger or alternate l Q L, t?st,sampft. These high points could be the basis of unconservativa design j hJ cjampallowables. Therefore, additional testing and/or clamp der & ting will be l performed for4 py clamp size and loading direction which seems questionable, l before restart. t .  !

Upon completion of the additional testing, for the points in question l a l reevaluation of the specific allowables will be performed and the i documentation revised as required.

Item 2 - The bolted joint of a Unistrut P2558 clamp is not an AISC friction

, connection. The AISC limitation on the use of A-307 tiolts in friction j

connections is intended to apply only to standard structural friction joints and not to specialty details such as the Unistrut P2558 clamp.

This interpretation is in agreement with a recent ASME code interpretation (Attachment B) on a similar issue.

On October 7, 1987, TVA had telephone conversations with Dr. John Fisher l (Attachment C), professor at Lehigh University, and a cell known " bolting" {

contributor to the AISC specification; and also with John H. Bickford {

(Attachment D), Raymond Engineeri"q, Inc., a well knoin' expert in the field of l bolting. Both of these expert: :Socurred with TVA's p0sition that the clamp / joint in question was not within the M 0pt of an AISC friction connection. They also agreed that if the bcilts and installation procedures l are calibrated to produce a required preload, then the A-307 bolts are acceptable for use with the Unistrut P2558 clamps and that no significant stress relaxation problems would occur. I j, i

I bm 3 - TVA feels that the interim or sh t-term (restart) use of the  !

Unismit P2558 clamps as presently designed (with no consideration for j t,mperat:re effects) is acceptable based on the following: j l'

1. h. majority of these clamps are used in relatively low temperature appl M tic 7s.

ITEM NO.: 03.5-3 (Continued)

TITLE: Unistrut Clamp Load Testing (Continued)

2. The net effect of not having incorporated a temperature factor into the design allowables is, in the worst case, possibly a reduced factor of safety against a test ultimate loading; and not necessarily a reduced operational safety margin, as the majority of these clamps (including the high temperature applications) are not expected to resist design loads which approach, or are at the maximum design allowable loads.

3. The allowable design loads have been conservatively derived using static tests, for what are mainly dynamic loadings; as dynamic tests, with inertially generated loads, will generally yield higher allowable design loads.

To justify the continued or long-term (nonrestart) use of these clamps, T/A will perform a case by case evaluation of the clamps installed at locations for which significant temperatures are possible. This evaluation will include an appropriate derating of the allowable clamp loads for the actual design temperatures at each clamp location.

Item 4 - TVA agrees that the pull-out test failure of the 3-inch Unistrut P2558 clamp was premature. However, the allowable tensile capacity of the 3-inch Unistrut clamp is based on the substantially lower premature failure load, resulting in the existing design evaluations using this clamp capacity being conservative.

1 TEM NO.: D3.5-3 (Continued) ,

TITLE: Unistrut Clamp Load Testing (Continued)

To verify that the apparent premature failure of the Unistrut clamp is an isolated occurrence, a series of seven tests (Reference 4) have been completed on the same 3-inch clamp configuration for the same loading condition which produced the premature failure. It should be noted that a sample size of 3 is generally considered acceptable for this type of test program. It was decided, however, to double the sample size for this program and test 6 specimens. An extra, or seventh test was performed as a result of a deflectometer adjustment problem.

The minimum ultimate load attained was in excess of 9000 pounds at which time the bolt tensile failure occurred. This provides indication that the test load on which design capacities were based is underestimated by a factor of 2 to 3. Although these tests do not provide an explanation for the premature failure in the initial TVA clamp testing, these tests, in conjunction with the manufacturer's test data, do provide a basis for determining that the premature strap failure was an isolated occurrence and that the failure load on which the design load capacities are based is conservative.

In addition, TVA has contacted the Unistrut Corporation for information relating to industry experience with the product which might indicate a manufacturing defect. Unistrut's response (Attachment E) was that they have no record or recollection of any customer complaint for any failures on the P2558 series of clamps.

Item 5 - The Unistrut P2558 clamp is used in either of two configurations: 1) a three-way piping restraint or, 2) a two-way piping restraint with washers installed between the clamp " ears" and the unistrut. The latter configuration provides a gap around the pipe, permitting axial pipe movement. The clamp bolts in both configurations are installed with a torque equal to or greater than the bolt torque used in the original qualification tests.

The configuration used in the qualification testing of the P2558 clamp was the three-way restraint. The failure mechanism of the clamp for loads applied normal to the unistrut and pipe is bolt tensile failure. Since the method by which the bolts are loaded remains unchanged, the addition of washers would have negligible effects, if any, on the ultimate clamp capacity in this direction. The failure mechanism for a load applied parallel to the unistrut is bolt shear. Although the separation of the shearing surfaces is increased by the addition of the washers, the effect on the ultimate bolt shear capacity will be insignificant.

The two-way restraint with the washers was not tested. However, TVA considers the as-tested lateral capacities of the three-way restraint to be acceptable load capacities for the twsway restraint.

l

ZTEM NO.: 03.5-3 (Continued) l TITLE: Unistrut Clamp Load Testing (Continued) .

Item 6 - At the present time, TVA does not allow the use of a square root of the sum of the squares formula to determine the interaction of loadings on Unistrut P2558 clamps, but only allows the use of a linear / straight-line method of interaction for evaluating these clamps. This requirement became effective with the issuance of Quality Information Release (QIR) CEB-87-099 (Reference 5).

1 For the corrective action plan, extent of con;ition, and action to prevent recurrence, as applicable to the square root of the sum of the squares interaction formula, see tha response to deficiency D3.3-5.

TVA considers the linear / straight-line interaction of applied load components

,to be an acceptable methodology, based on the following: 1

1. The decrease in the frictional resistance on a clamp or channel surface, which results from the application of a piping / conduit load directed away from that surface, will be compensated by a corresponding increase in the frictional resistance at the clamp or channel surface on the opposite side

'f the pipe or conduit.

o

2. The linear addition of each of the three orthogonal load components  !

divided by the respective allowable loads is a standard procedure widely l used in industry, and commonly recognized by codes.

.3. For the normal loading conditions, a perpendicular or lateral pipe / conduit ,

load is not expected to exceed the clamp / bolt preload. Thus, the clancing l force should not be reduced, and therefore the axial restraint capacity of l the Unistrut P2558 clamps should remain unchanged.

REFERENCES:

1. TVA, Memorandum R. O. Lane to G. G. Stack, 7/28/75

2. Unistrut Corporation, Test Report C-36-A, 5/13/77  ;

3. TVA, SME, SQN-Axial Load Capacity Testing of Unistrut P-2558 i Clamps, Report SME-STR-87-001, 5/29/87, 846 870529 001. (

4. TVA, SME, SON-Static Load Tension Test of Unistrut P2558-30 Clamp, SME-CON-87-073, 9/8/87, 846 870909 001. ,

5. TVA, Quality Information Release (QIR) CEB-87-099, l 841 870710 250, 7/10/87 1

1 l-I q

__-_-___-_-_A

i ATTACHMENT A TDI 3.5-3 Unistrut P-2558 Test' Data j

. SLIP Along Direction (Fi ) l 1/2" 3/4" 1" 1 1/2" 2" 2 1/2" 3" 4" 5" I >

- 1975 TVA Test 1960 1710 1890 1670gl5940 5570 5580 6350 5990 1 Unistrut C-36-A 1640 1280 1260 1540 5500 7400 8200 8000 7350 1440 1400 1590 1800- 6000 7400 7100 700C 6650' 1740 1655 2455 1140 8900 9000 9000 8600 7500 Pull Out Direction (F2 )  !

1/2" 3/4" 1" 1 1/2" 2" 2 1/2" 3" 4" 5" 1975 TVA Test 2620 2430 2770 2540 6570 7460 3200 7220 7250 Unistrut C-36-A 2780 3015 2620 2900 6000 8700 9380 8850 9600 3090 3080 2720 3015 5000 8100 9550 8600- 9450 3030 3140 2600 3040 5000 8200 8350 8350 9450 l

1

's

ATTACHMENT A XD.I 3.5-3 ,

Unistrut P-2558 Test Data Slip Through Direction (F3) j 1/2" 3/4" 1" 1 1/2" 2" 2 1/2" 3" 4" 5" l l l l l l l l l 1975 TVA l500 l 510 l 230 l490 l2450~l2220 l4300 l 3640 l2970 Test l l l l l l l l l l l l l l l l l l j l l l l l l l l l Unistrut l575 l 760 l 460 l890 l1040 l1200 l3145 l1690 l1200 C-36-A l620 l 880 l 475 l710 l1060 l1300 l2290 l1275 l1220 l530 l 975 l 375 l940 l1225 l1040 )2300 l1280 l1100 l I I I I I I I I 1987SHETestl l l l l1622 l1698 l3841 l3208 l (B46870109001)l l l l l1651 l1237 l3377 l3240'l DataShownisl l l l l1515 l3209 l3761 l3240 l For Piping l l l l l2917 l3185 l3605 l 3657 l Only l l l l l2232 l1153 l3908 l 3160 l l l l l l2737 l3193 l 3605 l 3165 l l l l l l4245 l3893 l 3964 l 3629 l l l l l l4312 l4173 l 3796 l2613 l l l l l l 4332 l4253 l3798 l3377 l l l l l l4496 l 3648 l3781 l 3221 l l l l l l 3965 l 3689 l3781 l 3149 l l l l l l4333 l 3881 l 3637 l 3404 l l

l

Codes and StandaNs , E S?

The American society of Mechanical Engineers

• ATTACHMENT 3 345 East 47th Street New York. NY 10011 IDI 3.5-3 October 2, 1987 i

C. R. Angstadt The Cleveland Electric Illuminating Co.

PO Box 97 Perry, OH 44081

Subject:

Section III, Division 1; Appendix XVII 1974 Edition with the W 1975 Addenda.

Item: N!87-021

Reference:

Your letter dated March 31, 1987

Dear Sir:

Our follows: understanding of the questfons in your inquiry, ,and are as our replies Question (1):

to develop friction that will resist the assembly's rotation ab n'\0ng a pipe 7 or movement Reply (1): Yes.

Question 2:

colts are used in the manner as desciribed in questien 1 above Reply 2: Yes, See NF-4725.

Very tr 1y yours, '

evin Ennis Asststant Secretary, Boiler and .

• Pressure Yessel Commtttee XE/  !

IM.'U::,','C,*.;",.;'.,.;,;,

, - , ~ . , .., ',;,yp;(;;;;<;7,,gg;"ll";;ll?'l*,;,';;,';j,'.G,0;%;.;'j',7

%=.

ATTACHMENT C IDI 3.5-3 TO  : W. E. Pennell FROM  : K. S. Seidle DATE  : October 8, 1987

SUBJECT:

SEQUOYAH NUCLEAR Pl. ANT - RECORD OF TELEPHONE CONVERSATION A telecon was held October 7, 1987, with Dr. John Fisher and the TVA employees noted on the attached list. Dr. Fisher,' professor Lehigh University, is a well-known expert in the design of bolts and bolted joints. He is co-author of the prominent text on this subject, " Guide to Design of Riveted and Bolted Joints." In addition, Dr. Fisher has been substantial contributor to the AISC Specification in the bolting area.

The purpose of this telecon was to discuss the recent NRC IDI concern (Deficiency D3.5-3) pertaining to TVA's use of A307 bolts in the Unistrut P2558 pipe clamp. The basis for this concern is NRC's perception that the AISC code prohibits the use of A307 bolts in a friction-type connection. This perception is based on 1) the non-provision of allowable stresses for A307 bolts in friction-type connections and 2) the AISC Commentary statement "

. . the clamping force developed by A307 bolts is unpredictable and generally insufficient to prevent complete slippage at permissible working stress."

Initially, the clamp detail, method of application and installation torque requirements of the A307 bolts were described at length by TVA. Dr. Fisher responded by stating his position on several issues related to the D3.5-3 Deficiency. His responses are grouped by subject and paraphrased below:

Basis for AISC Provisions - In general, A307 bolts used in standard structural connections have uncalibrated installation requirements such as " snug-tight,"

etc. Therefore, any clamping force would be low and unpredictable.

Conversely, if the incta11ation of A307 bolts are calibrated and a resulting bolt preload or clamping force determined with consideration given to variability, a reasonable basis has been provided for utilizing the joint frictional capacity to resist applied load.

Unistrut P2558 Clamp as an AISC Friction-type Connection - The clamp would not be classified, nor recognited as an AISC friction connection.

A307 Allowable Stress Basis - A307 is a ductile material with an expected yield stress of approximately 33 ksi (The second edition of the above referenced l

DNE4 - 1113Q i NEB - 10/21/87 l

l i

l I

t _ _ _ _ _ _ _ _ _ _ - _ - - _ _ -

i ATTACHMENT C

~

2 IDI 3.5-3 W. E. Pennell SEQUOYAH NUCLEAR PLANT - RECORD OF TELEPHONIL CONVERSATION text provides. test data related to A307 yield and ultimate strength). There ~

are no reservations about preloading this type of bolt up to the yield point.

Freload/ Torque Relaxation - At preloads at or near the A307 expected yield, plastic flow would potentially result in small relaxations on the order of 5 to 10%. If preload stresses are limited to 50% of yield or less, negligible relaxation would result.

K. S. Seidle i

i s

DNE4 - 1113Q NEB - 10/21/87 l

i ATTACHMENT C-IDI 3.5-3 TELECON PARTICIPANTS John Fisher Lehigh University Karl Seidle CEB - Knoxville Dennis Dombroski CEB - Knoxville Karen McE1haney NEB - Knoxville i' Dennis Lundy CEB - Knoxville Bill Neely CEB - Knoxville Paul Guthrie NEB - SME1 C. R. Brimer PS - SQEP Scott Bargerstock OES - SQEP 4

DNE4 - 1113Q I NEB - 10/21/87 l

4

'l C___ --_ _ _ - . _ - _ _

j

ATTACHMEi4T D t

^

IDI 3.5-3  !

'alex: Mr. Carl Seidle tennessee Yelley Ai:thority Knoxville, Tennessua

\

Telex No. 6156322887 or 6156326869 After sent, please verify by No. 6156326399

Dear Mr. Seidle:

j Michael joint. Looram and I have reviewed the design of your Unistrut o ed~

b lt We have discussed the qualification procedure used on this joint 'with TVA personnel . We have also reviewed the purposes of thi joint.  ;

I It is our {

this joint.understn.,

The procedure ding that was a load as foirating ows,. method was used to qualify l 1

A sample of A307 bo]ts were torqued to failure. ,

2 required to fall the bolts. Sample joints were then assembled orque using 50 3.

The assembled in Singleton joints were then subjected to appropriate loads laboratories.

4 All joints so tested passed the tests. '

procedures, we have reached the following conclusionsAs a i

1.

This joint was properly qualified.

2.

A Unistrut joint should not be equated to a convent $onal slip critical speoiticationsstructural don't steel apply joint. to As thisajoint. result the AISO structural 3.

safe use of A307 bolts.In our opinion this application represents an appr^p John H. Bickford ,

i l

_ _ _ _ _ _ _a

I l

l j

UNISTAUT UN BTRUT CORPOMAT ON 350o5 MioHIGAN AVENUE WEST 1 ATTACl&fENT E N$c"l(NyaY[o"'o"#

TELEX: 23 D457 IDI 3.5-3 l

October 15, 1987 j TVA '

400 W. Summit Hill Drive 4 West Tower {

Knoxville, TN j l

Attention: Mri Karl Seidle Room W9Cl26 subject: P2558 Series Conduit Clamps Gentlement The P2558 series of conduit or pipe clamps were first designed in May of 1957. They first appeared in our '

catalog No. 5 in 1958.

Our records do now show nor can we recall of any customer complaints for any failures on this P2558 series of clamps.

Yours very truly, Ellwood Irish Chief Engineer EI/did ref 87-086 i

i

1. 4 e e aim e

C l

. ITEM NO.: U4.09'(D4.2-1)  !

_ TITLE: ERCH Access Cell Design

SUMMARY

OF ITEM:

l l ERCH access cells were analyzad as a monolithic. structure. However, the concrete shrinks away resulting in its behavior as individual cells.

This item documents the concern.that the original analysis of the ERCH access cells may be' based upon an unconservative assumption. .The seismic analysis may need to consider each. cell as acting independently.

. i CLASSIFICATION: No Deficiency

RESPONSE

Based upon the following design features, the analyst established that the monolithic assumption represents a reasonable engineering simplification to a more complex structure and the structural responses that were calculated represents conservative ~ approximations of actual behavior.

o A steel member was designed and placed in the concrete of the cells to assure horizontal interaction. ,

o The individual circular cells are separated by a filler cell that partially surrounds them on either side (much like a ball and socket joint).

'o The cells are embedded to varying depths in rock fill (100 percent to a minimum of 35 percent). Credit for the rockfill damping effect was not taken in the original analysis, o Although some concrete shrinkage has possibly occurred, we do not agree with the assertion that the concrete has shrunk away from the interlocking sheet piling sufficiently to cause the cells to act independently.

Considerable force is needed to remove steel forms from circular poured-in-place concrete columns. If the concrete shrunk away from its forms as asserted, the forms would readily separate from the columns.

(TVA documentation related to the construction of cells is under review to ascertain if any record of shrinkage-induced gap was observed.)

To further respond to the NRC concern, TVA has analyzed one individual cell acting independently to calculate dynamic movements and to show that the .i structure is stable (reference 1). This is a bounding calculation that further demonstrates the adequacy of the structure.

I 1

C , l ITEM NO.: U4.09 (D4.2-1) (Continued)

TITLE: ERCW Access Cell Design (Continued)

SUMMARY

OF ITEM: (Continued)

Additionally, a confirmatory time history analysis of one cell is being I performed incorporating non-linear boundary conditions. This analysis will l show the following: )

l o The structural stability of the cell during cyclic motion.  !

I o The state of stress at the cell-rock interface. (The interface  !

stresses are expected to be far less than capacities of concrete and rock).

o The amplification of the earthquake through the cell by calculating the response spectrum at the elevation of the ERCW pipes. The response spectrum is being developed to assess the ERCW pipe response although these pipes will not be affected by seismic motion ,

since adequate thickness of compressible material is provided {

surrounding the pipe.

This is a unique structure and no generic issue is identified. This issue has no impact on the safety of the plant and no hardware or design changes are required.

REFERENCES:

1. Calculation (R0 B41 870925 003).

l l

s ITEM NO.: 04.01 (D4.2-2)

TITLE: Seismic Analysis of Shield Building

SUMMARY

OF ITEM:

This item addresses the fact that some sheets in the seismic analysis calculation package had not been initialed as checked. The concern was that a mistake in tne calculation of Amplified Response Spectra (ARS) may have gone undetected.

CLASSIFICATION: Documentation

RESPONSE

TVA performed an audit in. January 1987. During this audit, it was identified that TVA's civil calculations for the seismic analysis of the Shield Building did not conform to the requirements of SQN-QAP-III-1.3 in that every sheet di~d not show a checker. TVA therefore initiated a review of all seismic calculations (reference 3). TVA has generated alternate calculations (B41

.870917 008) to verify the adequacy of the original seismic analysis of the shield building.

In addition, eight (8) other calculation packages associated with controlled reports have also been checked and reviewed using either alternate calculations or rigorous verification of original calculations. This verification reconfirmed the adequacy of these calculations.

No further action is required to resolve this item because 100 percent of the civil seismic calculations were reviewed.

This finding has no impact on the safety of the plant and no hardware or design changes are required.

REFERENCES:

1. Quality Assurance Procedure SQN-QAP-III-1.3, " Preparation, Review, and Records of Design Computations," March 8, 1970.

2. Calculation (B41 870917008).

3. Significant Condition Report SCR SQNCEB8714:

I 1

1 L__________ _ ____ _ _ ___

C ITEM NO.: 04.07 (04.2-2)

TITLE: Seismic Analysis of Steel Containment

]

SUMMARY

OF ITEM:

This item eddresses the fact that most sheets in the original horizontal analysis had not been initialed as checked. The concern was that a mistake in the calculation of Amplified Response Spectra (ARS) may have gone undetected.

CLASSIFICATION: Documentation (

RESPONSE

TVA performed an audit in January 1987. During this audit, it was identified that TVA's civil calculations for the seismic analysis of the SCV do not conform to the requirements of SQN-QAP-III-1.3 in that every sheet did not show a checker. TVA therefore initiated a review of all seismic calculations (reference 3). TVA has checked calculations (B41 870918 006) to verify the

'(

adequacy of the original seismic analysis of the SCV. ,

In addition, eight (8) other calculation packages associated with controlled - - -

reports have also been checked and reviewed using either alternate calculations or rigorous verification of original calculations. This verification reconfirmed the adequacy of these calculations.

l No further action is required to resolve this item because 100 percent of the l civil seismic calculations were reviewed and verified.

This finding has no impact on the safety of the plant and no hardware or design changes are required.

REFERENCES:

1. Quality Assurance Procedure SQN-QAP-III-1.3, " Preparation, Review, and Records of Design Computations," March 8, 1970.

2. Calculation (B41 870918 006).

3. Significant Condition Report SCR SQNCEB8714.

1 l l

\. .

i 8

ITEM N0.: U4.14 (D4.2-3)

TITLE: Vertical Seismic Analysis for the Steel Containment Vessel (SCV) j

SUMMARY

OF ITEM:

The amplified response spectra (ARS) for the SCV were regenerated as part of the corrective action for PIR SQNCEB86S2. Comparison of the revised spectra to the old spectra indicated significant increases in the peak accelerations.

Additionally, there was a concern that the TVA design basis computer programs (0YNANAL, et.al) used in the generation of seismic leads and ARS underpredicted the peak spectra values.

CLASSIFICATION: Documentation.

RESPONSE

A discrepancy in the vertical ARS between digitized computer files and published data was discovered in August 1986 and led to the filing of PIR SQNCEB8652. The following is a chronology of events in the resolution of PIR SQNCEB8652.

DATE EVENT 1, 1971/1972 A seismic analysis of the SCV was performed. The original (Sept-July) ARS in the vertical direction were obtained by taking 2/3 of the horizontal spectra. This analysis is documented in report CEB-75-3 R0.

2. 1979-1980 Task initiated to update the seismic analysis with generation of vertical response spectra. Objective was to reduce conservatism in original vertical spectra. New digitized spectra for the vertical direction were generated. The data was loaded into the GETSPEC non-QA piping database and later into the FRAMS QA piping database when it became operational. This data was used in subsequent design work.

3. Aug 1986 When the discrepancy between the spectra data in FRAMS and CEB-75-3 was identified, PIR SQNCEB8652 was prepared and issued; and corrective action was initiated.

. L_________-__-______-_______ __ __ . _ _ _

I B

ITEM NO.: U4.14 (D4.2-3)'.' (Continued)

4. 1986-1987 The spectra available in FRAMS were examined for (Sept-Mar) consistency with available documentation and the existing, unverified, vertical spectra calculations were reviewed in detail. A new analysis to generate the vertical response  !

spectra was performed.. The FRAMS database was updated to include the revised vertical spectra in March 1987 and report CEB-75-3 R1 was issued.

5. March 1987 Since the vertical spectra showed increased response over a to date limited frequency range due to a shift in peak frequency, 11 " worst case" piping analysis problems were analyzed for the new vertical seismic spectra. Evaluation of other attachments to the SCV (i.e., HVAC, electrical penetrations, cable trays, etc.) have/are being performed for the revised vertical spectra. :All activities associated with this work are scheduled for completion by November 15, 1987.

In parallel with the regeneration of vertical seismic ARS for the SCV, all other response spectra co7tained in the FRAMS QA database-have been verified and compared against current CEB reports and corresponding calculation packages. No deficiencies were identified during this comparison. This review is documented by calculation package B41 870918 007.

Based on work performed to date, there is no impact on the safety of the plant.

l Issue 2 - DYNANAL versus STARDYNE l

To resolve the question on the adequacy of the design basis computer programs (0YNANAL, et al) versus present day computer codes, the seismic analyses of l the shield building and the auxiliary-control building have been rerun using

! present day programs. The frequencies, dynamic participation factors, moce shapes and responses are consistent and within the normal range of accuracy for this type of analysis. Any nominal differences are attributable to differences in the algorithm used to calculate the responses. Calculations (B41 870918 008 and B41 870925 006) documenting this comparison are complete.

~

l There is no impact on safety and no hardware and design changes are required.

No further action is required to resolve this item.

L__-__-______

8 ITEM NO.: -U4.14 (D4.2-3) (Continued)

REFERENCES:

1. PIR SQNCEB8652

2. "Sequoyah Nuclear Plant - Dynamic Earthquake Analysis of the Steel Containment Vessel and Response Spectra for Attached Equipment," Report CEB-75-03 (R0 841 801107 001).

3. DNE Calculations, " Seismic Analysis of the Steel Containment Vessel and Generation of Response Spectra" (R0 B41 870310 007).

4. DNE Calculations, "TPIPE Data File Storage and Verification for Category I Structures" (R0 841 870918 007). l

5. DNE Calculations, " Shield Building - Seismic Analysis - l Comparison of Response Spectra" (R0 B41 870918 008). '

6. DNE Calculations, " Auxiliary-Control Building - Seismic Analysis - Comparison of Response Spectra" (R0 B41 870925 006).

e L _ _____- _

ITEM NO.: U4.ll (04.2-4)

Seismic model of the auxiliary control building does not match the concrete outline drawings. Walls are not included and columns have been neglected.

SUMMARY

OF ITEM:

This item addresses the fact that some minor partition walls and columns were considered to have negligible influence on the mass and stiffness when the seismic model was generated. The concern was that this may underpredict amplified response spectra (ARS) and structural loads.

CLASSIFICATION: No Deficiency

, RESPONSE:

TVA agrees that some minor partition walls and columns were not considered in the seismic model. However, these columns and walls add very little shear strength; any additional structural stiffness and mass would be insignificant. Rreiew of outline drawings have shown that all structural i walls and slabs have been accounted for in the seismic model. Some columns l and minor partition walls were not included. j A comparison of the checked hand-calculated properties (RO) with the properties (RI) computed by INERTIA indicated differences of less than 2% for area and moment of inertia. A noted difference in the cross-sectional area at mass point 12 was resolved after it was determined the checker had indeed considered the effects of 2 U line walls in the R0 calculation. Calculation 841 870917 004 documents this comparison.

All calculations generated to resolve this concern have been included in the .

original seismic analysis calculation. TVA procedures in the generation of l seismic models are compatible with standard industry practice. I The Auxiliary Building represents the worst case enveloping condition and no further action is required to resolve this item. This finding has no impact on the safety of the plant and no hardware or design changes are required.

l l

4 l

l 1

I l

C ITEM NO.: D4.03 (D4.3-1)

TITLE: Hydrostatic Uplift Lcads

/

SUMMARY

OF ITEM:

For design of the Auxiliary Building anchor rods, the hydrostatic uplift pressure was reduced by 1.4 k/ft* to account for the weight of the bu l.l di ng. The hydrostatic pressure rhould be reduced by only the dead weight of the structural slab and fill slab.

CLASSIFICATION: Minor Calculation Error

RESPONSE

The reduction in hydrostatic uplift pressure for the base slab design was incorrect. Problem Identification Report (PIR) SQNCEB8781 was written to document the discrepancy.

Additional calculations (references 2 and 3) have been completed which indicate that even with the correct dead load the stresses in the anchor rods and the base slab design are acceptable and meet ACI 318-63 code requirements. Additionally, these calculations also confirm that the anchor rods are not required for overall building flotation or seismic overturning stability and that the depth of the anchor rods into rock is adequate.

A review of the calculations for other Category I buildings with anchor rods (Reactor and Control Buildings) has been completed (reference 2). It was determined from this review that the dead weight considerations were properly accounted for in the design of the anchor rods for these buildings.

There is no impact on safety and no structural modifications are required.

REFERENCES:

1. Problem Identification Report PIR SQNCEB8781 (B25 870902 006).

2. DNE Calculations SCGIS163 (B25 870916 451).

3. DNE Calculations SCGIS164 (B25 870916 452).

l

1 l

l C

ITEM NO.: 04.04 (U4.3-2)

TITLE: Development Length of Anchor Rods )

l

SUMMARY

OF ITEM:

{

\

The length of embedment of the anchor rods in the Auxiliary Building base slab {

does not fully develop the anchor rods in accordance with ACI 318-63.

CLASSIFICATION: No Deficiency l

RESPONSE

The embedment of the anchor rods in the Auxiliary Building base slab was obtained using the ACI 318-63 code bond stress allowables. The total required embedment was determined and this length was bent to keep the reinforcement within the confines of the concrete as shown in Figure 1. This application was not precluded by the ACI 318-63 code and the resulting design was consistent with the industry practice.

Additional calculations were performed using the requirements of ACI 318-83.

The embedment requirements for reinforcing were significantly reduced by ACI 318-83 based upon more recent tests. (The tests are referenced in the Commentary to ACI 318-83.) Figure 2 shows that the embedment of the hook into the slab to develop the full strength of the anchor rod would be 21.6 inches using the provisions of the 1983 code for 3000 lb/in concrete 2

(actual strength of the concrete will be greater due to aging). The actual length provided is 21 inches. For the actual load the required embedment would be 12.8 inches for areas of the base slab with fill slabs and 13.6 inches for areas without a fill slab.

A review of the design for other Category I buildings with anchor rods (Reactor and Control Buildings) has been completed. This issue is not applicable to the Reactor building since the anchor rods for the Reactor Building are straight bars. Figures 3 and 4 show that the anchor embedment for the Control Building is adequate.

Calculations using the code of record and the current code have shown that the embedment of anchor rods in the slabs is adequate and that no structural modifications are required.

REFERENCE:

Calculation (B25 870916 450)

ATTACHMENTS: Figure 1 - Auxiliary Building - Anchor Rod Embedment - Full Development for WSD Normal Stresses ACI 318-63.

Figure 2 - Auxiliary Building - Anchor Rod Embedment ACI 318-83.

Figure 3 - Control Building - Anchor Rod Embedment - Full Development for HSD Normal Stresses ACI 318-63.

Figure 4 - Control Building - Anchor Rod Embedment ACI 318-83.

S M

O

- (C y R A

B P

O

_ T

)i 4 )O

))

s . bc S pi ((

0 sp 31 1 E 0 G 33 00 S 0 - S 4, O 1 1 O 8 E 1

MM R (2 O 3 3OO T

/c )

2 V T.

I TTI S V i

• a CCC EE fs'7/ 6 7 1

ASS TL s 5 4  : NA )

A4 3 l

( 3 S E EM S C GMR S

_ , N E

NDO N E

R I

f R DE 3T E 1LBD F

E E U I

MS 6S -

R y

R B DEW8D 1

. - 4" . s

f. . U YO R 3 NO s .'a GRR OIS FIAL R FCA G a 7, I

,,4 O T 2 .

l, h t l I

XH N N U C E I

. - o S ANM P (

U

)

, A O 2 . . .

.~

. -e- L.

- 4 E 3 ' ._ G V 4 '3 ~

= ..

ST H

E U 4 .

EG D

(

4

,. .'~

A 'N-' . ~. ~

TN AE CL L 0

I L

.. DD 4

D x

.r . * -

r NE I

R U

F Ru l

]

[

O

_ AO -

N BE O R a ,e

_ M - ,, I E s:

\

2 i.

LE OH TI e;~ / - L- ST

_ I D

L F F (SI( l(

1 )

$, 6

~

2 j

3 l

)

T[

N h.l)

=

e

)

l ET DD 6, g

4. is M E E / f T 4 0 1

( p P O

R D (2( I g(

N )0 0 L u l I j

E 0 - 0 V )el) f

\)

f f 2 c 0 0 E G O " n nn M 1 p 2 3 V E R  ; i i i P x 1 E R P 2 6 6 G O D 9 5 9. 5 L 7. v (7. / y E o o D [A,A (

0 (l(

1 I O t(

V _~ ' E * =

E h R h 1 D O d 8 L ld -

I U.

g T

O L

U E N F R E GM

, M D I

2

/ D E 3

2L B 3

=4 N 5 I M8 c y 2

1 E UB E 8-o

=M 2 R D 1

' 5 U YO 3 2 2. GRRI C

)

c n l*.

1 S7 l

I F A R A I

l n *'b: N M L

O

O O I

' L:

I I XM n

TT CC UC

'l EE SS AN

)

7 1

3 D A M

2 =

8- A m%'l'

- d 3 "4 2 H 84

. 2 1 ST 1

33 4 E4 I

U + TM C 52

(

~

~

AE A1 x ICL  :

4 _ * ' ' D 'D S 0 I kE E

. 1 R C 4 RU I

N D a ,

E A Q jO

,, B E R O  : E N >F  : D R I F LE E

• /.Y:

M . OH R ST E

T I

E

- D I

- l

R

_ E S

M O

(C R

'y A S

P O

_ - T

- 4) 1

_ (

) ))

5 S

_ 5 . bc

((

P i E 31 1 0 sp 6 33 00 S 0 - S 0 1 \

4, O O 8 MM E

. 2O 1

3OO R 3 T q ]['iV I I ITT S tr CCEEC

'y 6 77 ASS TL f 5 4 .

NA )

f 1

[ 3 S

E EM S C C, M R S

= , N MD O E E E N R A

I R

J E 3L D B D 3T F iMS 6S -

E R EuEW8D RB 1 7 UL D O R3N O

- 1

.l y- . GOR R FC OIS 7 I a 8

4 F TR A G 2 .

NO T N OH N I

- c- C C E S

- l NM U P (

. > A O

)

L 2

4 .' - E 3

4 i

9 ~

G

= .'.> .

ST2 H V E

U

(

4 .

A ~ '

EC TN D

. AE L CL L 4 . i DD 0 _

• U 4 -

r ME F

. I R

D

. I

? # /~ RO O

1 AG N BE n D R n .

M LE E

T

.S gf f

OH ST I

I D

I

6 9 1

. ) j 2  :

2 1

=

T =

D M

)

" N / A 1 E hDD j

6 4 M . 4 s / s P EE /

3 T dO ( P O RD 2 N O ) O t I 5 E 2 'c 0 O L UV 2 E QO #

M 1 'f 20 O 1

V ER " o i P x (1 - E RP G N O

O

7. V 7 l3 D s , '

5 I L

E O O.

D AA 1

( T C

V _^ = E = E E h g R h S D d g I d ,

A V /

L t G

2 T L

U E 3 N F R 5 E 2

1 GM N ND D E I

O l T I

B

,7 C 4L M83

%N I

E ly S EU E -

R B D18 y

i s0 ,

n , :.

2 ULO3

.- .~, - 5 G RORIC lC.s

. I 2

$ n i 1 F TRA 5 *'%. N 2 ~

O NO n

q T I

OH C CC E F. I N

_" = S .

)

7 1

1

. A 2 =

38 7

d N -

3 "4 2 H 8O 1

2 r'l sT t 4 1 E4 3T U = ... IC

( +

TM CE s AE CL AS s I  :

4 _ 5 ' # DD S 0 ME E

. I R C

_ 4 RU I

N D a 7

E b

AQ O >. BE R O

N f .:. -- D R t

E F

_ . . - LE E M '. oH R E

T l/ .>' ST

_ I

_ I D

I I 1 l

I C

ITEM NO : 04.06 (U4.3-3)

TITLE: Negative Moment Reinforcement in Halls and Base Slab

SUMMARY

OF ITEM:

Negative moment reinforcement was not provided in the Auxiliary Building base slab and walls placed against rock.

CLASSIFICATION: No Deficiency I

RESPONSE

The, base slab for the Auxiliary Building has reinforcement in the top face of I the slab only. The use of only top face steel is not in violation of ACI code requirements. The slab is structurally adequate for all design conditions. ]

The design of the base slab is entirely based on resisting hydrostatic uplift for the various design basis flood elevations. Interior loads on the slab are transmitted directly into the underlying fill concrete and rock.

To resist the hydrostatic pressures, the span of the slab was reduced by installing a grid of anchor rods into rock. This reduced the span of the 2 foot thick slab to a maximum of approximately 5 feet. The resulting moments in the slab were therefore relatively small. The reinforcement quantitles in the slab were determined for top face only.

Additional calculations (reference 1) have been performed on the slab which show that the top face reinforcement is adequate. The quantity is significantly greater than needed for the positive moment that would develop in the slab if all negative moment capability of the slab were lost due to cracking at support point (anchor rod locations). The calculations also indicate negative moment in the slab is not sufficient to crack the slab. The tensile stress in the concrete was 59 lb/in* versus the allowable of 119 lb/in* for an unreinforced footing (allowable for maximum probable flood).

Cracking of the slab could occur due to thermal' effects. The calculations indicate that were this to occur, the opening of the crack would be very small and would not adversely affect the anchorage of the anchor rod. If a thermal crack occurred at an anchor rod, the crack width which would result from free rotation of the slab at the anchor rod has been very conservatively calculated as 0.0016 inches (assuming a one way span). This conservative crack width is  :

about 16 percent of the normal ACI 318-63 crack width for exterior exposure, and 2 percent of the minimum height of the anchor rod deformations.

Walls of the Auxiliary Building placed against rock have re' enforcement in the l interior face only. Additional calculations assuming one-way simple spans show that the interior face reinforcement is adequate. The controlling load for design of the walls is hydrostatic pressure. Interlock of the wall with l the rock provides restraint to uniformly distribute temperature and shrinkage l- .

cracking. The ACI 318-63 minimum reinforcement for walls is not applicable since structural analysis shows adequate strength.

.____________-__a

( e i i o ->

c i

ITEM NO.: 04.06 (U4.3-3) (Continued) 1i The only other feature affected by (tnis concern is the control building base 1 slab. Additional calculations have\teer, prcpared which show the slab is s i

adequate for all design conditions. All other structures required for safe l shutdown have reinforcement in the faces of walls and s} abs placed against rock. ,

Thereispoimpactonsafetyandnomodificationsarerequired.

REFERENCES:

1. Calculations (B25 870916 452). :l

\

2. Calculations (B25 871021 450).

1

)

\ ,

i 7

.\

j j i

)

J h

l f

L o,

i i

ITEM NO.: D4.05 (D4.3-5)

TITLE: Shear Stress Calculations for Concrete  ;

H

SUMMARY

OF ITEM: {

l No calculations of shear stress were included in the calculations for the Auxiliary Building El 7:d.0 roof slab, and Al-line and A15-line walls.

CLASSIFICATION: Documentation RESPONCE:

Shear stresses were normally checked for critical areas of concrete walls and slabs such as exterior walls at low r elevations and larger span, heavily loaded slabs. Shear calculations were not documented for areas where it was l

, judged, based on a review of the shear forces and corresponding member thicknesses, that the shear stresses were not critical. However, to document this judgment, new calculations were generated which demonstrated that the shear stress for the most highly loaded areas of the Auxiliary Building EL 778.0 roof slab, and the Al line and A15 line walls are significantly below the allowable stress.

Approximately 294 critical sections / elements which may be subject to significant shear stresses were identified in the various Category I buildings. These critical sections / elements included walls, floor and roof slabs, base slabs, beams, and columns. A review of the calculations associated with these sections / elements identified that shear calculations existed for approximately 274 cases. In response to this item, calculations were made for the remaining 20 critical sections / elements to document the acceptability of the shear stresses. In every case for all slabs, walls, beams, and columns in the reviewed structures, the shear stresses were less than the allowables. This review is documented in calculation package (B25 870917 450).

There is no impact on safety and no structural modifications are required.

REFERENCE:

Calculation B25 870917 450.

ITEM NO : 04.12 (04.3-6)

SUMMARY

OF ITEM:

1. TVA did riot provide minimum horizontal steel in walls in accordance with ACI 318-63.

2. For the calculations for the A5 line wall the minimum steel was based on the effective depth of the member instead of the total thickness.

CLASSIFICATION: 1. Minimum Horizontal Steel - No Deficiency

2. Implementation - Minor Calculation Error 4 l

RESPONSE

TVA is in compliance with the ACI 318-63 requirements for minimum horizontal steel in walls. SQN wall designs were performed in accordance with Section 2201 Paragraph A of ACI 318-63. This section states that "The limits l of thickness...shall be waived...." Calculations show the walls to be structurally adequate. '

Minimum Horizontal Steel Used by TVA The 3 foot thick sections of the Al and A15 line wall used TVA Temperature and Shrinkage design standard to determine the area of horizontal steel to control crack widths. The use of this standard resulted in the use of 0.55 percent horizontal steel in the lowest lif t, 0.37 percent in the second lift, 0.30 percent for tae next two lifts, and 0.20 percent for the top 5 lifts. The lifts average approximately 10 feet. Therefore, application of the TVA temperature and shrinkage standard has resulted in code compliance when the -

average reinforcement over the height of the wall is considered.

Use of Effective Depth for Determination of Minimum Steel.

TVA has determined that the minimum quantity of steel for the A1, AS, All and A15 line walls was incorrectly calculated. The effective depth "d" (the distance from the compression face to the tension reinforcing bars) instead of the total thickness "t" of the member was utilized. Uting the effective-aepth of the member instead of its thickness slightly underestimates the amount of reinforcement. .

1 Calculations for structural slabs and walls in the Auxiliary Building, Reactor Building, ERCH Pumping Station, and Control' Building have been reviewed

( ). Occurrences of the use of "d" for calculation of minimum i steel were reexamined and adequate reinforcement has been provided.

I i

B  ;

. ITEM NO.: U4 10 '74.3-7)

-TITLE: Roof Fi'exibility

SUMMARY

OF ITEMS:

The. seismic design of the Auxiliary BLilding roof systems assumed the roof to be rigid and used the corresponding vertical acceleration from the response spectra. However, it is not a rigid structure and should not be analy ed that way.

CLASSIFICATION: Minor Calculation Error. . a RESPONSES:

TVA agrees that the roof of the Auxiliary Building is not rigid. However, the original design was consistent with industry practices which was to utilize ,

the roof ZPA. The increase in vertical loading due to the increased vertical response is equivalent to 17 psf which is 6 percent of the original roof

. design load. LDue to the long spans of this structures resulting in lower

-frequency and the increased magnitude of the floor response at this elevation, the percent increase due to flexibility for this roof system is potentially the highest that will be seen for any floor or roof structures.

An. issued calculation has determined that the roof design has adequate margins to meet design-requirements considering this additional lead except in the

-areas of roof-mounted tanks which are being addressed .in issue D4.3-8.

The generic implications of this issue on main steel building framing will be dispositioned by CAQR SQP871386. Final Calculations will be available by November 2, 1987.

REFERENCES:

1. CAQR SQP871386 IDI

2. TVA Design Calculation - SCG-CSG-87-189 - Auxiliary Building Roof Framing EL 791' - Resolution of CAQR SQP871386

l l

l ITEM NO.: U4.17 (D4.3-8)

IITLE: Tank Overturning Moment

SUMMARY

OF ITEM:  !

The analysis of the framing which supports the tanks on the Auxiliary Building roof did not consider the overturning of the tanks in the design of the framing.

CLASSIFICATION: Documentation RESPONSES:

The Auxiliary Building roof supports four Seismic IL (two-over-one) tanks.

This issue is related to the seismic forces exerted by these tanks on the supporting structures.

The original design considered the vertical tank seismic loads in the design of the roof. It also considered horizontal tank seismic loads based on roof ZPA in the design of the pedestals which support the tanks. This was consistent with the common industry practice at the time of the seismic design for non-Category I tanks.

However, the overturning moments were excluded in the design of the roof slab and supporting trusses.

Seismic forces from the tanks were determined in an analysis of the tanks which included modeling a significant portion of the Auxiliary Building roof and thereby. incorporated the roof flexibility cencern (IDI D4.3-7). This seismic response spectrum analysis was performed using damping values defined in the FSAR. The tanks were analyzed at their maximum fill level for normal operation to ensure an upper bound load prediction (reference 3).

The dead and seismic forces from the tanks and roof dead, live, and seismic loads were applied to the roof structure through a conventional framing analysis. Calculations performed to date indicate that all affected aspects of the Category I roof structure (beams, trusses, and concrete roof slab) meet the design requirements. (Final calculations by November 2, 1987.)

The concrete pedestals and anchor bolts which support the tanks are considered Category IL (seismic non-category 1) because they support Category IL tanks.

The design calculations for the concrete pedestalt and anchor bolts for the raw service water tanks have demonstrated that the pedestals meet design requirements. Calculations performed to date for the concrete pedestals and anchor bolts for the demineralized and the cask washdown tanks for the SSE loading indicate that the pedestals and the anchor bolts have sufficient ductility so that failure of the pedestal will not occur. (Final calculation by November 2, 1987.)

I

A ITEM NO.: U4.17 (D4.3-8) (Continued)

TITLE: fank Overturning Moment (Continued) l l

Further analysis of the demineralized and cask washdown tanks is being '

performed to verify that these tanks will remain intact and stay in place under design basis loading and that there will be no impact on the Category I roof. (Final calculation by November 2, 1987.)

i The effects of tank overturning moments on tank anchorage is addressed I generically in the response to IDI D4.6-2. j

('

REFERENCES:

l. CAQR SQP871387 1

2. TVA Design Calculations - Auxiliary Building Roof Framing SCG-CSG-87-188

3. DNE Tank Analysis Calculations (041 871026 002, B41 871026 003, and B41 871026 004)

, J A I

~

H ITEM NO.: U4.l'3 (04'4-1)-

TITLE: ERCH Missile Protection System

SUMMARY

OF ITEM:

-The calculations for th'e'ERCH Pumping Station missile protection were jincomplete in that they do not explicitly cover the adequacy of the beam i grillage system as it relates to beam shear resistance, web crippling, and flange buckling. In addition, other missile impact angles, impact locations, l and penetration resistance to smaller missiles should be evaluated. The potential generation ~of secondary missiles should be addressed in the

. calculations.

]

CLASSIFICATION: ' Documentation'.

RESPON5E: l TVA has completed calculations (B41 870916 002) to address the issues raised by the IDI team with the exception of impact at the end of.the beams.

Additional. calculations have been initiated to evaluate.the beam-end impact.

Results of these calculations to date indicate that for the midspan vertical missile impact, several beams interact to stop.the impacting missile. These calculations are based on the conservative assumption that impact energy will be-absorbed by weak axis bending only. Further analyses to verify these .

conclusions and to address beam end-impact have been initiated and are- i scheduled for completion by November 30, 1987. Calculations which demonstrate l the adequacy of the missile protection for smaller missiles have also been  !

completed.

Restart Significance:  !

The additional analysis completed to date, coupled with the extreme low probability of tornadoes in this area (fall and winter seasons), provide adequate assurance of the structural adequacy of the missile protection system for plant restart. TVA will perform any additional analysis of the missile protection system which is considered necessary to further assure the adequacy of the missilo protection system as a postrestart action by April 1, 1988. ,

REFERENCES:

1. PIRSQNCEB8785

2. Calculation B41 870916 002) f

l ITEM NO.: U4.10 (D4.4-2) 1 l

TITLE: Dike Stability _i

SUMMARY

OF ITEM:

During a review of_the ERCW' piping support slab a question was raised l regarding the stability of the rock fill ERCH access dike in which the slab and piping are buried and the consequences to the piles supporting the slab should a slope failure occur in the dike. Subsequer.t questions related to the ,

issue include:

1. Why are factors of safety less than the minimum required safety factor' ]

apparently accepted for some sections? q

2. What'is the justification for the higher material property (0 - 45*) {

used?

3. Was a vertical earthquake coefficient used in the stability analysis?

4. Was the probable maximum flood (PMF) used in the analysis?

5. FSAR should be reviewed to show that a friction angle (0) of 45* was used for the rockfill.

CLASSIFICATION: Documentation

RESPONSE

The original. ERCH access dike calculations were 'done using a conservatively assumed value of internal friction angle-(0) = 35' for the rockfill material. Using this assumption, there were two instances out of the more than 45 combinations of loading cases and cross sections analy:ed that had factors of safety (FS) less than required (Criteria Requirement 1.05:

calculation 0.97 and 1.02). Additional analyses performed at that time for -i' those two cross sections indicated that a O c? 39.5* would be necessary to meet the minimum required FS. Concurrent with the analysis, the actual rock fill material to be used for constructing the dike was being tested by the Corps.of Engineers South Atlantic Division Laboratory. The test results )

indicated that 0 - 45' (B41 870312 004). These test results were included in the original calculations but no definitive statements were made to clearly document what was done or the decisions made.

A recent revision of the calculation has been made (B41 870910 006) which includes statements to correct the documentation deficiency described above.

The critical case for each cross section was also recomputed using 0 - 45* ,

for the rock fill. The minimum FS obtained was 1.12. Therefore, the design of the dike is acceptable.

1

)

, i i

______.___________________________________J

TTEM NO : U.410 (04.4-2) (Continued)-

, TITLE: Dike Stability (Continued)

RESPONSE: (Continued)

The FSAR will be revised to reflect - 45' for the rock fill material at the next annual update.

! i The original calculations did not consider vertical earthquake accelerations.

The pseudo-static analysis procedure was used with only the earthquake peak horizontal acceleration values. This approach is considered to be adequately I conservative and a standard industry practice. Justification.for the use of only the peak horizontal acceleration is provided in the revised calculation.

This justification consists of standard Corps of Engineers practice,

. literature by professionally recognized individuals and A-E firms, and data derived from 40 actual earthquakes (see references 1 through 4). A CAQR (SQN 8714101D1) has been issued to reflect the discrepancy between the FSAR (FSAR states vertical component was used) and the calculation package. The FSAR will be revised at the next annual update to reflect the calculation method l

actually'used. l The effect of the PMF was not addressed in the original calculation. The revised calculation evaluated the effect of the PMF upon the stability of the dike. (It should be noted that considering the PMF to act simultaneously with the SSE is not a required design case.) The revised calculation shows that the water levels associated with the PMF will not affect the structural integrity of the dike because: (1) the dike is constructed of highly i permeable material which prevents the development of differential water I pressures and (2) the reduction in effective material weight (buoyancy) as a  ;

result of submergence affects both the driving and resisting forces such that l there is essentially no net effect. For the seismic load cases, the horizontal seismic coefficient was applied to the total weight (rock and entrapped water) when computing horizontal forces resulting from SSE.

This issue has no impact on the safety of the plant and no hardware or design changes are required.

4 This is a unique structure and therefore, has no generic implications.

i

REFERENCES:

1. Corps of Engineers, " Engineering and Design Stability of ]

Earth and Rock-Fill Dams," Manual No. EM 1110-2-1902, Office of the Chief of Engineers, Dept. of the Army (1970).

2. Seed, H. B., and Martin, G. R. (1966) "The Seismic Coefficient in Earth Dam Design," JSMFD, ASCE, Vol . 92, No. '

SMe, May pp 25-58.

3. Sarma, S. K. (1975) " Seismic Stability of Earth Dams and Embankments, "Geotechnique 25, No. 4, pgs. 743-761.

l l

ITEM NO.: U.410 (04.4-2) (Continued)

TITLE: Oike Stability (Continued)

REFERENCES:

(Continued)

4. Calculation (B41 870910 006)

5. TVA Condition Adverse to Quality Report (CAQR ) No.

SQP871410I0I.

6. Calculation (B41 870312 004).

l I

l l

.i A-

" ITEM NO.: 04.18 (04.6-1)

TITLE: Equipment Drawing versus Calculation Discrepancy

SUMMARY

OF ITEMS:

The design drawings for equipment foundations and supports are not in agreement-with the calculations.

CLASSIFICATION: Minor calculation error, i

RESPONSE: j i

The foundation embedment designs for the following equipment were included in i the IDI review:

1. Component Cooling Hater Heat Exchanger (CCHHx) -

TVA agrees that the baseplate thickness shown on the drawing (48N1269 R11) at the top of the concrete pedestal for the CCHHx is less than required by the calculations. A 1-/2-inch thickness is specified on the drawing while the' calculations require 3/4. inch. CAyn SQP871450IDI has been written to address this discrepancy.

This issue is being resolved in conjunction with NRC IDI D3.06 (D3.4-3).

The CCHHx has been reanalyzed, and reverified loads were used to qualify the embedded anchorages. The 1-/2-inch embedded plate meets design criteria requirements.

2. Component Cooling Hater surge Tank (CCHST)

TVA agrees-that a drawing discrepancy exists between the CCHST support calculations and the drawing (48N1271). Specifically, the drawing detail is unclear with respect to number and spacing of studs required.

The CCHST anchorage has been reviewed in response to NRC IDI D4.6-2 (tank 12 of attachment to IDI D4.6.2). The anchorage details on the design drawing have been evaluated with reverified loads, and the anchorage meets design criteria requirements.

1 i

I

A ITEM NO.: D4.18 (D4.6-1) (Continued)

TITLE: Equipment Drawing versus Calculation Discrepancy (Continued) i

3. ' Containment Spray Heat' Exchanger (CSHx)

TVA agrees that the diameter of the concrete anchors and the plate thickness shown on drawings 48N1266 and 48N1267 for the CSHx anchorage are less than required by the calculations. This discrepancy had been 4 previously identified by TVA under Detailed Technical Review of Miscellaneous Steel Structures (SQN-CEB-87-02). As a result of this review, CAQR SQP870188 was written against the heat exchanger support structure.

The CSHx and its support structure are bcing evaluated and are scheduled for completion by November 30, 1987. ,

Generic Review

/

The generic issue of this IDI finding is that design drawings for equipment foundations and supports are not in agreement with the j calculations and may result in structurally inadequate supports. To  !

evaluate this issue generically, equipment has been classified as tanks, '

heat exchangers, and other equipment. I 1

Category I tacks identified by the Design Baseline Verification-Program, I as required for safe shutdown, have been reviewed for. calculation and I drawing compliance and supplemental calculations have been developed where required in resolution of NRC IDI D4.6-2.

l Heat exchangers are similarly being reviewed in' response to NRC.IDI D3.06 )

and D3.12.

The remaining equipment is being evaluated by reviewing an engineered sample of 60 supports for calculation / drawing compliance. Any discrepancies that are identified will be resolved through supplemental calculations, and conclusions will be drawn regarding the remainder of the l population. This work will be completed by November 30, 1987. 1 i

i I

i 1

I 1

I

A ITEM NO.: D4.20 (04.6-2)

TITI.E: Tank Anchorages

SUMMARY

OF ITEM-Shear was not considered in the design of the anchor bolts for the component cooling water surge tank.

CLASSIFICATION: Documentation.

RESPONSE

The specific issue raised by the IDI finding was that the vendor calculation did not consider shear-tension interaction in the design of the threaded rod for the component cooling surge tank. Additionally, NRC IDI finding D4.6-1 identified a discrepancy between anchor sizing and spacing for the same tank.

To resolve these specific issues, the design drawing, vendor calculations, and TVA calculations were reviewed and the TVA calculations were revised as necessary. The calculations regenerated also resolve any discrepancies between drawings and calculations. The calculations that have been completed l demonstrate that the existing design of threaded rods and anchorage meets design criteria requirements. Final calculations will be completed by October 30, 1987.

To resolve the generic issue of considering shear-tension interaction, all anchorages for Category I tanks identified by the Design Baseline Verification Program as required for safe shutdown were reviewed (see attachment to 04.6-3). New calculations for the anchorage were generated utilizing reverified loads. For all tanks, the threaded rods and embedded anchorages were shown to be adequate when compared to design basis allowables. Final calculations will be completed by October 30, 1987.

The generic issue of tank overturning is being resolved through the evaluation of the tank anchorages including overturning effects. Final calculations will be completed by October 30, 1987.

The design of the tank anchorages reflect no impact on safety and no structural modifications are required.

1 i

1 i

l ITEM NO.: 04.19 (D4.6-3)

TITLE: Seismic Analysis of Steel Tanks

SUMMARY

OF ITEM:

Category I tanks which are required for safe shutdown may not be gid and ,

should be reviewed to ensure seismic adequacy of the tanks and the r anchorage. This issue was defined on the basis that the original .endor calculation for the component cooling water (CCH) surge tank neglected shear deformation when determining the fundamental tank frequency. The Category IL tanks on the Auxiliary Building roof are discussed in TVA's response to item U4.17 (D4.3-8).

CLASSIFICATION: Documentation

RESPONSE

In response to this issue all seismic Category I tanks required for safe shutdown (DBVP Phase I tanks) have been reviewed. This review included determination of fundamental frequencies and the appropriate accelerations to generate seismic loads to the anchorage. TVA's response to item D4.20 (D4.6-2) addresses the tank anchorages.

Twelve tank designs were identified for review (attached list). Original dynamic analyses were retrieved and supplemental calculations were performed as necessary. The analyses yielded fundamental frequencies (including bending and shear deformation effects) and seismic accelerations were obtained. This review confirmed that the the tank designs used appropriate seismic acceleration levels. Of the twelve tank designs, seven (numbers 2,4,6,7,9,10, and 11) were nonrigid (F less than 25 Hertz) and are appropriately analyzed as nonrigid tanks. Sloshing effects are considered where appropriate. All documentation of this review and the supplemental calculations have been issued (reference 3).

The refueling water storage tank (RHST) and its foundation are the only soil-supported category 1 tank structures at SQN. Soil structure interaction was appropriately considered in the seismic analysis of the RHST system (reference 4).

The SQN tank design criteria, SQN-DC-V-13.6, will be revised to permit nonrigid tank assemblies by November 2, 1987. There is no SQN FSAR commitment for rigid tanks. This response clarifies TVA's response to NRC Inspection Report 50-326/327-86-27 item D3.3-5 (reference 2) in that regard.

The reviews and confirmatory calculations have shown that the seismic category I tanks listed in the attached list have been properly desigr.ed. Therefore, this issue is considered closed and no additional work will be performed.

There is no impact on safety and no hardware changes are required.

REFERENCES:

1. SQN-DC-V-13.6

2. TVA's response to NRC Inspection Report 50-326-327-86-27 dated July 28, 1986 (L44 860729 801)

3. Tank Assembly Calculations (B41 871013 007)

4. RWST Calculations (B41 8706323 010)

ATTACHMENT: , List of Category I tanks required for safe shutdown

l CATEGORY I TANKS REQUIRED FOR SAFE SHUTDOWN OF UNIT 2*

1. 68,000 Gallon, 7-Day 011 Tanks for Diesel Generators (2 Embedded Tanks in Concrete Below DG Building Floor)

2. 550-Gallon Day' Tanks for Diesel Generators (4 Skid Mounted Tanks With  ;

Bruce GM DG Package)  ;

3. ' Starting Air' Tanks for Diesel Generators (4 Vertical Tanks in DG '

Buildings)

• • 4. Haste Gas Decay Tanks (5 Vertical Tanks at EL.669 in Aux Building)

5. Control Air Receiver Tanks (2 Vertical Tanks at EL.734 in Aux Building)'

• • 6. Boron Injection Tank (1 Vertical Tank at EL.690 in Aux Building)

• • 7. SIS Accumulator Tanks (4 Vertical Tanks at EL.693 in Reactor Building) t

• • 8. UHI Surge Tank (I Horizontal Tank at EL.740.5 in Additional Equipment Building) I

• • 9. UHI Hater Accumulator Tank (1 Vertical Tank at EL.706 in Additional l Equipment Building)-

• • 10. UHI Gas Accumulator Tank (1 Vertical Tank at EL.706 in Additional ,

Equipment Building)

11. Refueling Water Storage Tank (1 Soil Supported Tank in Yard)

12. CCH Surge Tank (1 Vertical Tank at EL.734 in Aux Building) l I

i i

• DBVP Phase I Tanks

• • These tanks were in Westinghouse's scope of supply. .

l 1

)

I 1

1 l

1

-- _ _ _ _ - _ _ - - - - _ _ _ J

B

. ITEM NO,: D4.15 (04.7-1)

TITLE: Use of Rawl Anchors

SUMMARY

OF ITEM:

Discuss reduction in strength for Rawl versus Phillips. . Discuss apparent discrepancy in 79-02 report if Rawl anchors were used. What allowed construction to use Rawl anchors as substitute for Phillips?

CLASSIFICATION: Documentation

RESPONSE

CAQR SQ870101 issued on June 16, 1987, identified discrepancies in the-documentation relating to the use of Rawl brand self-drilling expansion anchors at SQN. 'The discrepancy was identified by the NRC during an audit of

.the employee concern program. TVA's response to NRC OIE Bulletin No. 79-02 states that "to the best of our knowledge no Rawl anchors were used at SON."

However, Rawl anchors are called for on drawings for pipe supports that were prepared for TVA by Basic Engineers. Also, a memorandum prepared by the construction division indicates that Rawl anchors were used.

. Based on the investigation and tests discussed in the following text, TVA has concluded that the use of Rawl anchors in lieu of Phillips was authorized by design and that the Rawl anchors are effectively equivalent to the Phillips anchors.

Use of Rawl Anchors in Lieu of Phillips TVA General Construction Specification No. G-32 covered the installation of expansion anchors. It was issued in 1972 and was therefore applied to virtually all anchor installations at SQN. The specification required the use of Phillips Self-drilling anchors or equal. It was found during our review that Phillips self-drilling anchors were substituted for Rawl anchors where Rawl anchors had been specified on design drawings.

No formal documentation has been located that specifically instructed the use of Phillips in lieu of the Rawl anchors called for on the Basic Engineers drawings. However, TVA Civil Design Standard DS-C6.1 for Anchorage to Concrete listed " approved" self-drilling anchors. Phillips and Rawl anchors were both on the list. This list and the requirements of G-32 were the probable basis for the use of Phillips anchors.

1 D4.15 (D4.7-1) (Continued)

Manufacturer's Test Data The manufacturer's data for the Rawl anchors indicates that some sizes of the Rawl self-drilling anchors have capacities less than Phillips self-drilling anchors. However, the tests performed for the manufacturer on the Rawl anchors exhibited failure by splitting of the test blocks. This mechanism for failure would not occur in anchors installed in slabs or other members with adequate edge distance. Therefore, the manufacturer's tests are not directly applicable to the conditions that exist at SQN and the Rawl anchors at SQN wodJ be expected (see attachment) to develop greater strength than those obtained in manufacturer's tests.

1 Usage of Rawl Anchors at SQN l Although the Basic Engineers drawings call for Rawl self-drilling anchors, l Phillips anchors were probably used for most of the installations. The consensus of personnel knowledgeable in anchor usage is that Phillips anchors were generally used although some Rawl anchors may have been used during the early stages of construction.

No evidence was found in TVA's procurement records to indicate purchase of the Rawl anchors. however, this is not considered to be conclusive evidence.

Anchor proof load records are available. However, TVA inspection procedures did not require the brand of anchor to be recorded on the report until 1979.

Proof load reports after 1979 indicate use of only Phillips anchors.

A support which was installe.d in 1975 was inspected to determine the brand of anchor installed. The support (1APBH-425) was designed by Basic Engineers and called for Rawl anchors. The inspection of the anchors showed that the anchors were Phillips (as indicated by the distinctive red cone expander).

TVA Teuing of Rawl Anchors Since it is not possible to determine conclusively that the Rawl anchors were or were not used, testing was perforned by TVA to determine if the Rawl anchors meet the G-32 requirements. The tests compared the capacities of the Rawl anchors to the Phillips anchors.

Basic Engineers performed ths original design for pipe supports outside cor. tai nment . Basic Engineers drawings called for Rawl ancnors for about one-half of the supports cnd Phillips for the other half. Approximately 5 perceret of the drawings call for 7/8" anchors, 25 percent 3/4", 20 percent 5/8", 50 percent 1/2", and less than 1 percent 3/8". During the on-going regeneration effort for pipe supports, no 1/4" anchors have been identified.

. _g7_

1 B

04.15 (04.7-1) (Continued)

The tests were performed in a test slab at Bellefonte Nuclear Plant since an existing slab was available and the concrete properties are similar to SQN.

The concrete was made with a similar coarse and fine aggregate (dolomitic limestone). The actual properties of the concrete are not of particular significance since the tests are intended only for comparison of the two j anchor brands. The concrete at BLN would also be more representative of SQN I concrete than concrete used for testing by a manufacturer in another part of the country.

i Parallel tests were performed on Phillips and Rawl self-drilling anchors. J

-Three anchors of each size and brand were loaded to failure to determine the 1 ultimate tensile capacity. The ratio of the average results of tne Rawl and j Phillips anchors are given in the attached table for the Rawl sizes called for J by Basic Engineers. The ratios vary from 0.94 to 1.05. One-quarter inch I anchors were tested before determination that they were not used. For the 1/4" anchors the Rawls had lower capacity (Ratio - 0.89).

The range of results is similar to the ratios that are obtained by comparing i the manufacturer's results to the G-32 requirements. However, the sizes above )

and below the G-32 requirements are almost reversed. For the manufacturer's tests, the 5/8-inch and 7/8-inch were greater than 1.0 and the 1/2-inch and ]i 3/4-inch were less than 1.0. This indicates that deviations in results in the order of 5% would be expected. Therefore, deviations of this magnitude are not significant. i In order to provide evidence that the Rawl anchors meet the engineering requirements, the attached table also provides the ratio of the Phillips anchor capacities to the G-32 requirements. These ratios are based en the )

qualification test results obtained for the Phillips anchors at Watts Bar Nuclear Plant. Again, this concrete is similar to the concrete at SQN and would be more representative of the concrete than a manufacturer's tests.

The attached table shows that the Phillips anchors exceed the ultimate tensile capacity requirements of G-32 by about 10%. The table further shows that the calculated ratio of the Rawl anchor capacities to the G-32 requirements based i on the two sets of tests exceeds the requirements.

Application of Increased Concrete Strength to Anchor Capacity i The design loads for expansion anchors provided in TVA Design Standard DS-C1.71 are based on a compressive strength of 3000 psi. This is the minimum specified strength used for structural members. Since the actual specified strength of the concrete in which the anchors are installed may be greater than 3000 psi and since the concrete gains significant strength with age, the design standard allows the use of strengths greater than 3000 psi for the l

evaluation of existing anchorages. t B ,

D4.15 (D4.7-1) (Continued)

The NRC inspector noted that an increase-in anchor tr1sile capacity was used for a pipe support calling for Rawl anchors even the y h the manufacturer's tests for-Rawl anchors were performed in 4500 psi concrete. The inspector

' concluded that the. increased capacity was inappropriate.

l The strength increase is appropriate because the qualification of the Rawl j anchors was based on their equivalency to Phillips anchors and the Phillips l anchor qualification is based on their capacities 1.n 3000 psi concrete. I The allowable loads for self-drilling anchors used by TVA were originally  !

based on the Phillips manufacturer's tests and tests performed by TVA on two sizes. The TVA tests were performed in 1972 in concrete with a strength of 3270 psi. TVA applied an across-the-board decrease to the manufacturer's capacities for all sizes based on these tests. The HBN qualification tests, mentioned in the previous paragraphs, were performed on concrete with an average strength of 3265 psi. Therefore, the desipq loads provided in the TVA standard are based on 3000 psi for both the Rawl ahJ Phillips anchors.

The use of a capacity increase for installed expansion anchors is recognized in NRC IE Bulletin No. 79-02. The Bulletin accepts an increase based on the quality control tests for the concrete. Since the quality control tests are on moist-cured specimens, TVA has limited the increase that can be used for small members which may not achieve the increase indicated by moist-cured cylinders due to drying. Specific provisions are included in DS-C1.7.1.

Conciusion The Rawl and Phillips self-drilling anchors are effectively equivalent.

Deviations in capacity for some sizes of approximately 5% are compensated for by the fact that Phillips anchors have been shown to exceed the minimum required capacities by approximately 10%.

There is no impact on safety and no modifications nor design changes are required.

REFERENCES:

1. Test Report - Comparison of Rawl to Phillips Anchors SME-CON-87-065 (B46 870821 003) - Tests performed at Bellefonte.

2. Hatts Bar Bulletin 79-02 Report - Appendix G - Ultimate Anchor Capacity (CEB 841210 002) - Hatts Bar qualification results.

3. TVA General Construction Specification No. G-32 (B41 870701 040),

4. TVA Civil Design Standard DS-C6.1 (See Appendix A of Reference 2).

5. TVA Civil Design Standard DS-C1.7.1 (841 870728 009) l

B D4.15 (04.7-1) (Continued)

SEQUOYAH NUCLEAR PLANT COMPAP! SON OF TESTS ON RAWL AND PHILLIPS ANCHORS' Size Ratio ei Ratio #2 Ratio #3 Raw 1/Phillips Phillips/C-32 Rawl/G-32 3/8 1.05 not tested -

1/2 0.95 1.09 1.04 5/8 0.95 1.15 1.09 3/4 1.04 1.08 1.12 7/8 0.94 1.09 1.02

1. 1 - Based on comparison tests at BLN.

1. 2 - Based on qualification on' tests at WBN.

1. 3 - Product of #1 and #2.

l l

l

-100-

1 TEM NO.: D5.2-3 TITLE: Ineffective ERCH Alarms

SUMMARY

OF ACTION:

The NRC noticed on their July 28, 1987, walkdown of panel 0M27A that 13 out of 35 alarm windows were activated and NRC concluded that the number of invalid alarms represented an excessive number of nuisance alarms for a safety system (ERCH).

CLASSIFICATION: Operational discrepancy.

RESPONSE

TVA has analyzed nuisance alarms under the control room design review program and problems were found with OM27A alarms. TVA has analyzed the specific conditions NRC has found and is prepared to address the cause and correction.

However, TVA does not agree that this item is a deficiency except as a Human Engineering Discrepancy. The causes of most of the alarms present on OM27A are (1) a number of alarms are valid "old pumping station" (intake pumping station) alarms which were not removed when the ERCH pumping station was put in operation, (2) the remainder are maintenance-related alarm where maintenance needed to be performed to clear the alarm, and (3) low flow alarms from the present station. l Hindows on OM27A which can be cleared by performing maintenance will be )

cleared before U2 restart. All windows which fall into the category of "old pumping station" alarms (intake pumping station) will also be removed. TVA has, under Control Room Design Review commitments, committed to review and eliminate nuisance alarms and to clearly identify alarm priorities. This annunciator study is a specific commitment '.o NRC under the TVA Control Room Design Review summary report (reference R. L. Gridley to B. Youngblood dated November 16, 1986.

TVA does not agree with the NRC in that the condition of the annunciators on OM27A yields a system for which no credit can be taken for operator action initiated by an alarm. NRC has presented no information that is evidence that the SQN operations were confused. Since the alarm response procedures are written on an individual window basis and the operator responds to a new alarm when the horn sounds and window flashes, operator action will still be valid regardless of how many windows are lighted normally.  ;

TVA does not agree with this item being a deficiency. NRC's basis for the deficiency on the ERCH annunciator is incorrect. The section of the FSAR '

3.1.2.3 is specifically referring to Criterion 20 of Appendix A of 10 CFR 50 which deals with the " Protection System" which is defined in IEEE-279 as

. . . all electric and mechanical defices and and circuitry (from sensors to actuation device input terminals) involved in generating those signals associated with the protection function. These signals include those that l

1

-101-1

_ _ _ - _ _ _ - _ l

ITEM NO.: D5.2-3 (Continued)

TITLE: Ineffective ERCH Alarms (Continued)

RESPONSE: (Continued) lk actuate reactor trip and that, in the event of a serious reactor accident, actuate engineered safeguards such as containment isolation, core spray, safety injection, pressure reduction, and air cleaning." It is clear from this definition that the " Protection System" referred to in FSAR 3.1.2.3 and GDC 20 does not include ERCH or the ERCW annunciator system since there exists no reactor protection system sensors or actuation circuitry as part of ERCH.

The annunciator system is designed to be a commercially available annunciator with no specific regulatory requirements or industry standard applicable to its design. Since the basis for the deficiency is in error, this item shculd l not constitute a deficiency. Additionally, this item does not meet the TVA i restart criteria as accepted by the NRC in TVA's Nuclear Performance Plan.

Since no safety feature is involved (the annunciator system is not taken credit for in safety-related actions), the item as found by the NRC does not meet any of the specific criteria for becoming a restart item for SQN unit 2.

-102-

1 1

ITEM NO,: 05.2 TITLE: Inadequate Electrical Isolation of Non-Class lE Traveling S:ree Speed Switch

SUMMARY

OF ITEM:

NRC concluded on the basis of information given under ECN 5637 that no analysis had been performed by TVA that would assure that the additional isolating fuses would clear a fault before the control circuit fuse would blow and that this violates IEEE-279. I CLASSIFICATION: Design Deficiency. j

RESPONSE

TVA agrees that the condition found is a specific isolated deficiency against the TVA design criteria and FSAR commitments on single failure criteria.

However, TVA does not agree with the NRC's basis for the deficiency or the ,

extent. l The design for ECN 5637 was done correctly to consider the selection of fuses which coordinate properly. However, at the time of this change there was no design documentation of the exact fuse manufacturer and catalog number generated. When the fuse program which documents the class lE fuses in most applications was put _in place and the installed fuses were input to the fuse tabulation per SQN Procedure SQEP-34, the installed fuses were documented.

When analyzed, these fuses were found to have a coordination problem which is reflected in a TVA Condition Adverse to Quality Report. The cause of the ,

problem was the lack of control by plant procedure and DNE procedure to '

require DNE analysis of fuse substitutions which were listed in the approved Design Standard. Therefore, plant maintenance was allowed to replace fuses with substitutes listed in the Design Standard without contacting DNE. This  :

substitution caused the coordination problem. ,

An ECN has been developed to resolve the above coordination problem. This work will be completed before U2 restart.

NRC's conclusion that all Class lE circuits using fuses as isolation devices should be reviewed by TVA to assure that sufficient margin is provided in the coordination of the isolation fuse and other Class 1E fuses or protective devices does not take into account work already performed which examined common mode failure potential of isolation type fuses. TVA has analyzed the power systems (ac and dc) for common mode failure of non-Class 1E circuits connected to the power system under 10 CFR 50, Appendix R, analysis and as a part of the power system coordination studies. Additionally, TVA has performed a cascade fuse analysis study which considered series fuse combinations in both the Class 1E 120Vac and 125V de power systems, l

-103-

ITEM N0.: D5.2-4 (Continued) ,

TITIr* Inadequate Electrical Isolation of Non-Class 1E Traveling _;reen Speed Switch (Continued) 1 1

RESPONSE: (Continued)  !

The remaining area of potential Class IE and non-1E fuse-to-fuse coordination

~

is in the control circuits fed from'480/120Vac transformers as part of Class 1E motor control centers. The series fuse combinations which represent 1E/non-lE isolation in motor control center control circuits will be investigated to prove coordination as'a postrestart activity.

~

This investigation can be limited to MCCs because of the other analysis identified above.

To prevent reoccurrence of this type of problem in the future, TVA has ,

instituted a calculation checklist under TVA EEB Procedure Methods as part of the change control process which will detect the need for evaluations / calculations on items such.as fuse coordination. Additionally, since this problem arose out of fuse substitution, a review of present procedures reveals that the present SQEP-34, Revision 0 issued December 5, 1986, plant AI-16 Revision 12 issued for June 9, 1987, and Design Standard DS-E8.1.2, Revision 3 issued January 13, 1987, provide adequate control of fuse substitution of Class IE circuits so that once corrected, this problem should not occur. These procedures and standard require documentation and analysis by DNE of fuse substitutions.

TVA believes that NRC's description of this condition as a deficiency on the basis of the FSAR is incorrect. The section of the FSAR 3.1.2.3 is specifically referring to criterion 20 of 10 CFR 50 which deals with the

" Protection System" which is defined n IEEE-279 as ". . . all electric and mechanical devices and circuitry (from sensors to actuation device input terminals) involved in generating those signals associated with the protection >

function. These signals. include those that actuate reactor trip and that, in the event of a serious reactor accident, actuate engineered safeguards such as containment isolation, core spray, safety injection, pressure reduction, and air cleaning." It is clear from this definition that the " Protection System"  !

referred to in FSAR 3.1.2.3 and GDC 20 does not include the circuitry used to control equipment such as the travel screen motor since there exists no reactor protection system sensors or actuation circuitry as part of this motor control. The NRC description of this condition as a violation of TVA's design criteria SQN-DC-V-12.2 is valid.  ;

'l I

-104-

ITEM NO.: D5.2-5 TITLE: Inadequate Separation of Redundant Main Control Board Wiring

SUMMARY

OF ITEM:

During the July 28, 1987, NRC walkdown of the panel 0M27A, the NRC observed that in several places cables enclosed in braided metal sheaths from redundant divisions were touching, or could migrate with time to touch, in violation of FSAR :ommitments to assure independence of redundant safety-related circuits.

CLASSIFICATION: Construction deficiency.

RESPONSE

TVA agrees that the cases found do represent a violation of the FSAR and TVA l design criteria for electrical separation. TVA takes credit for the metal braid in lieu of 6-inch separation only in main control panels and auxiliary control room panels which contain redundant safety-related cabling.

TVA had written CAQR SQP871325IDI, before the NRC noting the problem during a walkdown, to document a case where the cables of redundant divisions were tied together. TVA has since done a QC inspection of the wiring internals to all panels in the Main Control Room to identify where other cases may exist of braided redundant cables touching or being able to migrate to touch. TVA will rectify these discrepancies by installing spacers to separate and hold the position of cables to achieve some air space separation. TVA will resolve all cases found before U2 restart for U2 and common panels which contain safety-related control cables.

l

-105-

_-- ._- /

b i

ITEM NO.: 'US.2-6 a TITLE: Adequacy of Electrical Separation of Isolation Device Inputs anc )

Outputs'

SUMMARY

OF ITEM:

TVA drawings 6858D75 and 6858066 show that non-1E wiring connecting to the  ;

non-lE side of the electrical isolator has been bundled together with Class 1E signal wiring and wiring for the 125 volt DC switchgear control power. The NRC_ believes since the Class lE and non-Class 1E wiring is bundled together, ,

the Class IE control power may be degraded due to the propagation of a  !

credible fault in the non-Class 1E circuit to the Class lE control power l

, circuit. '

CLASSIFICATION: No Deficiency.

P.ESPONSE: i NRC's potential basis for the. design condition noted is not correct. The I section of the ESAR 3.1.2.3 is specifically referring to Criterion 20 of 10 CFR 50 which deals with the " Protection System" which is defined in  ;

IEEE-279 as ". . .all electric and mechanical devices and circuitry (from sensors to actuation device input terminals) involved in generating those signals associated with the protection function. These signals include those that actuate reactor trip and that, in the event of a serious reactor accident, actuate engineered safeguards such as containment isolation, core spray, safety injection, pressure reduction, and air cleaning." It is clear  !

!' _from this definition that the protection system referred to in FSAR 3.1.2.3 and GDC20 does not include the control circuitry in question in this item since no reactor protection system sensors or actuation circuitry which is  ;

part of RPS is involved in the circuits in question. Also, NRC basis that the  !

design practice noted is a violation of the TVA Design Criteria SQN-DC-V-12-1 is Invalid. The criteria quoted by the NRC SQN-DC-V-12-1, paragraph 4.4.7,  !

applies to isolation devices, not to the wiring. However, all safety-related !

circuits should meet Design Criteria SQN-DC-12-1.

TVA does not agree with the findings for the following reasons:

Routing of Class lE and non-Class lE circuits for short length in special i cases like this application is based on our Design Criteria SQN-DC-V-12.2, i section 4.4, and FSAR section 8.3.1.4.3 which states:

" Cables for nonessential functions are not run in conduit used j for essential circuits except at terminal equipment where only one conduit entrance is available." j

" Cables for nonessential circuits may be run on cable trays with 3 those for essential circuits with the following restrictions. When I a nonessential cable is routed .in a tray with essential (GSPS) cables, tnat cable or any cable in the same circuit has not been subsequently routed unto another tray contain'ng a different division of separation of essential cables."

J l

-106-l

___ - a

E . .

L ITEM NO.: :US.2-6'(Continued) ,

TITLE: Adequacy of Electrical Separation of Isolation Device Inputs and Outputs (Continued)

RESPONSE: (Continued)

FSAR section 7.1.2.2.1 (as stated below) permits similar use.

" Circuitsfor non-redundant functions should be run in cable trays or conduit separated from those used for redundant circuits. Where this cannot be accomplished, non-redundant circuits may be run in a cable tray, conduit, etc., assigned to a redundant function. When so routed, it must remain with that'particular redundant circuit routing and will not cross over.to other redundant groups."

This design is consistent with the FSAR. 1 In 1978, TVA provided a response to the NRC for a similar application for RTS Actuation Logic (NRC 07.52) and Isolation Boards in the Solid State Protection  :

System (NRC Q7.53). (See attachment.) These responses address that no credible failure of non-Class lE circuit shall prevent the Class IE circuit from performing its design basis function.

TVA considers this item closed.

)

l I

f l

-107-

- - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __o

1 l

ITEM NO.: U5.2-7 TITLE: Use of. Common Penetrations for Redundant Instrument Lines

SUMMARY

'0F ITEM:

NRC reviewed TVA Design Criteria SQN-DC-V-10.5 which allow redundant i instrument lines to be routed through a common penetration when separate penetrations cannot be used. NRC's concern is that the lack of spatial separation could cause a loss of redundant instrument lines as a result of an accident, internally generated missile or other hazard.

CLASSIFICATION: No Deficiency.

RESPONSE

The basis for NRC concern is a potential violation of IEEE-279-1971, section  !

4.6, which in itself does not meet TVA restart criteria agreed to by NRC in the TVA Nuclear Performance Plan.

TVA design criteria SQN-DC-V-10.5 clearly delineate the use of barriers where redundant sense lines use the same wall, floor, or containment penetration.

Since the primary concern with sense line separation is pipe whip or jet impingement, the exact evaluation of hazards must be performed by the pipe break analysis team. The team evaluated as-installed configuration of instrument lines in areas where the potential for unacceptable interactions with postulated pipe breaks are present. This field evaluation would identify the need for barriers or for additional separation if a common hazard was found. The TVA criteria and documentation requirements for these analyses are covered in TVA Design Criteria SQN-DC-V !.l.11 and SQN-DC-V-2.13.

Presently and in the future, since the sleeves for wall, floor, or containment penetrations are shown on the 47W600-series drawings, designers can identify if a proposed route might cause the use of a common penetration for redundant sense lines. Also, design output drawing 47W600-133 requires that lines be verified to meet the separation requirement of this drawing and the TVA design criteria by TVA's DNQA. In addition, the pipe break analysis team is included in the .ECN review process and will ensure that any change to instrument lines which creates a new line route will be evaluated for impact of postulated pipe breaks.

i l

l I

l l

-108-l

! l

' ITEM NO.: U5.2-7' TVA has sufficient design' control in place to assure that common p 1etrations are used only when necessary and that appropriate barriers or.other protection I be installed when pipe break hazards are present.

TVA believes that NRC's statement that the allowance of common sleeves for

< redundant instrument line may lead to the allowance of common structures.for redundant instruments'is unfounded. TVA's design output for locating I instruments'are the 47H600 series of drawings which control location of all instruments (not on skids). These drawings clearly show that no trained or protection set instruments are located on a common instrument panel with a i redundant' instrument.

This' issue is considered closed by TVA.

1 I

]

i 1

]

i I

i

. I

-109- l 4

1 TEM NO.: D5'.2 TITLE: Adequacy of ERCH. Instrumentation Provided for Detection of BreaA in Non-Seismic ERCH Piping

SUMMARY

OF ITEM:

l In the event'of a pipe break in the non-seismic portion of the ERCH piping, operator action is.taken based on a high flow alarm and status light in the control room'that monitors each ERCW' header.

CLASSIFICATION: No Deficiency.

' RESPONSE:

l Calculations were provided to the NRC showing that under a " critical. crack" ]

there.was sufficient time available to use manual (operator initiated)

. isolation of the headers. TVA: contends that a " critical crack" break is the largest break which must be assumed in design of this piping.

Fur further response, refer to Mechanical item D2.2-5. J The NRC's concern about the seismic qualification of the information display to.the operator for line break detection is unfounded. It is true that'the annunciator is not seismically qualified, but the excess flow indication is-provided to the operator via indicating lights on 0-M27A (XI-67-206, i XI-67-209). These indicating lights are seismically qualified and are used in lieu of the annunciator under a seismic event if the annunciator fails to ,

. operate.. The Abnormal Operating Instruction (AOI) will be modified to  !

-specifically call out these lights for operator action.

NRC assumes that the operator will be responding to the annunciator system and-only the annunciator system to isolate the break in a timely manner. However, if specific procedure steps call for isolation of 0-FCV-67-205- and 0-FCV-67-208 upon a seismic event when the 0-XI-67-206 and 0-XI-67-209 are lit, the' operator. response is not to the alaro but is to the specific .

sequenced steps in the AOI. Therefore, operator confusion by other alarms '

should not be a factor in the operator's decision process to isolate the ERCH lines.

)

1

-110-l 1

l

-!E 2 TEM.NO.: D5.3-1

. TITLE: InadequateTShutdown Capability Outside Control Room: Traveling Screen /Screenwash Circuits

SUMMARY

0F ITEM:

Sho'rt circ'ults-in the' traveling' screen backwash indicating lights circuits'due to a fire'in control: room panel 27A could prevent operation of the screen-1 backwash pumps during remote' shutdown due to. blown fuses causing.a loss of- .i control' power. ,

i CLASSIFICATION:

RESPONSE

The response will be provided during the week of November.2, 1987.  !

t u

l 1

i i

l I

-111-  ;

ITEM NO.: D5.4-1 ,

TITLE: Adequacy of Freeze Protection for Instrument Lines in the ERCH

SUMMARY

OF ITEH:

NRC's concern is based on the loss of engineered safety features equipment under environmental extremes under the loss of environmental control facilities. NRC's specific concern is that freezing of instrument lines used

.for the strainer differential pressure measurement could defeat automatic operation of all four strainers and defect alarms which alert the operator to a clogged strainer. They state the basis to be FSAR 3.11.4. NRC states that redundant and qualified environment controls are a requirement.

CLASSIFICATION: No Deficiency.

RESPONSE: See response to 02.2-7.

I 1

1 I

-112-I

l l I:

l c ITEM N0.: U5.4-3 ,

TITLE: Adequacy of Seismic Qualification for Field Located Relays, Timers and.

. Terminal Blocks

SUMMARY

OF-ITEM:

Connection diagrams (reference 1-5) show installation of'various components (relays, timers, terminal blocks, etc.) incorporated by several ECN modifications. The installation instructions for these items include the notations: " field to locate" or " field to. order and. locate."

The installation _ instructions appear to direct the field personnel to use their own judgment when. installing additional equipment in Class IE panels 1

! -without proper analysis of the effects of such installations upon seismic l qualification of the components or the panels. i k

Since many ECNs have been completed this way and if each ECN added some ." field I to locate" equipment, it is possible that the panel qualification might have been voided.

)

1 There is presently no evidence that TVA has considered the significance of the l changes or performed an analysis to verify continued qualification.  ;

1 CLASSIFICATION: No Deficiency I

RESPONSE

The installation of the various components (relays, timers, terminal blocks, {

etc.) is supported by engineering change notices (ECN), a procedurally )

controlled (reference 6) method of engineering review. The ECN process j ensures that the affect of such installations upon seismic qualification o. j the components and panels is properly analyzed. The ECN cover sheet, which 'l indicates when seismic analysis is required, was prepared per reference 6 for i the subject modification. I As applied to the components and panel, the engineering review (ECN) established that the modification is not location sensitive and does not degrade the seismic qualification of the component or panels because the i component weight is insignificant compared to the panel weight. Thus the modification was approved and recorded on the appropriate design drawings. i When current design changes are proposed, the engineering procedures require an evaluation of all design changes on the design drawings. Thus, the individual change and the accumulative affect of all changes, including field selected installation locations is evaluated. Additionally, the current ECN  ;

procedure provides for a field walkdown, or other appropriate' action, to verify plant configuration for the area of the proposed change.

There is no impact on safety and no hardware changes are required.

I

-113-

i

ITEM-NO.: US.4-3 (Continued)-

TITLE: Adequacy of, Seismic Qualification for Field Located Relays, Timers and.

Terminal Blocks '(Continued)

REFERENCES:

1. TVA Drawing No. 33-470350-664 Connection Diagram Panel-8, ERCH Pump'F-B-Rev. 2 (Contract 54499) 1

2. TVA Drawing No. 6858066,. tow. Voltage Metal Enclosed Switchgear/480V Shutdown Board 2B1-B,.Rev. 8 (Contract 54523):

3. TVA Drawing No. 6858D75, Low Voltage Metal Enclosed Switchgear/480V Shutdown Board 2B2-B, Rev. 8-(Contract 54523)

.. 4. TVA Drawing No. 33-47035-D474 - Connection Diagram Panel 8, ERCH Pump Rev. 1 (Contract 54499)

.5. TVA Drawing No. 6858046, Low Voltage Metal Enclosed Switchgear/ Shutdown Board 2Al-A, Rev. 9 (Contract 54523)

6. EN DES-EP 4.02 R9 (ESS 790828 202) i

-114-

ITEM NO.: D5.6-1 ,

TITLE: Inadequate Specification of Background Radiation for ERCH Effluent Liquid Radiation Monitors

SUMMARY

OF ITEM:

NRC concluded from a review of the ERCH effluent radiation monitor's specification that under accident conditions the calculated background of 1.0 R/hr would yield these monitors unable to perform effluent monit( ing under accident conditions. Also, NRL concluded that these monitors were not specified on designed properly for the postaccident monitoring function.

NRC's basis for this conclusion is FSAP Section 11.4.2.1.2.

CLASSIFICATION: Documentation.

RESPONSE

FSAR Section 11.4.2;1.2 says.that " Redundant monitors continuously monitor the ERCH effluent to ensure radiation is not released to the environment. This monitor serves as an accident monitor to detect leakage from either the component cooling heat exchangers or containment spray heat exchangers (during accident)." This statement is incorrect as presently written and will be revised in the next update of the FSAR.

The cause of this discrepancy is a lack of coordination in the review of various FSAR sections which deal with the same subject. FSAR Chapter 11, Section 11.4.2.1.2, states that these radiation monitors are used postaccident to monitor leakage which is inconsistent with the FSAR Section 7.5 or the Chapter 15 accident analysis. These monitors are designed to act as normal system leakage monitors and have not been designed to function under the background environment postaccident. The monitors were specified properly for their intended function as a normal effluent monitor.

TVA will change the FSAR in Chapter 11 to correctly reflect these monitors' usage in the next annual update. TVA will complete an FSAR verification review which will correct all cases of similar non-coordination of various sections of the FSAR covering the same subject.

TVA has evaluated other postaccident radiation monitors and has found no otner monitors to have a similar problem as the one found by the NRC with regard to background radiation.

-115-i i

I j

I ITEM NO.: D6.2-2 l TITLE: Insufficient Demonstration of Adequate Class IE Motor Starting and Running Voltages

SUMMARY

OF ITEM:

The SQN calculation OE2-EEBCAL001, Rev. 8, "AC APS Voltage and Loading l Analysis," determined the starting and running voltage levels at both 6.9kV j and 480V motor terminals for conditions in which offsite power is available.

The study did not consider the condition when AC power supply is from the onsite diesel generators.  ;

{

Section 1.2 of both the medium voltage and the low voltage sections of the i study indicates that the design basis condition considers unit 2 with'a LOCA and. Phase B containment isolation and unit I with a full-load rejection.

However, Section 5.A of the low voltage portion of the study states that unit 1 is considered in cold shutdown.

CLASSIFICATION: Documentation

RESPONSE

This condition had previously been identified as a result of the SQN design baseline verification program and by surveillance testing. TVA issued calculation SQN-E3-Oli (B43 871002 903) entitled " Diesel Generator Voltage Analysis" on October 2, 1987, for diesel generator lA-A, which verified tne voltage to be adequate for equipment to perform its intended safety function.

Voltages for the remaining diesels will be verified by testing to be no worse than those calculated before unit 2 restart.

Calculation OE2-EEBCAL001, Revision 8 (B43 870608 905) was specifically prepared to support unit 2 restart. Section 1.0 of the calculation states the design basis for which the calculation is to be performed and agrees with the design basis as stated in the FSAR Section 8.3.1.1. The low voltage portion of the calculation was performed with unit I assumed in cold shutdown as stated in Section 5.A.3 of the calculation because not all the unit 1

! as-configured field data was available at the time the calculation was prepared. The calculation will be revised before unit I restart to reflect the established design basis.

REFERENCES:

1. DNE Calculation OE2-EEBCAL001 (B43 870608 905).

2. DNE Calculation SQN-E3-Oli (843.871002 903)

3. FSAR Section 8.3.1.1.

l l

1

-116-

p JTEM NO.: D6.2-4 j TITLE: ' Absence of Neutral Grounding and Ground Fault Detection or 480V

. Auxiliary Power Systems j

SUMMARY

OF ITEM: .I

'l TVA states in FSAR Section 8.3.1.1 (page 8.3-25) that the 480V auxiliary power distribution substations are equipped with ground fault detectors. The existing detection system may not be-adequate to detect an arcing ground fault, t

CLASSIFICATION: Desi.gn Deficiency 1

RESPONSE

TVA agrees in part with the deficiency and will take appropriate corrective actions. CAQR SQF870181IDI has been initiated to document this deficiency.

1 TVA uses a delta-to-delta ungrounded system for.480V, Class lE power. This -j sytem was specified by TVA and provided by the switchgear and MCC vendor and is a design commonly.used in industrial and power plant applications.

.This system was designed and specified in the early 1970s and was typical of power distribution system design during this timeframe. As stated in FSAR l Section 8.3.1.1 (page 8.3-25), the SQN auxiliary power system is adequately  !

designed to protect individual circuits and the system for phase-to-phase j ground faults. Ground fault indication' circuitry with local test switch is j provided on each:480V shutdown board, as shown on schematic drawing 45N779-1, Revision J, and each ERCW motor control center, as shown on connection diagrams, 35H736 series. This indication, which meets the intent of the FSAR j commitment, will show the existence of a solid ground fault on the system and l identify the faulted phase. The highest possible voltage that can be experienced during a solid ground fault for a delta connected system is 173 )

percent of the phase or line voltage which for this case would be 277V X 1.73 or 480V, rather than the 830V mentioned in the NRC report. An intermittent  ;

arcing. ground fault which is much less likely to occur could cause a high transient overvoltage above the phase or line voltage. The existing installed circuit is presently tested once each shift in accordance to Operations Section Letter A99. This circuit may not adequately detect an arcing ground 4 fault which is intermittent. l A system with a high-resistive ground would counteract the system capacitive charge present during an arcing fault and would prevent high overvoltages.

ICEA Standard Publication No. S-66-524, S-68-516, and S-61-402 recommends that cables have insulation levels of 173 percent for circuits that may be required to operate longer than one hour continuously with one phase grounded. TVA has, in all cases, specified cable with 133 percent insulation level which is

-117-

_ _ _ _ _ _ _ _ _ _ _ _ _ l

ITEM NO.: D6.2-4 (Continued)

TITLE: Absence of Neutral Grounding and Ground Fault Detection on 480V l Auxiliary Power Systems (Continued)

RESPONSE: (Continued) recommended where the faulted cable will be deenergized in a time not exceeding one hour. Calculations using actual cable size information contained in Electrical Design Standa*d DS-E12.1.13 show that average cable insulation thicknesses for 600V cable vary from 36 mils for a 3c #12 to 159 mils for a single conductor 750 MCM. Table 3-1 of ICEA S-66-524 recommends insulation thicknesses for 600V cable of 30 mils for a 3c #12 and 110 mils for a single conductor 750 MCM. The values for a 2,000V cable are 45 mils for a 3c #12 and 120 mils for 750 MCM. In all cases except for 3c #12 and 3c #10, the TVA average insulation thicknesses meet or exceed the ICEA recommended values for 2,000V .able. In addition, TVA has also contacted two major cable suppliers, Rockbes.os and Okonite, who have confirmed that 600V cable with 133 percent insulation level is suitable for continuous operation with single ground fault provided the cable ampacity is not exceeded. Therefore, TVA concludes the cable insulation levels are adequate for continuous operation with one phase solidly grounded. j 1

TVA considers an undetectable arcing ground fault of such magnitude to cause high system overvoltages in one train of ac power simultaneous with a LOCA and a single-active failure of the loss of the opposite train of ac power to be an l incredible event. TVA has performed a probablistic analysis that shows the 1 probability of occurring to be 6.7 x 10-* per year. Based on this low l

-probability, TVA believes the testing presently performed by OSLA99 is j adequate. Since the condition does not present a condition that will  :

jeopardize safe plant operations, TVA will evaluate the need to unrade the existing circuitry to provide main control room alarms after plant startup. ,

1 l

REFERENCES:

1. TVA.15E500 Drawing Series

2. TVA 35W716 Drawing Series

3. TVA 45N749 Drawing Series

4. ICEA Standard S-66-524, S-68-516, and S-61-402

5. TVA Condition Adverse to Quality Report CAQR SQF870181IDI (B05 871005 302)

6. TVA 45N779 Drawing Series

7. TVA 35H736 Drawing Series ,

8. FSAR Section 8.3.1.1

9. TVA Electrical Design Standard DS-E12.1.13 l

l

]

)

-118-I

i

'! TEM NO.: 06.3-1 i

TITLE: Inadequate Voltage Drop 1 Calculation for 125V DC and 120V AC Control Circuits

SUMMARY

OF ITEM:  !

TVA has made several studies and calculation to demonstrate that at least the minimum level of operating voltages, as identified by suppliers, will be applied'to 125V de and 120V ac Class lE devices and equipment _in the event of ,

a design basis incident. The NRC team identified the following items which they considered shortcomings in the voltage drop calculations: (1) all.the calculations have unverified assumptions, one of which, the minimum operating ,

voltage of one type of relay in SQN-VD-VDC-1, may be critical to the validity of the results; (2) corrective actions identified by Significant Condition Reports SRNSQNEEB8605 and SCRSONEEB8532 regarding calculations SQN-VD-VDC-1 and SQN-VO-VAC-2, are yet to be completed; (3) calculation SQN-VD-VDC-1 considers a dc board voltage level of 120V dc when analyzing the 6.9kV and 480V switchgear control devices. The calculations should have used the battery discharge voltage of 105V dc.

CLASSIFICATION: Documentation

RESPONSE

Calculations SQN-VD-VDC-1 and SQN-CPS-001 are technically adequate, but the purpose and approach used in performing the calculations was not clearly defined. Items (1) and (3) of the shortcomings identified above are duplicate items referring to the 6.9kV and 480V switchgear breaker control relays.

Calculation SQN-VD-VAC-2 was revised by revision 1 to substantiate the adequacy of corrective actions for SCR SQNEEB8532 which were implemented by ECN L6698.

Sections 8.3.2.1.1 and 8.3.2.2.2 of the FSAR states that vital battery voltage is 105V de after a two-hour discharge. This rating is based on a total loss of all ac power with the battery supplying power to connected loads.  !

Immediately upon loss of ac power, the battery voltage is 120V dc which was {

the voltage value used for the switchgear breaker controls and for certain  !

solenoid valves required to operate immediately.

The summary contained in Section 5.0 of SQN-CPS-001 and Section 5.5 of )

SQN-VD-VDC-1 confirms the technical adequacy of the calculations except for -

l those items identified in Section 5.5(3) of calculation SQN-VD-VDC-1. These I items were documented by SCR SQNEEB8605 and were further evaluated in l calculation SQN-VD-VDC-2 (B43 870612 901) entitled "125V.DC Vital Instrument l Power System j l

l l

1

-119-

1 l

ITEM NO.: D6.3-1 (Continued) ,

TITLE: Inadequate Voltage Drop Calculation for 125V DC and 120V AC Control Circuits (Continued) l RESPONSE: (Continued) j Design Verification - Further Analysis." This calculation in conjunction with the 120V ac vital inverter testing performed at Hatts Bar Nuclear Plant (documented in the corrective action portion of SCR SQNEEB8605 R1, ,

843 860220 910) confirmed technical adequacy. I 1

To resolve this issue, TVA has revised SQN-CPS-001 to adequately document the design basis for which the calculation was prepared and to document the reasons for the voltages used. TVA has revised SQN-VD-VDC-1 to state that the switchgear breakers will have adequate voltages to operate after the designed battery two-hour discharge period (105V dc). The switchgear manufacturer's data shows that the switchgear breakers can operate as low as 90V dc. Since the calculations are technically adequate, no hardware changes are required by these revisions.

In each of the calculations TVA has eliminated all unverified assumptions except one for which a CAQR has been initiated in accordance to NEP 9.1.

Based on these actions, no further corrective action is required.

REFERENCES:

1. DNE Calculation SQN-CPS-001 (843 870601 902)

2. DNE Calculation SQN-VD-VDC-1 (B43 861210 903)

3. QN FSAR Section 8.3.2.1.1 and 8.3.2.2.2

4. DNE Calculation SQN-VD-VDC-2 (843 87.0612 901)

5. Significant Condition Report SCR SQNEEB8605

6. Significant Condition Report SCR SQNEEB8532

7. DNE Calculation SQN-VD-VAC-2 (B43 870518 969)

-120-

l:

-! TEM N0.: D6.6-1.

U TITLE: Unsubstantiated Motor-Operated Valve Performance at Deg: .;d Voltage l

SUMMARY

OF ITEM:

L

' Calculation OE2-EEBCAL001, Revision 8, which analyzes the capabi;ities of the auxiliary power system, assumes motor operated valves will open with 80 percent rated voltage applied to motor terminals. Also, the study assumes motor operated valves without brakes will close at 75 percent of rated terminal voltage, and valves with brakes at 80 percent. The calculations show that the~ motor operated valves' required to operate for a LOCA with Phase B containment isolation will werate, based on these assumptions. The team reviewed procurement docurwtation indicating that while some ERCH motor operated valves were specified to operate with a minimum terminal voltage of

. 75 percent, the purchase specification of other safety-related ERCH MOVs had no degraded voltage requirement.

CLASSIFICATION: Design Deficiency

RESPONSE

TVA's AC Auxiliary Power System Voltage and Loading Analysis (TVA Electrical Engineering Branch. calculation No. OE2-EEBCAL001, B43 870608 905) demonstrates that the ERCH system MOVs have at least 80 percent of rated voltage on starting. TVA has evaluated all ERCH system electrically operated valves-that do not have documented minimum starting voltages. There are 57 ERCH system valves of which 26 of these have documented starting voltages of 80 percent of '

460V. This documentation is contained in TVA's Environmental Qualification Binders for MOVs. Of the remaining 31 MOVs, 13 have their power disconnected; 8 have been determined to be able to' start at 80 percent due to the similarity 3 of operators used for the MOVs for which documentation is available; 4 have '

calculated worst case voltages of at least 90 percent of rating; and 6 are not active MOVs required to mitigate a design basis event. l This evaluation is documented by Quality Information Release QIR EEB87460 (B43 870923 901) and verified the starting capability of the MOVs to be .]

adequate. Since MOVs used in other systems are the same type and/or of '

similar construction, TVA has a very high confidence level that a similar evaluation for other systems will show the same results as that for the ERCH system. To substantiate this, TVA will perform a similar evaluation of the l starting capability of all active valves in other safety systems after restart.

]

REFERENCES:

1. DNE Calculation OE2-EEBCAL001 (843 870608 905)

2. Quality Information Release QIR EEB87460 (B43 870923 901) l

3. Environmental Qualification Binders SQNEQ-MOV-001, j SQNEQ-MOV-002, SQNEQ-MOV-003, and SQNEQ-MOV-004. )

-121-l

_ _ _ _