ML20127J588

Responds to Deficiencies & Common Concerns Raised in Chapter 7 of Insp Rept 50-400/84-48 Re Integrated Design Review. Independent Evaluation Team Concluded That Identified Items Resolved & Overall Design Process Adequately Controlled

ML20127J588
Person / Time
Site:	Harris
Issue date:	06/13/1985
From:	Watson R CAROLINA POWER & LIGHT CO.
To:	Taylor J NRC OFFICE OF INSPECTION & ENFORCEMENT (IE)
References
RTR-NUREG-0737, RTR-NUREG-737 HO-853100-(E), NUDOCS 8506260704
Download: ML20127J588 (226)

Text

r- -

CD&L Carolina Power & Light Company Shearon Harris Nuclear Project P. O. Box 165 New Hill, NC 27562 June 13, 1985 File: HXDE-XXX-XXX-XXX Letter No.: HO-853100 (E)

Mr. James M. Taylor, Director

, Office of Inspection and Enforcement U. S. Nuclear Regulatory Commission Waanington, DC 20555 INTEGRATED DESIGN INSPECTION 50-400/84-48

Dear Mr. Taylor:

Your letter of April 15, 1985 transmitted NRC Inspection Report No. 50-400/84-48 concerning the Integrated Design Inspection of the Shearon Harris Nuclear Power Plant. You requested that we respond to the deficiencies and unresolved items and the common concerns addressed in Chapter 7 of the report. Attached is Carolina Power & Light's response. For your convenience, our response is submitted in a format similar to your report and addresses in detail the ite,ms which appear in your report.

I and other CP&L senior managers have closely monitored the Integrated Design Inspection and associated activities. We were pleased with the NRC staff's conclusion that the overall Shearon Harris plant design process has been adequately controlled pending corrective action of the items identified in the report and the root causes. Necessary correc-

. tive actions and our response to the report have been given a very high priority. The six weaknesses identified in Chapter 1 of the report that raise the possibility that similar weaknesses might be found in systems other than those inspected by the IDI team have been investigated in relation to other systems. The results of our investigation, as well as actions taken, are provided in our response. Our investigation and actions demonstrate that items identified in your Chapter 1 are not generic in nature and do not extend to other systems. In additi,on, the concerns identified in Chapter 7 that may be common to more than one discipline have been investigated and addressed.

You commended our intent to shift certain design work to our in-house engineering organizations. Our objective is to create a smooth transi-tion from construction to operations. Furthermore, Ebasco is working with us' to make this transition as effective and smooth as possible.

1 6

8506260704 850613 PDR ADOCK 05000400 l\

G PDR

r--

Mr. James M. Taylor HO-853100 (E)

We agree that there is a need to assure that appropriate skills and technical guidelines exist within our engineering organization and we have an orderly, controlled program to accomplish this task. We presently provide a mix of in-house design and A/E engineering managed by CP&L in support of our three operating units, and will be applying the same controlled procedures in support of Harris. As a result of the IDI team's concern, we have initiated actions to strengthen the overall capability of'our organization with emphasis in the Electrical discipline.

As noted in our response, Carolina Power & Light Company established an Independent Evaluation Team with technical capability to review and assess the adequacy of the responses and corrective actions proposed by Harris Project personnel. This Team included engineers from our Corporate Nuclear Engineering and Licensing and Quality Assurance Departments, from Ebasco, and from an outside consultant. These engineers were independent of the Harris Project. This team considered those issues raised by the IDI team. As documented in the attached memo from the team leader, the Independent Evaluation Team concluded that our responses satisfactorily resolve the items identified by the IDI; that the overall Shearon Harris plant design process has been adequately controlled; and that none of the identified deficiencies, either collectively or indi-vidually, are such that the overall adequacy of the plant design is called into question.

We consider the IDI to have been a positive experience for the Harris Project and look forward to your follow-up inspection, at which time we are confident you will be able to confirm your conclusion that the overall Shearon Harris plant design process has been adequately controlled.

Very truly yours, fhut/

R. A. Watson, Vice President Harris Nuclear Project Department RAWh4FT/ jam Enclosures IBHM-3326-OS4 cc: Mr. B. C. Buckley (NRC) Mr. Wells Eddleman Mr. G. F. Maxwell (KRC-SHNPP) Mr. John D. Runkle Dr. J. Nelson Grace (NRC-RII) Dr. Richard D. Wilson Mr. Joe Joyce (NRC-ICSB) Mr. G. O. Bright (ASLB)

Mr. Travis Payne (KUDZU) Dr. J. H. Carpenter (ASLB)

Mr. Daniel F. Read (CHANCE /ELP) Mr. J. L. Kelley (ASLB)

Wake County Public Library

..g._ . . _ , , _ ._ ._ _

Mr. James M. '. Taylor . HO-853100 (E) t 1 bc: Mr. H. R.-Banks Mr. D. C. McCarthy Mr. R.- G. Black, Jr. Mr. S. McManus Mr. C. S. Bohanan - Mr. C. H. Moseley, Jr.

' Mr. H. W. Bowles Mr. D. L. Nordstrom (LIS)

Mr. . C. Carmichael (2) Mr. R. M. Parsons Mr. N.' J.. Chiangi Mr. C. A. Rosenberger

. Mr. A. B. Cutter _ Mr. M. Shannon (Westinghouse)

Dr. T. S. Elleman Mr. Sheldon D. Smith Mr. , D.- E. Hollar/Ms. S. F. Flynn , Mr. A. C. ..

Tollison

- Mr. C. ~' L. Forehand Mr. E. J. Wagner Mr. B. J. Furr Mr. C. C. Wagoner Mr. J. F. Caribaldi (Ebasco) Mr. R. A. Watson.  !

Mr. J. L. Harness Mr. B. H. Webster a Mr. W. J. Hindman Mr. B. M. William's Mr. P. C. Hopkins Mr. J. L. Willis Dr. J. D. E. Jeffries Mr. T. A. Baxter (Shaw, Pittman, Mr. L.' I. Loflin Potts & Trowbridge) -

Mr. R. E. Lumsden Mr. M. F. Thompson, Jr.

Mr. L. H. Martin Central File Mr. R. L. Mayton, Jr.

4 f

e 1

4 i

s c yy . e. e ,--r, r +- *r"-~***~-+-- ' - ' ~ " ' - ~ ' ' + -w

• wa n - ----- ve r- er w e r wW rw r**m<ee-f'c-e*=-****'- "=r~ + - -* - - - * -'-*=---*-m-r--~-*'=-----*'

r-i Form 244 Carolina Power & Light Company Company Correspondence File: H85-502-AU-A617 NELD-G-474

. fJUN 07ENE MEMORANDUM 'IO: Mr. R. A. Watson FROM: M. G. Zaalouk

SUBJECT:

Harris Nuclear Project Independent Evaluation of the Harris Plant Responses to NRC IDI The CP&L Independent Evaluation Team has completed its review of the Harris Nuclear Project responses to the NRC IDI Report. The Team has concluded that the overall response to the NRC IDI Report (No. 50-400/84-48),

dated April 15, 1985 is acceptable and will satisfactorily resolve the items identified by the IDI; that the overall Shearon Harris Plant Design Process has been adequately controlled; and that none of the identified deficiencies either collectively or individually are such that the overall adequacy of the plant design is called into question.

The Evaluation Team has considered all issues raised in six engineering discipline areas (Mechanical, Stress Analysis, Supports, Civil, Dose, and Electrical /I6C) and five programmatic areas (Computer Sof tware Quality Assurance, Design Capability of the Harris Plant Engineering Section, Design Verification Process, Minor Design Change Designation, and Design Interface Problems). The above conclusion is based on the responses received by the Team from the Harris Nuclear Project and on the assumption that stated corrective actions, where indicated in the responses, vill be performed in a timely manner, technically correct, and properly documented.

The Evaluation Team included engineers from the Nuclear Engineering &

Licensing Department, Corporate Quality Assurance Department, Corporate Nuclear Safety Department, Ebasco Engineering, and Duke Power Company.

These discipline engineers and experts were independent of the Harris Project and were selected in accordance with the Team Charter (Operating Plan), dated March 29, 1985.

r:-

e

).

• Mr. R. A. Watson - JUN 07W85 NELD-G-474 I would like to take this opportunity to thank your Staff for their cooperation and support to the Team Members during the completion of this effort.

Please let me know if you have any questions.

,/pg . d

• Dr. M. G. ouk - Chairman CP&L Indepen Evaluation Team MGZ/gvc (0245) cc: Mr. A. B. Cutter Mr. E. M. Harris Mr. L. I. Loflin Mr. M. F. Thompson, Jr.

Mr. E. J. Wagner Team Members Mr. John W. Audas Mr. R. T. Biggerstaff Mr. John Christoddins (Ebasco, N.Y.)

Mr. Angelo Demagistris (Ebasco, N.Y.)

Mr. William Fan (Ebasco, N.Y.)

Mr. Steve Floyd

• Mr. George Hackett (Ebasco, N.Y.)

Mr. Ibrahim Hassam (Ebasco, N.Y.)

Mr. Jeffrey D. Hologa Mr. Tom James Mr. David M. Koss Mr. Ravi Raghavan (Ebasco, N.Y.)

Mr. Kent Russell Mr. Gerald A. Speight, Jr.

Mr. Kriton Valleras (Ebasco, N.Y.)

Mr. R. F. Wardell (Duke)

Mr. R. R. Whitt (Duke)

J: ,

~

.GS ' . .

SHEARONHARRISNUCLEARPROJECT INTEGRATEDDESIGNINSKCTIONRESP0 TEE SIEMITTED IN RESPONSE T0:

REPORTND: 50 f400/8tN18 DOCKETND: 50f400 mL

TABLE OF CONTENTS 1

1 INTRODUCTION AND SUPMARY l 1.1 Objectives and Summary j

.1. 2 Carolina Power & Light Response 1.3 Response Format ,

1.4 Major Conclusions I

! 2 MECHANICAL SYSTEMS 2.1 Design Information i 2.2 Chemical and Volume Control System Design

! 2.3 Containment Spray System Design 2.4 Fire Protection Design 2.5 - Radiation Protection Design Analyses 2.6 Engineered Safety Features Ventilation System 2.7 Seismic Design of Components Not Required for Safe Shutdown 2.8 Field Installation of Equipment in Close Proximity to Piping Systems 2.9 Emergency Service Water System ,

3 MECHANICAL COMPONENTS 1 3.1 Design Information 3.2 Pipe Stress Input and Output 3.3 Piping Stress Procedures

3.4 Pipe Supports.

3.5 Mechanical Equipment 4 CIVIL AND STRUCTURAL 4.1 Design Information 4.2 Seismic Analysis 4.3 Containment Building Foundation Mat

4.4 Containment Building and Internal Structures 4.5 Reactor Auxiliary Building 4.6 Diesel Generator Building 4.7. Tank Building

4.8 Main Dam Spillway i

4.9 Analysis and Design of Mechanical Component Supports 4.10 Design Of Electrical Cable Tray / Conduit Supports

4.11 Design of HVAC Duct Supports 5 ELECTRIC POWER SYSTEMS l 5.1 Design Information

5.2 Personnel and Guidance

! 5.3 Standby Emergency Power Supply f

5.4 Motor Control Centers l 5.5 Direct Current System

5.6 Containment Electrical Penetrations l 5.7 Motor Electrical Penetrations l 5.8 Cable Design and Analysis l

TABLE OF CONTENTS (Cont'd) 5 ELECTRIC POWER SYSTEMS (Cont'd)

- 5. 9 Elementary / Interconnect Diagram Review 5.10 Design Activities Including Design Change Control 5.11 Electrical Separation 6 INSTRUMENTATION AND CONTROL 6.1 Design Information and Personnel 6.2 Protection System and Protective Action System 6.3 Instrumentation and Panel Separation 6.4 Environmental and Seismic Qualification 6.5 Balance of Plant Instrument Setpoint Calculations 6.6 Control Systems 6.7 Indication and Annunciation 6.8 Balance of Plant Design and Field Changes 7 DESIGN CONTROL ASPECTS RELATED TO MORE THAN ONE DISCIPLINE 7.1 Computer Software Control 7.2 Design Capabilities of Harris Plant Engineering Section 7.3 Design Verification Process 7.4 Minor Design Change Designation 7.5 Design Interface Problems Appendix A - Response to Deficiencies, Unresolved Items and Observations l

1. INTRODUCTION AND SLMtARY 1.1 OBJECTIVES AND SIMtARY The Nuclear Regulatory Commission (NRC) on December 3,1984 initiated an Integrated Design Inspection (IDI) of Carolina Power & Light Company's Shearon Harris Nuclear Project. The IDI team's inspection was primarily concerned with reviewing the adequacy of design details as a measure of how well the design process had functioned in a selected sampling area.

The IDI team concentrated its inspection in five engineering disciplines within the project: mechanical systems, mechanical components, civil-structural, electric power, and instrumentation and controls. The IDI . team selected the Chemical and Volume Control System for this inspection. The IDI team also inspected other systems where necessary to evaluate an aspect not afforded by the sample or to expand the inspection to determine the extent of observed deficiencies. After a review of thousands of pages of technical documentation taking approximately 8 weeks by an experienced inspection team, consisting of 13 experienced NRC inspectors and consultants, the NRC staff concluded that the overall Shearon Harris plant design process has been adequately controlled, pending resolution of the IDI items. This report contains Carolina Power

& Light Company's responses to these items.

1.2 CAROLINA POWER & LIGHT RESPONSE Carolina Power & Light Company's senior management has given the IDI Report high priority in the Shearon Harris Nuclear Project effort. The Shearon Harris Nuclear Project has devoted extensive resources to review, respond to and act on the concerns and recommendations contained in the IDI report. Our efforts were immediately initiated as preliminary findings were first communicated by the IDI team during the inspection.

In this regard, many of the corrective actions identified in this report were completed prior to the completion of the inspection and exit interview on February 13, 1985.

Carolina Power & Light is committed to use the Integrated Design Inspection in a constructive fashion. The responses to the individual deficiencies and the corrective actions defined therein are a means of strengthening what is viewed by the IDI team as an already effective design effort. The depth of our commitment can be gauged from the resources we have dedicated to the responses and resolutions of inspection items. To date, our personnel have expended over 2500 workdays in this effort.

In addition to the above described activities initiated by the project, an Independent Evaluation Team has been assembled to assess independently the results of the inspection. This team is composed of individuals from various disciplines and with sufficient technical expertise to reach independent conclusions regarding the adequacy of the resolution of IDI items. To assure independence, the team consists of personnel that have not been directly involved in the supervision or completion of safety related design work on the Harris Project. In order to assure that adequate resolutions to all items have been initiated, the team was required to perfom one or more of the following tasks:

1

r

1. INTRODUCTION AND SUP9tARY (Cont'd) 1.2 CAROLINA POWER & LIGHT RESPONSE (Cont'd)

1. Review proposed responses prepared by Harris Project personnel.

2. Discuss IDI items with the personnel that originally performed the design work.

3. Independently evaluate the technical basis for the proposed responses.

4. Verify corrective action and the implementation of preventive measures. ,

5. Additional inspection of similar items not covered by the NRC.

The Independent Evaluation Team placed particular emphasis on detennining the generic implications of inspection items. Generic issues raised by the NRC during the exit meeting were also evaluated. This report reflects the conclusions of the Independent Evaluation Team.

1.3 RESPONSE FORMAT

-This response has been prepared in response to the NRC's Integrated Design Inspection report number 50-400/84-48. Detailed responses to all Deficiencies, Unresolved Items and Observations discussed in the IDI report are contained in Appendix A. In addition, all major concerns expressed by the IDI team are addressed by discipline in Chapters 2 through 6 of this report and on a generic basis in Chapter 7. As such, this report is similar in format to the report prepared by the IDI team. Where applicable, corrective actions initiated to address the concerns expressed by the NRC are also described.

1.4 MAJOR CONCLUSIONS The Integrated Design Inspection team conducted a thorough, in-depth review of the major disciplines engaged in the design of the Shearon Harris Plant. During the course of the inspection, which included a review of thousands of pages of technical documentation, the IDI team identified several deficiencies. The IDI team concluded, however, that "none of the identified deficiencies, either collectively or individually, are such that the overall adequacy of the Shearon Harris Plant design is called into question, pending satisfactory resolution of the items identified".

We do not agree that all items identified can be legitimately classified as deficiencies, but the Shearon Harris Nuclear Project is confident that the corrective actions undertaken will resolve all concerns to the satisfaction of the IDI team. Conclusions reached with respect to individual disciplines and cross-discipline activities are addressed in Chapters 2 through 7 of this response.

i 2 i

2. MECHANICAL SYSTEMS 2.1 = DESIGN INFORMATION The IDI team evaluated the basis of mechanical systems design information and guidance in the mechanical systems area. The IDI team paid particular attention to the design comitments contained in the FSAR and related correspondence submitted in support of the

. operating license application to determine whether or not the actual Shearon Harris Nuclear Plant's design conformed to the licensing commitments.

The IDI team observed that Ebasco has procedures that describe the methodology for selecting and documenting design inputs and performing a design verification review on design output documents.

The IDI. team also observed that guidance, in the form of mechanical engineering department guides, is available for use. These guides provide mechanical-nuclear design criteria, technical directives,

.and general engineering activities.

The IDI team concluded that these guides, coupled with Ebasco's Company Procedures provide sufficient guidance to accomplish design activities within the. mechanical design engineering discipline effectively.

2.2 CHEMICAL AND YOLUME CONTROL SYSTEM DESIGN The IDI team evaluated the information exchange between Westinghouse and Ebasco pertaining to the design of the chemical and volume control system and to the performance of design analyses for that system. The IDI team confimed that the external interface between Ebasco and Westinghouse was identified; that the responsibilities were defined and documented in sufficient detail to cover the preparation, review and approval of documents involving design interfaces; that a systematic method exists for communicating needed design information across external design interfaces; and that established procedures exist to control the flow of design information between the organizations.

The IDI team identified a minor deficiency (D2.2-1) with respect to the control of three interface drawings. We have reviewed this finding and do not consider this item to be a deficiency, because the design drawings were regularly transmitted to Westinghouse for their information and use. Periodic transmittal of individual design changes (DCNs/FCRs/PWs) to Westinghouse was initiated to enhance an already adequate design interface system. For a detailed discussion of the individual deficiency refer to Appendix A of this response, Item D2.2-1. The IDI team also identifled two minor deficiencies with respect to Westinghouse proof-of-design calculations (D2.2-2) and concluded these deficiencies would have no design impact due to the conservative nature of the errors. The calculations have been appropriately corrected. For a detailed discussion of this deficiency, refer to Appendix A of this response, Item D2.2-2.

3

r 2.3 CONTAINMENT SPRAY SYSTEM DESIGN The IDI taan reviewed the Containment Spray System to assess system design adequacy by examining design details of the system and by evaluating the tecnnical accuracy and control of design analyses to support system design bases.

The review of the containment emergency recirculation sump hydraulics was primarily based on the Shearon Harris Nuclear Project's constitment to comply with the guidelines of Regulatory Guide 1.82. This regulatory guide identifies several design considerations affecting the design of the containment emergency sump fine screen (to preclude excessive flow velocity and potential for clogging by debris), structural integrity and submergence. The IDI team felt that the containment emergency sump screen design was inconsistent with the guidance of Regulatory Guide 1.82 (Deficiency D2.3-1). The IDI team was concerned the emergency sump is not fully submerged during minimum containment water level conditions and that the approach velocity of water at the face of the sump screen does not " approximate" the value identified in the regulatory guide. The team found that no objective consideration was given to the effects of floating debris on sump screen structural integrity or to the potential for increased screen blockage.

The IDI team reviewed the design analysis supporting the structural design of the fine screens and found that the screens were designed to withstand a safe shutdown earthquake, but found no consideration

-of functional loads.

The IDI team also questioned the design methodology employed to calculate the height of sump water required to prevent vortex formation at the emergency sump inlets (Deficiency 02.3-2). The IDI team found that an unsubstantiated assumption was used for the volume of water required to remain in the refueling water storage tank to prevent vortex formation (Deficiency D2.3-3).

i i The IDI team also noted that the capacity of the refueling water storage tank available for injection into the containment building via the safety injection and/or containment spray systems following a LOCA is not consistently stated in the FSAR (Deficiency D2.3-4).

The IDI team was concerned because they felt that Ebasco has a

~

! company procedure which permits the FSAR to be used as a source of

design input.

The IDI team found an incorrect methodology used in the calculation of the hydrodynamic conditions of the eductors. This resulted ".um an inappropriate assum> tion and a mathematical error made dr ing an iterative process whic1 prematurely stopped the analysis before arriving at the correct solution (Deficiency D2.3-5).

4

m , ,

2.3 CONTAINMENT SPRAY SYSTEM DESIGN (Cont'd)

To assess' the ability of the containment spray system to withstand a single. failure, the IDI team examined the system with respect to single failure conditions. The IDI team found a single, normally closed, manual valve used to isolate an emergenWy sodium hydroxide fill connection from the suction side of both eductors (Deficiency D2.3-6). The IDI team concluded that if this valve were mistakenly opened and remained open undetected, a common mode failure of both trains could occur, preventing the addition of sodium hydroxide o.'

possibly causing air binding of the containment spray pumps.

The IDI team concluded that the errors found represent a significant weakness in the implementation of the design control process with respect to containment spray system design.

We have reviewed the IDI team's findings against the containment spray system and have concluded that the system as presently designed satisfies the design requirements outlined in applicable regulatory guidance. Specifically, the system meets or exceeds the requirements provided in USNRC Regulatory Guide 1.8.f. We are fully confident that the design will perform the proper safety function necessary to mitigate the consequences resulting from a LOCA. In addition, since we do not believe that these items are deficient, we therefore do not consider them evidence of a significant weakness in the implementation of the design control process. I As supported by our discussion in the response for item D2.3-1, we do not agree with the IDI team's position that the recirculation sump must be fully submerged under all expected modes of operation.

We do agree that the sump structure should be designed for complete submergence to account for the loads resulting from the forces imposed on the structure. Although the containment recirculation sump is not fully submerged during the minimum containment water level conditions, the velocity at the fine inner screens does approximate the value identified in the regulatory guide (0.2 feet per second).

Because the velocity is low in accordance with regulatory guidance, and the outer trash rack prevents large floating debris from reaching the fine inner screen, the " functional loads resulting from floating debris are judged negligible.

With respect to the IDI team's finding regarding an incorrect design methodology employed to calculate the height of sump water required to prevent vortex formation at the sump discharge, D2.3-2, we note that most analytical methods published on the subject of vortex prevention underestimate the required submergence. As such, the literature recommends that dependence must be placed upon ex >erimental investigations. The tests performed by Alden Research La>s specifically for Shearon Harris and the work done for the USI-A43 Resolution Position support the published literature on this matter. As a result of the plant specific test, the Shearon Harris Nuclear Project is in the process of making the design modification necessary to further minimize the potential for vortex fomation in the sump.

5

U 2.3 CONTAINMENT SPRAY SYSTEM DESIGN (Cont'd)

With respect to the IDI team's finding regarding an unsubstantiated assumption noted in calculaticn EQS-2, for vortex prevention in the refueling water storage tank (RWST), calculation EQS-2 was prepared to establish the RWST level instrumentation set points. The value noted by the team of 54,430 gallons is not stated as the required volume >of water to prevent vortexing in the tank, rather it states that this value is the water available for vortex prevention.

Calculation Tank-13 Rev. 2, which was reviewed by the IDI team, established the minimum water level required to prevent vortexing.

(See the response to Deficiency D2.3-3).

With respect to the IDI team's position that the minimum available capacity of the refueling water storage tank is not consistently stated in the FSAR, we note that the FSAR is not a detailed design tool, but reflects data from controlled documents and sources. When design changes are made on primary documents, the FSAR necessarily lags. The calculation of minimum available capacity of the refueling water storage tank is contained in calculation TANK-13, which was based on guidance provided by Westinghouse and supercedes data presented-in the SHNPP FSAR. The Shearon Harris Nuclear Project has initiated an FSAR consistency review, and the FSAR will be revised to reflect changes made in primary documents. (See the response to Deficiency D2.3-4).

The IDI team's finding concerning the calculation performed for the containment spray system eductor is correct; however, the error was minor and did not result in a significant change in the flow rate.

This error had no impact to other calculations, analyses or design.

We consider this item to be isolated and not indicative of a weakness in the design verification process. (See the response to Deficiency D2.3-5).

With respect to the IDI team's single failure analysis conclusion regarding an inadvertent opening of a nomally closed valve in the emergency NaOH addition line, we note that this valve is nomally closed and is not expected nor required to be opened at any time.

Therefore, the erroneous opening of this valve is considered an unrealistic failure scenario. However, in order to assure a more positive method of maintaining the valve in the closed position, locking devices and administrative controls have been specified for the valve. (See the response to Deficiency D2.3-6).

Considering the above and the detailed discussions presented in the responses to the individual items, the " errors" found by the IDI team do not represent a significant weakness in the implementation of the design control arocess. We do not agree that the type of

" errors" found should 1 ave been identified and corrected by the design verification process. The design of the containment spray system satisfies the design requirements of the system as outlined in the regulatory guidance.

6

__ -1 1

2.4 FIRE PROTECTION The IDI team reviewed fire protection activities related to preparation  :

and maintenance of the fire hazard and safe shutdown analyses. The IDI team found that the documentation reviewed represented an extensive engineering and design effort which was properly reflected in the plant design. Overall, the IDI team found the design analyses they reviewed to be well documented and controlled.

The IDI team reviewed the design analysis perfomed to calculate the combustible load contained .in a typical cable tray and questioned the conservatism of certain assumptions. The justification for the assumptions utilized in this analysis are provided in the response to Unresolved Item U2.4-1.

The IDI team was also concerned that in selected instances, Electrical Engineering approves cable tray fill percentages that exceed the maximum established values without prior review by the Fire Protection group for impact to the combustible load calculations. As outlined in the response to Deficiency D2.4-2, the combustible load calculations are sufficiently conservative to permit deferral of the evaluation of actual cable fill percentages until after cable routing is essentially complete. Fire Protection does not review cable tray overfills prior to release to construction, but will perform a general review prior to fuel load.

The IDI team found an inconsistency in the FSAR between the total combustible load in a given fire zone and the combustible load per unit area. As noted in the response to Deficiency D2.4-3, these inconsistencies have been eliminated and had no effect on the adequacy of the design.

With regard to the field change request which was not reviewed for impact to the plant's Fire Hazard Analysis, an engineering evaluation is being perfonned to assure that there has been no impact to design. In addition, Project procedures have been revised to ensure that the Fire Hazard Analysis is specifically assessed for impact prior to approval of Field Change Requests. Site engineering personnel have been trained on the specific requirements of the procedure and the content of the Fire Hazard Analysis. Additional detafis can be found in the response to D2.4-4, 2.5 RADIATION PROTECTION DESIGN ANALYSES The IDI team reviewed radiation protection design analyses. Particular attention was directed towards the equipment qualification program and the shielding design.

The IDI team identified deficiencies as listed in Appendix A (D2.5-1, D2.5-3, D2.5-4, D2.5-5, D2.5-6 and D2.5-7), and concluded that these are evidence of a significant weakness in implementation of the design control and design verification process by the applied physics department. Our detailed responses addressing many of the specific and general NRC findings appear in Appendix A, and will not be repeated here. After a close examination of the findings, we have concluded that the significance of the discrepancies has been overstated.

7

2.5. RADIATION PROTECTION DESIGN ANALYSES- (Cont'd)

This may be put in perspective by grouping the deficiencies into the following categories: - '

+

1. Deficiencies that we do not agree with or are inconsequential to the results of the calculation. These include findings which, in the professional opinions of the originator and verifier, are  :

correct as originally stated in the calculation, or which have been questioned because additional justification of assumptions were deemed necessary by the NRC. Inconsequential items include minor errors which do not materially affect the results, as well  :

as errors of form (e.g., not fully stating assumptions, or '

documenting data sources).

2. Deficiencies where the calculational data or methodology is  !

incorrect, and the results of that calculation could be materially affected. t Counting the major subdivisions contained in the six deficiencies, '

and taking care not to double count several findings appearing in more than one deficiency, the total number of separate findings is  ;

about thirty. Using the categorization presented above, and in view r of the detailed responses in Appendix A, eighty percent of the items fall into Category 1.

The shielding calculations reviewed are complex, involving hundreds j of pieces of input data, assumptions, and separate computations, and h were, by the IDI team's admission, extremely lengthy. The fact that a number of Category 1 deficiencies were cited should not indicate inadequate design verification control, because many of these items are either disputed or minor in nature. ,

Just six of the items are considered to have a material effect on the calculational'results; the most significant being the inadvertent reduction of the iodine source term. However, these items are i compensated for by other conservative assumptions present in the origina1 ' calculations. Ebasco has redone the calculations affected by these errors, and finds that there is no design impact, and equipment qualification will not be affected except for one area of the RAB which is still under evaluation.

Because the majority of deficiencies noted did not affect the I

results, either because they were inconsequential or were offset by other conservatisms, the conclusion cannot be drawn that the design i i verification was systematically inadequate. Except as noted in one l area of the RAB, all doses and shield walls have been demonstrated to be adequate.

i l

i 8 L

- - - _ , ~ , , , , , , - - _ . _ _ _ - - - - . - - - - - - _ _ . _ _ . . , . , _

'2.5 RADIATION PROTECTION DESIGN ANALYSES (Cont'd)

Nevertheless, in response to the IDI team's findings the applied physics department has taken the following steps:

1. Calculations with significant errors identified by the IDI team were completely redone.

2. The impact of these errors will be assessed to see if they affect any other calculations and, if so, revisions will be made

, where necessary.

3. A departmental program has been initiated to increase both the engineers' awareness of QA requirements and supervisors' participation. This has included a series of QA training seminars, information memos, and group discussions to instruct the engineers on both the form and technical documentation necessary in preparing calculations and the degree of thoroughness required for design verification. Supervisors and other key members of the applied physics staff have been made aware of the IDI findings, and they have been directed to upgrade and monitor continually the key aspects of design control in their groups.

4. A sample of 10 applied physics calculations for Shearon Harris (5 in shielding, 3 in thermal hydraulic, and 2 in applied mechanics) are being audited internally to ensure that the deficiencies identified in the IDI audit are not systematic. If this sample review is not conclusive, the scope of the program will be expanded accordingly.

5. A program has been instituted whereby all nuclear calculations of active projects in the applied physics files are undergoing a page by page review to determine ttat the calculatfor, conforms with Procedure E-30 and that they have been signed by a verifier.

6. A departmental procedure has been amended to require the department file custodian to review all future calculations for proper form and signatures prior to filing.

The IDI team noted that several audits of applied physics, conducted by Ebasco's Quality Assurance Department, were considered unsatisfactory and the lack of improvement was representative of a programmatic concern. The Ebasco Quality Assurance Program, as defined in the Ebasco Topical Report, ETR-1001, has been effectively implemented by Ebasco for several nuclear clients for more than ten years. Ebasco has always been responsive to NRC and client needs and requirements, and has incorporated refinements into the program over the years where an area of possible weakness was identified.

9

2.5 RADIATION PROTECTION DESIGN ANALYSES (Cont'd)

The Ebasco Quality Assurance Internal Audit Program has, since its

~ inception, effectively monitored and improved Ebasco's overall Quality Program performance. As with other areas of the QA Program, Ebasco often had requisite systems in place before they became Regulatory requirements. As new requirements or interpretations became applicable, program modifications have been made to upgrade specific areas of the system, thereby strengthening the overall programs. QA internal audits are scheduled and conducted using documented detailed audit plans, and the results are compiled and reported to the appropriate levels of management on a regular basis.

Corrective actions to deficiencies cited, are verified through the follow-up of the audit.

Ebasco also has recently established a computerized Trend Analysis Program, which tracks the perfomance of individual projects, groups, i

departments, etc. (Prior to establishment of the computer data base, the Trend Analysis was performed by analyzing the hard copy raw i data). All internal audit activities are entered into the computer data base and periodically reviewed, with the results documented in a i

report submitted with recomendations to the appropriate levels of 4

management. A review of the last several semi-annual Trend Analysis

, Reports has not identified any other engineering discipline as

, unsatisfactory.

f With regard to the concern expressed in the IDI Report relative to the audits performed in applied physics, a review was made of all QA

! internal audits of this discipline since 1976, in order to determine:

1) areas found discrepant, 2) what remedial action was required and i 3) whether the remedial action was verified as having been

completed. Of thirty-one audits performed and documented, twenty-six

!' required some form of corrective action that was carried out and verified. Many of the audits were designated as unsatisfactory as a result of Ebasco revising its internal audit program in 1975, when an approach was instituted which established a determination of

" satisfactory" or " unsatisfactory" based solely on a numerical i.

percentage of documents considered deficient using a " weighted" deficiency scale. For example, an audit of four sets of calculations j with a deficiency identified against one set of calculations could

! lead to a 25% deficiency rating and a designation of unsatisfactory.

This could be unreasonably harsh when one considers the overall

! amount being audited and the quantity and nature of the condition j cited (eg: procedural violation on isolated pages in a single 4

multipage document). Our review has established that the discrepancies were corrected, but even if subsequent communication

produced an agreement that an overall rating of unsatisfactory was j unreasonable, the original audit report remained unchanged.

l l

i

[

i j 10 i

2.5 RADIATION PROTECTION DESIGN ANALYSES (Cont'd)

The IDI Report also indicates that subsequent to the 1975-78 period, only one follow-up audit in the area of radiation protection was performed (in 1982). That audit concluded that the work reviewed was unsatisfactory and no record of satisfactory follow-up could be found. A review of our records indicates that satisfactory corrective action was verified for Audit Number 1792, on July 20, 1982 and the audit was closed. Additionally, our records indicate that audits of applied physics (shielding) were performed on April 25,1979 (audit report 1033) and November 6,1979 (audit report 3

1131).. In both cases, minor deficiencies were noted, with corrective action taken and subsequently verified by QA.

Ebasco believes that the audit program in place has effectively responded to Ebasco, client and regulatory requirements since its inception. However, based on our own observations, as well as those

included in the IDI Report, some refinements in the QA Audit program will allow the system to better address positive corrective action

! where such action may be indicated by successive QA audits.

Therefore, Ebasco has initiated steps to provide for the following:

1. Semi-Annual Summary Reports from QA to Management or audited groups or disciplines, which will include recomendations for corrective action to preclude recurrent discrepancies (where

warranted), and will require a documented response as well as follow-up to verify implementation. This will provide a " closed

loop" comunication as opposed to the current system.

2) A revision to the current Ebasco Quality Assurance Implementing Procedure for Internal Audits has been made in order to provide clarification and emphasis that each audit must provide for a determination of the need for corrective action to preclude i recurrence of conditions noted.

3) The Ebasco Management Audit Committee will be requested to perform a review to determine if any other modifications to the QA Internal Audit Program are warranted.

We believe that these additional measures will provide for tighter

, administrative controls, improve overall effectiveness, and preclude a recurrence of the condition identified in the IDI Report.

2.6 ENGINEERED SAFETY FEATURES VENTILATION SYSTEM DESIGN 4

The IDI team performed an evaluation of the seismic analysis of safety related air handling units and the installation of th+ units in the plant. The IDI team's review found the seismic analysis to be adequate but identified a concern with respect to the l installation of these units (D2.6-1).

i 1

11

(

- - - . . . w%._.-,, __~m_ _ . - - ,._.,,.w.---._,,.-p. ..~_.yr. . , _ . . . g.,,, ~,r-,...-,_,.9,y ,-,c.,w.-m... e_m_w.-w.y-

2.6 ENGINEERED SAFETY FEATURES VENTILATION SYSTEM DESIGN (Cont'd)

Of specific concern was the torque values used for the anchor bolts of the air handling units. While the anchor bolts were not torqued to assumed values as specified by the AISC Manual of Steel Construction, the Shearon Harris Project consultant, who perfonned

-the analysis, has confirmed that the anchor bolts torqued to a " snug tight" condition as defined by the AISC, 7th edition is adequate.

" Snug tight" is required by the installation procedure where torque values are not shown on the installation drawings. Alleged incomplete field installation packages for the air handling units have been reviewed and are complete and the installation procedure has'been determined to be acceptable.

2.7 SEISMIC DESIGN OF COMPONENTS NOT REQUIRED FOR SAFE SHUTDOWN The IDI team evaluated the adequacy of design activities performed to demonstrate that postulated failures of non-seismic components would not damage safety related equipment during a safe shutdown earthquake.

The IDI' team found that a non-seismic piping interaction study was' performed by the mechanical-nuclear discipline. The team found these design analyses to be unchecked preliminary calculations (Deficiency D2.7-1) and concluded that the review of unchecked design analyses would not provide useful conclusions as to the adequacy of the design control process. The results of these unchecked and unverified calculations appeared to be the bases for reporting in the FSAR that failure of any non-seismic Category I piping system would not cause failure of seismic Category I structures, piping, or equipment essential for safe shutdown.

The IDI team alleged that a well controlled design process was not in effect at Ebasco to ensure that systems and components, whose functions may not be required for safe shutdown following i safe shutdown earthquake but whose failure may prevent a system important to safety from functioning, were designed to survive the earthquake. The IDI team concluded that previous experience at other facilities has indicated that walkdowns are difficult to perfonn and should not be relied upon to identify and resolve problems during late stages of construction.

We have reviewed the IDI team's finding with respect to the considerations of seismic design of components not required for safe '

shutdown. Based upon our review, details of which are presented in the response to D2.7-1, we agree with the IDI team's finding that unchecked calculations should not be used to form the basis for reporting commitments in the FSAR. However, we do not agree that a seismic interaction study based upon existing design documents should be performed.

12

2.7 SEISMIC DESIGN OF COMPONENTS NOT REQUIRED FOR SAFE SHUTDOWN (Cont'd)

Perfomance of the previously mentioned non-seismic piping interaction study (albeit unchecked), established design criteria for.each major discipline, and the design control process, provide assurance that adequate attention has been focused during the design phase on seismic design of components not required for safe shutdown. In addition, the Shearon Harris Nuclear Project is currently performing a Seismic II/I walkdown of safety related areas of the plant to ensure further that unacceptable seismic interactions are avoided.

The IDI team's conclusion, that previous experience at other facilitie:; has indicated that walkdowns are difficult to perform and should not be relied upon to identify and resolve problems late in construction, is not consistent with industry direction in this area. Walkdowns have proven effective in identifying interactions that would otherwise not have been identified during the design phase of the project. Walkdowns that are scheduled consistent with thes overall integrated plant construction, testing start-up and operation schedules will have minimal impact on final plant operation.

Initial results of the Seismic II/I Walkdown indicate that there is not an excessive amount of rework, thereby substantiating our position.

2.8 FIELD INSTALLATION OF EQUIPMENT IN CLOSE PROXIMITY TO PIPING SYSTEMS The IDI team conducted a limited walkdown of the installed chemical and volume control system piping to confirm that sufficient clearance is provided between installed equipment for seismic and thermal movement of piping systems.

The IDI team found an ASME III, seismically supported charging and volume control pipe in physical contact with two heating and ventilating dampers (Deficiency D2.8-1). Although clearance dimensions were identified on an Ebasco prepared drawing, it appeared to the IDI team that the field construction forces were not observing these requirements.

Because the clearance requirements for seismic and thermal movement are not being enforced as inspection criteria, walkdowns to identify and resolve pipe clearance problems will have to be perfomed. The IDI team is concerned that excessive reliance is being placed on field walkdowns late in the construction.

The IDI team believes that a quality assurance inspection / verification program should be instituted to ensure that adequate piping tolerances are maintained as piping or other equipment are installed.

13

I 2.8 FIELD INSTALLATION OF EQUIPMENT IN CLOSE PR0XIMITY TO PIPING SYSTEMS (Cont'd)

Although specific guidance is provided to construction to minimize the number of instances where equipment comes in close proximity to piping systems, it is difficult to guarantee that adequate interdisciplinary clearances exist. As such, the Shearon Harris Nuclear Project is currently performing an interdisciplinary clearance walkdown to identify areas where inadequate clearances have been provided. We do not agree that excessive reliance is being placed on this field walkdown. As part of the construction inspection program, the Shearon Harris Nuclear Project checks for piping tolerances and interferences.

2.9 EMERGENCY SERVICE WATER SYSTEM The IDI team conducted an inspection of the emergency service water system to determine that adequate NPSH at runout conditions and the minimum critical submergence for the pumps are provided.

The IDI team performed a limited review of the emergency service water pump hydraulics. One aspect of the inspection involved a comparison between the pump purchase specification and the pump vendor's certified drawing. The IDI team fcund that the pump vendor's drawing and purchase specification disagreed with respect to the critical submergence (Deficiency D2.9-1). The vendor's drawing was found to have an incorrect elevation for the critical submergence.

As discussed in the response to D2.9-1, the pump minimum submergence is correctly noted on the pump vendor's drawing. The vendor, however, incorrectly stated the water elevation corresponding to the minimum submergence required. Since the specification and the drawing consistently stated the minimum submergence required, we do not agree that the elevation error should have been detected during review of the pump vendor drawing. Notwithstanding the above, the drawing and instruction manuals have been revised to reflect the correct elevation corresponding to the minimum submergence.

14

i

• i 1

3. MECHANICAL COMP 0NENTS 3.1 DESIGN INFORMATION The IDI team evaluated design documents for confomance to FSAR commitments, consistency and technical accuracy, and reviewed the ,

analysis and design of equipment, piping and supports for '

conformance to the governing design specifications. The IDI team concluded that the Shearon Harris Nuclear Project specifications ,

are acceptable.  !

L The IDI team recommended that Ebasco Shearon Harris stress analysis i calculation procedural guidelines be controlled. This recommendation is being implemented. Under the calculation '

transfer program, all stress analysis calculations are scheduled to be transferred to the site by the end of August 1985. As such, the  !

l HPES Instructions Manual will be used to control the Calculation

Procedural Guidelines.

l The IDI team also expressed some concerns with the Harris Plant

, Engineering Section Manual of Instructions. The Piping i Support / Restraint Design Guideline and the Site Hanger Problem Resolution Guidelines have been integrated into a single document with deflection criteria consistent with a pending FSAR Amendment

. to Table 3.9.3-7 (D3.1-4). '

i.

Nuclear restrictions for non-seismic use of tabular data for pipe t

support supplementary steel have been corrected in the above >

l engineering guidelines. These restrictions, taken from the

!' Bergen-Paterson Project Instructions, were clearly non-seismic and }

were only used by Bergen-Paterson as design criteria for new design '

within their scope (D3.1-5). The check for maximum shear stress in a pipe due to torsion and shear loads has been corrected to combine directly torsional and direct shears. However, since supports are usually deflection limited not stress limited, there is little 1 j potential for significant errors. A review is being performed to

confirm that this incorrect stress check has not been incorporated i into site pipe support design work (D3.1-6). Also, the straight i line interaction equation to select U-bolts subjected to combined

tension and shear has been confirmed as appropriate by

. Bergen-Paterson. Qualification of these U-bolts was by linear i analysis (D3.1-7).

t T

! l i -

1

' [

15 4

3.2 PIPE STRESS INPUT AND OUTPUT The IDI team reviewed the input design infomation used for the stress analysis of the chemical and volume control system. The IDI team also reviewed the evaluation of the output data. Based on this

, review, the IDI team questioned the effectiveness of the checking process.

Stress analysis of nuclear safety class piping requires the entry of hundreds of data inputs to each of many computer calculations.

Accuracy of results is contingent upon the stringent selection of appropriate inputs by the isometrician and analyst as well as correct entry by the keypunch operator. Use of current design documents and selection of appropriate loading data and criteria are essential.

The IDI team has noted in its review that the Harris Project engineers and designers have been diligent and successful in this process.

Recognizing the fact that even with extreme caution, errors of judgement and/or carelessness are inevitable, the Code of Federal Regulations requires checking to verify inputs to safety calculations. Furthermore, recognizing that even with independent verification of calculations, some percentage of errors is unavoidable, analytical fomulas and criteria are designed to contain a generous degree of conservatism.

The IDI team identified several examples of isolated, non-programmatic errors of data input in the sample group of calculations they reviewed. Some of the perceived errors cited have been explained by Ebasco and are indeed not errors. Admittedly, a pressure input, a c.g. location, a nozzle thermal displacement, etc were found to be in error. Many of these were obviously of little impact to the stress calculation results. It is quite possible that these were, indeed, found by the checker and deemed negligible.

While it is clear the documentation of such judgements should have been provided, a lack of notation does not compromise the accuracy of the calculation.

The errors brought to Ebasco's attention have been evaluated by reanalysis and, as expected, they have not affected results to the point of requiring design modifications. This demonstrates how criteria conservatism augments the limits of accuracy inherent in the complicated task of bringing so much and varied data into one calculation. Indeed, the system has been proved effective.

Through procedures, and quality and training of persor.nel, a high degree of accuracy has been achieved in the conduct of safety analyses. Therefore we are confident in their results.

The IDI team also identified isolated filing problems where check lists of equipment evaluation sheets were missing from calculation packages. However, the overall status of the stress analysis files was proven adequate. It must be noted that missing documentation, when relocated or reproduced, demonstrated design and analytical acceptability versus appitcable criteria.

16

3.2 PIPE STRESS INPUT AND OUTPUT (Cont'd)

The following responses address specific concerns identified by the IDI team:

1. Equipment Overloads The IDI team noted the Ebasco policy of releasing piping drawings for fabrication and analysis when the supporting calculations show overloads of equipment nozzles, valve g-loading in excess of criteria, flange overloads or penetration overloads. This policy is consistent with general industry practice.

During early design stages, much of the data entered in piping calculations are assumed (insulation weight, equipment or building movement, valve dimensions and weights, etc.) due to the fact that not all of these components are procured at the time these initial analyses are performed. These calculations confirm the acceptability of preliminary routings and layouts. As more data is finalized, equipment and valve loadings are checked versus the best available criteria, which may later be modified. Often criteria is violated slightly, but experience has shown that when all data is finalized and allowables are stripped of excess- conservatism, the design is proven satisfactory.

Rather than hold construction for resolution of each overload prior to drawing release, it has become general practice to follow-up on overloads via specialized comprehensive programs and resolve any problems analytically or by physical changes. The latter option is rarely excerised. This decision minimizes repetetive interfacing with equipment vendors and the associated monetary impact. This is a prudent course of action and represents sound engineering judgements.

Although some older calculations are without vendor criteria evaluations and others fail to meet vendor allowables, the programs already in-place prior to the IDI assure the ultimate resolution of these items.

2. DBE Evaluation The IDI team questioned Ebasco's assumption in qualifying DBE loading cases by the piping's ability to meet upset criteria under OBE loading. Although the IDI team agreed that this is generally a safe assumption, they felt more rigorous evaluation is warranted.

To justify the assumption that the upset condition is limiting, a tabulation of maximum stress ratios for the upset and emergency conditions was assembled. Ten randomly selected calculations, in which DBE spectra were analyzed, were reviewed. This review found the emergency stress levels ranged from 0.99 to 1.24 times the upset levels. Since the emergency stress limit is at least 1.5 times the upset allowable, the DBE condition was found not to be limiting, thus demonstrating the validity of our judgement. (See response to Deficiency D3.2-1).

Nonetheless, Ebasco has revised stress analysis procedures to require the evaluation of the DBE case in all future calculations.

17

3.2 PIPE STRESS INPUT AND OUTPUT (Cont'd)

3. The IDI team questioned the validity of the g-load calculation for the valves of Calculation 141-2. Although the method of analysis was not apparent during the inspection, subsequent review by Ebasco confimed the adequacy of the technique used. The technique was then documented and inserted in the calculation package.

4. The IDI team cited a valve in Calculation 3001 (D3.2-5) for which the current weight was not used. However, the analyst used the weight which was correct in accordance with the then-current revision of the vendor drawing.

The Shearon Harris Project has reviewed all concerns identified by the IDI team and concludes that the stress analysis program was ef fective. This is demonstrated by the fact that all additional analyses performed subsequent to the IDI have demonstrated design adequacy. The actions described above in conjuction with the checking activities being performed as part of the stress calculation transfer program make the existing calculations suitable for use in the Harris Plant 79-14 effort.

3.3 PIPING STRESS PROCEDURES The IDI team evaluated the modeling procedures which Ebasco used for piping stress analysis. In this regard, the team specifically reviewed Ebasco specification CAR-SH-M-71, the Ebasco Shearon Harris Procedural Gudielines, and the Ebasco piping program PIPESTRESS 2010. The IDI team concluded that the design information and instructions contained in these documents should enable Ebasco to perform piping stress analysis in a consistent and timely manner.

The IDI team noted that Ebasco procedures did not address flexible valves in piping systems since all valves on the Harris Project were specified to be rigid. ITT-Grinnell, however, has determined that certain valves originally thought to be rigid are actually flexible. This has necessitated the development of an analysis nethod capable of addressing flexible valves. This is addressed in the response to Unresolved Item

. U3.3-1.

3.4 PIPE SUPPORTS The IDI team evaluated the modeling procedures used by Bergen-Paterson to analyze and design pipe supports and found them acceptable, except for the design of slender struts. The IDI team considered the design of slender struts to be systematically deficient becuase of the absence of a design basis that enables an evaluation of the effects of dynamic excitation and eccentricity.

We do not agree with this assessment. Project specific licensing comitments, as outlined in the FSAR (Section 3.9.3.4 and Table 3.9.3.7),

do not require vibrational analysis of pipe support structures. To control the vibrational responses of the structures, displacement limits are specified in the FSAR. This approach is consistent with industry practice.

18

3.4 PIPE SUPPORTS (Cont'd)

The design of long structural members is further controlled by limitation of the effective slenderness ratio (KL/r) to a maximum of 200. In addition, when a cantilevered member is used in conjunction with a strut or a snubber, an out of plane load is incorporated into the design of the member.

To verify that the present design criteria provide adequate protection against self excitation for long cantilevered structural members, several sample cases which by inspection appear to be limiting were evaluated. Dynamic analyses using applicable floor response spectra were performed. Stresses and displacements from the dynamic analysis were superimposed on those resulting from design loads. The results, which show total stresses and displacements to be less than half the allowable, demonstrate the conservatism of the present design criteria.

However, to ensure Support / Restraint structural integrity, we are revising the design guidelines to incorporate additional limits on unbraced catilevers. All safety related supports undergo a final calculation review after installation. Those revised prior to incorporation of the cantilever limits will rechecked to assure compliance. (See response to Deficiency D3.4-1).

3.5 MECHANICAL EQUIPMENT The IDI team evaluated the Westinghouse NSSS interface for equipment located in the Chemical and Volume Control System concentrating on the following major components: (1) charging pump, (2) regenerative, letdown, excess letdown, moderating, letdown chiller and letdown reheat heat exchangers, (3) boric acid tank, and (4) cation bed demineralizer.

The IDI team concluded that Ebasco implemented the interface criteria detailed in the Ebasco Specification No. CAR-SH-M-71 in a consistent manner. Westinghouse-procured equipment nozzle thermal displacements and spring-mass models that Westinghouse fomulated for flexible equipment were properly coded into the stress calculations.

19

4. CIVIL / STRUCTURAL 4.1 DESIGN INFORMATION The IDI team reviewed the civil / structural design information and personnel qualification and training.

The IDI team concluded that the design information reviewed meets the FSAR commitments and regulatory requirements. They reported that Ebasco training programs in civil / structural disciplines are acceptable.

4.2 SEISMIC ANALYSIS The IDI team reviewed the seismic analysis of the Shearon Harris Project major structures (Containment Building, Reactor Auxiliary Building, Diesel Generator Building, Emergency Service Water Intake Structure and Tank Building) to verify that the FSAR commitments and NRC-regulations were complied with.

In general, the IDI team determined that the seismic analysis design process, from the formulation of mathematical seismic models to development of floor response spectra curves, is controlled: the artificial earthquake motion enveloped the provisions of Regulatory Guide 1.60; the structural damping values used for cach component agreed with the provisions of Regulatory Guide 1.61; and model responses were combined in accordance with the provisions of Regulatory Guide 1.92. The documentation of the Ebasco in-house computer program DYNAMIC 2037 was also reviewed and found acceptable by the IDI team.

A total of four (4) deficiencies were reported by the IDI team, with two (2) unresolved items and one observation.

The unresolved items (U4.2-1 and U4.2-3) dealt with coupling effects of the'three-directional (2 horizontal and 1 vertical) earthquake on the seismic responses, including the development of floor response spectra curves.

The seismic analyses of symmetrical buildings have been performed by computer program DYNAMIC 2037 in accordance with the FSAR

commitment. For these structures, the coupling effects are not significant. For unsymmetrical structures having significant i eccentricities between mass and rigidity centers, a i three-dimensional torsional analysis, by STARDYNE, as required by

! FSAR commitment, was performed.

1 j At the request of the IDI team, a new three-dimensional seismic i analysis was performed on the Reactor Auxiliary Buf1 ding (with i significant eccentricities) to evaluate the coupling effects on the

! seismic responses. The results indicated that the coupling effects j were not significant and that the original FSAR commitments were adequate. Based on this analysis and justification provided to the l IDI team during the inspection, both unresolved items (U4.2-1 and U4.2-3) have been closed.

20

4.2 SEISMIC ANALYSIS (Cont'd)

The IDI team reported two items requiring reanalysis to justify the validity of assumptions concerning exclusion of mass moments of inertia in symmetrical structures, e.g., Containment Building (Deficiency D4.2-2) and shear area used (Deficiency D4.2-7) in the seismic analysis.

Confirmatory analyses were performed to assess the effect of mass moments of inertia in symmetrical structures. The results confirmed ,

that the effect of exclusion of mass moments of inertia on structural responses in symmetrical buildings is insignificant (see

. response to Deficiency D4.2-2).

Confirmatory analyses were also performed for three structures (Tank Building, Emergency Service Water Intake Structure and Main Dam Spillway) utilizing the actual shear areas. These results confirmed that the existing design is adequate (see response to Deficiency D4.2-7).

The other three deficiencies / observations (D4.2-4, D4.2-5 and 04.2-6) identified by the IDI team were related to documentation aspects which had no impact on the design. These . items identified inconsistencies between design documents and 'similar minor discrepancies, which have been corrected.

The IDI team concluded tnat the current design is adequate, even though reanalyses were required. These analyses. confirmed the adequacy of existing'FSAR commitments and the design.

4.3.CONTAINMENTBUILDINGFOUNDATIONMAb- . .

s - ~ , . . - - .

TheIDIteamreviewdoihestructurai'xdesign and analysis of the containment comitments andbuilding NRC reguiat6ns. fouMatfon; irit to verify'.confSrmance e ,'" - with FSAR The !IDI team fou id'that,. the f andath mTt a ad o ately. designed

.in accordance' with

the design criteria: loadi and leading ,

l combinations used for design and analysis were in c5nformance with the FSAR commitant;,the factors of3 safety against overturning, sliding and flotation were acceptable;' radial and circumferential

. reinforcercnt nre provided as required i3y anglysis and design. '

s '. - ~

a .. 3 One deficiencyJ(D4.3-1) was found by the IDI team concerning documentation of rad!al reinf@ceraant- St7ess calcjlatioiis. This L deficiency had no: Impact'o'rEtne design'.

, .. w. , t ." A s l The IDI te' ad; conc 1d.ied t'nat the desf<jnCof the fo'uiidation mat is acceptable. - t N S ^

4

.s 1 sgs, s

j y s s .

e.

h** <*  % ,N 4

• 1 , i

• % A,,  %

e a _

.. ~

,, ,y h

j .  ; y ' ., hj^ g.%  % ~' ,

y ~

y ,, y>

l . .-. i__ _

_ _ Ju.a . l_ .__ _ . .[ $'? . . _ _ ( , _ _ _ _ . . _ _ , , _. _

'l 4;4 CONTAINMENT BUILDING AND INTERNAL STRUCTURES The IDI team reviewed the design of Containment Building and Internal Structures including the Primary.and Secondary Shield Walls, Reactor Pressure Vessel supports, steam generator supports,

. and operating structural steel floors and platforms.

The IDI team determined that the criteria usec in the design and analysis of the structures was consistent with established criteria, i codes, specifications and FSAR commitments. The IDI team also detemined that design information was correctly transferred to the drawings and that the reinforcing steel details were accurately i

shown.

The IDI team reported one concern (D4.4-1) with regard to Cesica of the containment building polar crane runway ar.d structural steel platforms. The IDI team's description of the deficiency is not entirely correct. The structures were designed for a conservative seismic acceleration of 1.0g in all three directions (with the exception of the polar crane runway where seismic stress due to the runway girder's own dead weight in the vertical direction was neglected). The design is conservative because 1.0g envelopes all the applicable g values and the runway girder dead weight is

! negligible as compared to weight of the polar crane. Design calculations have been revised to document the comparison of g i values and loading combination. The design is adequate.

An observation (04.4-2) was also made by the IDI team concerning a documentation problem. The original calculations for steam i

generator support anchor plates and tank anchorages, which were inadvertently voided, were reinstated (see response to 04.4-2).

The IDI team concluded that the structural design of the containment building and internal structures is acceptable and meets the FSAR commitments.

4.5 REACTOR AUXILIARY BUILDING The IDI team reviewed the design of the Reactor Auxiliary Building.

! The review included design calculations for reinforced concrete

, floor slabs at various elevations, and analysis and design of safety

_ related masonry walls.

The IDI team reported that the overall design of the Reactor

! Auxiliary Building meets FSAR commitments. The-IDI team determined that some local areas required additional analyses to confirm the structural adequacy of these portions.

l

' The IDI team identified six (6) concerns. Three addressed the design of floor slabs while the others addressed masonry walls.

l L

) 22 L

4.5 ' REACTOR AUXILIARY BUILDING (Cont'd)

Deficiency D4.5-1 addressed use of the " Direct Design Method" of the ACI 318-71 code. This method was used for manual design of the slabs in the Reactor Auxiliary _ and the Fuel Handling. Buildings. For certain slab panels, the limitations of the method were exceeded.

The IDI team determined that this deficiency appeared to be systematic. Based upon our review, the dt.ffciency was not found to be systematic but limited to the cited buildings. Confimatory calculations were perfomed using alternate methods which have verified that the existing design is adequate (see response to Deficiency D4.5-1).

Deficiency D4.5-2 addressed an inconsistency between the calculations and the drawings concerning spacing of reinforcement.

This inconsistency did not affect the design since the actual reinfortement provided on the drawing was greater than the design reinforcement requirement (see response to Deficiency D4.5-2).

Deficiency D4.5-3 addressed the load combinations for portions of the slab at El 236.0 ft which do not meet the FSAR requirements.

The deficiency was a result of changes in the design criteria. The original design, based on conservative boundary conditions is adequate; however, no statement to this effect was provided in the calculations. Calculations for the revised loading combinations were performed and documented (see response to Deficiency D4.5-3).

Deficiency D4.5-4 addressed the design of solid masonry walls in the vicinity of volume control tank for which there are no seismic analysis calculations. These walls were qualified "by similarity" to another seismically analyzed masonry wall. However, due to height differences in the walls, the applicability of comparison was questioned by the IDI team. Therefore, seismically designed masonry walls analyzed by similarity are being reviewed for applicability to the typical wall panels (see response to Deficiency D4.5-4).

Deficiency D4.5-5 addressed the use of unbroadened floor response spectra curves for the hollow masonry walls around stairway A-4.

The wall was subsequently redesigned using the broadened response spectra in connection with an "as-built" condition which voided the original design (see response to Deficiency D4.5-5).

Deficiency D4.5-6 also addressed the masonry wall around stairway A-4, where the final resolution of a Permanent Waiver (PW AS-1045) contained information different from that used in the design analysis. The final resolution was reviewed prior to release to the field and found to be acceptable. The design calculation, however, was not revised (see response to Deficiency D4.5-6).

Based on the above, the IDI concerns have been fully addressed and the local areas requiring review are being documented.

i 23

4.6 DIESEL GENERATOR BUILDING The IDI team reviewed the design of the Diesel Generator Building including the design specification, manual computations, computer analysis and structural drawings.

The IDI team determined that extensive analysis has been perfonned for the building and that the design calculations were well organized and easy to follow.

The IDI team concluded that design of the building was in full compliance with the FSAR commitments and the associated Regulatory Guides. No deficiencies were reported.

4.7 TANK BUILDING The IDI' team reviewed the structural design and analysis of the Tank Building including the tanks.

The IDI team determined that the structural design and analysis of the Tank Building are acceptable and consistent with FSAR commitments, NRC regulations, and ACI318-71 code requirements.

The IDI team noted that the loads and loading combination pertaining to the tanks and anchor bolts were revised, after the original design was completed, due to a change in the tank vendor. The new tank loads, which were less than those used in the original design, were reviewed and documented by Ebasco.

The IDI team reported no deficiency. However, an observation (04.4-2) relating to documentation of the design calculations for tank anchorages, was reported. With the reinstatement of the original calculation, this condition has been corrected (see response to 04.4-2).

4.8 MAIN DAM SPILLWAY The IDI team reviewed the design specification, seismic analysis, structural analysis and design of the Main Dam Spillway.

The IDI team identified three concerns. Deficiency D4.2-7 questioned the conservatism of the shear areas used in seismic analyses of the structure. Confirmatory analyses were perfonned l which demonstrated the adequacy of the existing design. (See i Section 4.2, also response to deficiency D4.2-7).

24

4.8 MAIN DAM SPILLWAY (Cont'd)

I Deficiency D4.8-1 noted that the load combinations for the design included the Design Basis Earthquake (DBE) condition but did not include the Operating Basis Earthquake (0BE). Our position is that the simplified conservative design procedure used was judged to be adequate for the OBE condition.- Additional analyses have been

, performed for the OBE condition. The design is adequate (see response to D4.8-1).

Deficiency D4.8-2 described an arithmetical error which resulted in

the overestimation of the weight of the abutment. The abutment wall was reanalyzed using the correct weight and found to be adequate (see response to D4.8-2).

j The IDI team determined that the Ebasco design specification is in accordance with the FSAR commitments. The IDI team did not believe that the above deficiencies would impact the original design. This has been confirmed.

.t 4.9 ANALYSIS-AND DESIGN OF MECHANICAL COMPONENT SUPPORTS

, The IDI team has reviewed the supports for various tanks and equipment, (including the charging safety injection pump, regenerative heat exchanger and air handling units in the Reactor Auxiliary Building (RAB) and Fuel Handling Building (FHB). The IDI team also reviewed the design of the embedded plates associated with the chemical and volume control system pipe supports.

The IDI team determined that the design of these supports is acceptable and in accordance with the FSAR commitments.-

One concern (D4.9-1) was identified by the IDI team. The IDI team was concerned that the design of the slab panel at El.236.0 ft (FHB) supporting the boron recycle hold-up tank, did not consider local j effects of horizontal seismic loads of the tank. The slab was

! designed for 1.25 times the applicable dead and vertical seismic i

loads of the tank, which were judged to be adequate for such local effects. New calculations for the applicable slab panels, including horizontal seismic loads, demonstrate the correctness of that

'. judgement and the adequacy of the existing design (see response to l Deficiency D4.9-1).

l l

i 25

4.10- DESIGN OF ELECTRICAL CABLE TRAY / CONDUIT SUPPORTS

^

The IDI team reviewed the design criteria, cable tray qualification report, and supports design. One concern (D4.10-1) was reported which identified a particular type of longitudinal bracing without documented frequency calculations.

Cable tray supports were designed by analysis, similarity or by inspection / judgement. Specific references defining the design and method utilized were not in all cases-provided.

In general, the same group of experier.ced design engineers have been involved in the production effort of cable tray restraint analysis and design. This continuous participation in establishing and evaluating frequency characteristics and the state of stress for large numbers of supports has provided the lead design engineers with the ability to predict expected characteristics of the support without performing an analysis. Similarity to supports which were previously analyzed and found to be acceptable was used as a design basis. The proper references to similar and previously qualified supports or clarifying statements in support of engineering judgements used, however, were not always provided.

The adequacy of the present design for those supports which may not have a specific qualification analysis is not in question. The inherent conservatism in the approach and the experience and judgement of the design engineers insure a safe design.

Nevertheless, a program to evaluate the calculations for all supports shown on sample drawings has been initiated. This program will identify supports which have not been specifically qualified by stated similarity and either identify and document the similar cases, or perform and document the actual analysis (see response to Deficiency D4.10-1).

4.11 DESIGN OF HVAC DUCT SUPPORTS The IDI team reviewed the design criteria, duct qualification and design of HVAC duct supports. Three (3) concerns were identified.

Deficiency D4.ll-1 noted that in some cases, duct frequencies below 33 Hg could result from spans exceeding the maximum allowed. The i requirement of 33 Hz for ducts is conservative due to the fact that cable tray and HVAC ducts supports are covered by the same criteria and cable trays are more flexible than ducts. However, f' program was initiated to identify all such cases. Evaluation to i confirm support and duct adequacy has been initiated as recommended

by the IDI team.

Deficiency D4.ll-2 addressed duct loads which in some cases were not c

in agreement with actual conditions. All load calculations have i

been reviewed and all cases where the actual loads are higher than

! - those originally used in the supports design were identified.

( Supplementary reviews and analyses to confirm design adequacy are in progress.

i

(

26

4.11 DESIGN OF HVAC DUCT SUPPORTS (Cont'd)

Deficiency D4.ll-3 addressed cases of HVAC duct supports without documented frequency calculations. HVAC duct supports were designed by analyses, similarity or by inspection / judgement. Specific references defining the design method utilized were not in all cases provided.

In general, the same group of experienced design engineers have been involved in the production effort of HVAC restraint analysis and design. This continuous participation in establishing and evaluating frequency characteristics and the state of stress for large numbers of supports has provided the lead design engineers with the ability to predict expected characteristics of the support without performing an analysis. Similarity to supports which were previously analyzed and found to be acceptable was used as a design basis. The proper references to similar and previously qualified supports or clarifying statements in support of engineering judgements used, however, were not always provided.

The adequacy of the present design for those supports which may not have a specific qualification analysis is not in question. The inherent conservatism in the approach and the experience and judgement of the design engineers insure a safe design.

Nevertheless, a program to evaluate the calculations for all supports shown on sample drawings has been initiated. This program will identify supports wnich have not been specifically qualified by stated similarity and either identify and document the similar cases, or perform and document the actual analysis (see response to Deficiency D4.ll-3).

< r e

27

5. ELECTRICAL' POWER 5.1 DESIGN INFORMATION The IDI team reviewed the design criteria and the transfer of design i information within and among disciplines. More specifically, the l IDI team reviewed the Ebasco company procedures for division of l responsibility, preparation of calculations, _ design input and verification, interface with the NSSS vendor, and the Project Design Basis Documents.

The IDI team concluded that there was a good basis for design information and control in the Ebasco electrical group and that the ,

Project Design Basis Documents did provide the necessary documentation for design infomation about the electrical systems.

The IDI team identified one area of concern regarding Electrical interface with Westinghouse (NSSS) regarding approval of a departure ,

from the Westinghouse guidelines for cable pulling tension and length (US.1-1). As stated in the response to US.1-1, we do not consider this item a failure to obtain Westinghouse approval since a letter was issued to Westinghouse delineating Ebasco's exceptions.

" Approval by omission" was the format of the letter which we concur may not be the best method to approve / verify, but a transmittal of correspondence did occur. Fomal approval by Westinghouse has subsequently been received.

The IDI team was also concerned that the electrical Design Basis Documents were not in general use in the site electrical group.

The Design Basis Documents (DBDs) define the function, general design basis, system test acceptance criteria, design margin and document references for electrical systems. It is the main objective of these documents to identify, collect and organize the design basis documentation required for permanent retention and use in future design work and plant operations.

At the time of the IDI,ll of the approximately 15 DBDs related to electrical were written and under various stages of review, however-nona were yet approved. The completed DBD's will form a substantial part of our design control process when they approved and invoked. ,

5.2 PERSONNEL AND GUIDANCE The IDI team evaluated the degree of education and actual design '

experience of personnel engaged in the design effort on the Shearon Harris Nuclear Project and also reviewed the guidance provided to the personnel.

28

\

5.2 PERSONNEL AND GUIDANCE (Cont'd)

The IDI team concluded that the technical guidance and training-provided by Ebasco for its personnel fully meets the intent of ANSI 45.2.11. However, the IDI team was concerned with the guidance, design experience and training available in the Harris Plant Engineering Section (HPES) electrical group. In this regard, programs have been initiated to increase the overall capability of the group by developing guidelines for areas within the HPES electrical scope of work such as, conduit seals, cable sizing, cable and conduit list interface, containment electrical penetrations, communications, protective relay settings and coordination, load monitoring, voltage and short circuit considerations, heat tracing, FCR resolution, tray covers and radiation monitoring isolated ground bus. Additional guidelines will also be developed as is necessary and as the electrical unit scope expands. Upon completion of the discipline guidelines, personnel within the HPES electrical section will be trained as to purpose and application of these guidelines.

Furthennore, by utilizing to the fullest extent the information contained in the Design Basis Documents as well as project criteria and calculations, the capability and information available to the HPES electrical unit will be greatly enhanced to support project requirements. This concern is addressed further in Section 7.2.

In addition, to provide technical guidance and supervision as well as to enhance the overall comunication and transition of work between the Ebasco and HPES electrical sections, the Ebasco electrical supervisor has been assigned to the site on a full time basis and assumed the duties and responsibilities of Electrical Unit Manager for the HPES electrical unit.

5.3 STANDBY EMERGENCY POWER SUPPLY The IDI team reviewed the applicable design documents to ensure that the diesel generators have sufficient capacity, capability and reliability to perform their safety function.

The IDI team found the design to be adequate but noted that non-safety loads were not shed on LOCA alone, but only on LOCA plus a loss of offsite power (LOOP) (US.3-1). This had been identified by the project prior to the start of the IDI. Safety Analysis Report Change Notice F-162 had been generated to modify the condition under which the 6.9KV non-safety buses would be isolated from their associated 6.9KY safety buses. This design modification, which is more fully discussed in the response to Unresolved Item US.3-1, was not completed at the time of the IDI.

29

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

5.4 MOTOR CONTROL CENTERS The IDI team evaluated.the design and application of the safety related motor control centers and expressed several concerns.

With regard to the concerns identified in this section, we agree

.that an additional procedure and verification as stated in the report is required for selection of thermal overloads for motor

operated valves. Corrective actions as described in the specific response to deficiencies D5.4-2 & D5.4-3 will assure that the proper

. overload heater element will be selected for the respective motor operated valve.

Although we agree with the IDI team that adequate overcurrent i protection is required for station service transformers, power

! centers and feeder breakers to motor control centers, we. disagree 4

that the selection of the overcurrent relay settings should be based

, upon consideration of connected load. Rather, this selection should 3 be based on consideration of the maximum anticipated steady state r i

current with sufficient margin for transient overloads, such as the.  :

starting of a large motor or overloads due to voltage variation. In any case, the continuous rating of the station service transformer, power center or motor control center bus will not be exceeded.

Furthermore, in response to the concerns identified in deficiencies D5.4-1 and D5.4-4, a guideline for protective relay settings and ,

coordination will be developed and safety related breaker

overcurrent relay set points will be re-analyzed in accordance with

the criteria. set forth in the guideline.

• i '

The IDI team identified a concern with the procurement of electrical boxes as non-safety related. The Shearon Harris Project will verify that the boxes satisfy applicable design requirements (US.4-5).

i

' I 5.5 DIRECT CURRENT SYSTEM The IDI team reviewed the method used to size and specify the Class l lE batteries and battery chargers and evaluated the overall dc  ;

. system.

y  ;

l The results of the review showed that the battery and battery  !

t chargers were adequately sized and the associated calculations were I

clear and complete with references attached. There were, however, l specific inconsistencies identified in the battery sizing

! calculation, but they did not affect the size of the battery. These items are more specifically addressed in the response to Deficiency

5.5-1. The IDI team also determined that the battery discharge "

voltage profile was not completely conservative (Deficiency 5.5-3).

When applying this battery discharge voltage profile to determine

the allowable voltage drop in switchgear control circuits, the IDI team identified an assumption that did not result in the most conservative condition (Deficiency 5.5-4).

30

5.5 DIRECT CURRENT SYSTEM (Cont'd)

New calculations were recently developed as part of a comprehensive de system update that was conducted by the Ebasco electrical group.

In many cases the load changes were significant enough to require the preparation of a new calculation. These new calculations address the above concerns identified by the IDI team.

The IDI team also showed a concern that part of the de system design was not consistent with the statements in the FSAR. The IDI team was concerned that adequate documentation was not available to support an FSAR statement that all de equipment was qualified to operate at 140V dc (Deficiency 5.5-2). Regarding D5.5-2, several alternatives are listed in the detailed response that will provide assurance that the upper voltage limit of the de system is within the tolerances of the equipment.

The IDI team also identified that surveillance of the battery charger current was not available on the main control board as stated in the FSAR (Deficiency 5.5-5). In the response to 05.5-5, an explanation is provided why the intended design (to provide local indication for the battery charger output current) is at variance with the FSAR.

An additional observation related to dc undervoltage alarms is addressed in detail in the response to 05.5-6.

5.6 CONTAINMENT ELECTRICAL PENETRATIONS The IDI team evaluated the containment electrical penetration specification, reviewed the available fault current and examined the interdiscipline input to the specification.

The IDI team found the specification to be correct but expressed a concern with respect to backup protection for the reactor coolant pump power feeder. In response to this concern, the subject circuit has been redesigned to incorporate two sets of overcurrent relays each connected to an independent source of 125Y de power. A more detailed discussion of the circuit design can be found in the response to Deficiency 5.6-1.

5.7 MOTOR ELECTRICAL PROTECTION The IDI team reviewed and evaluated the motor starting and running conditions under different plant modes of operation and evaluated the type of motor protection provided. The IDI team found that protection for the 460 volt motors was set based on assumed data (D5.7-1) and a random error was identified whereby the 480 volt bus undervoltage relay setting did not agree with the relay setting calculation. (DS.7-2) The IDI team also observed that the motor checkout procedure did not require monitoring nor recording of the motor's starting parameters. (05.7-3) 31

m - . _._ _ _ _ _ _ _ _ _ . _ - _ . . _ . . _ _ _ _ _ _ . _ . - _ . . - _ _ . . _

5.7 MOTOR ELECTRICAL PROTECTION In response to these concerns, the relay calculations for each of the 460 volt Class 1E motors will be updated and verified using motor specific starting times and safe stall times. The relay calculations will be incorporated in a Standard Relay Calculations and Data Sheet as required by the Harris Plant Engineering Section guidelines for. relay protection, which is currently being developed. Accordingly, any missing assumptions will be identified and verified. Concerning the 480 volt Class lE bus undervoltage relay setting, the latest voltage study, " Adequacy of Station Electric Distribution System Voltage" (Feb,1985), has determined

that the relay dropout settings of 100 volt for two of the 480 volt Class 1E power centers were acceptable and the relay dropout settings for the remaining two 480 volt Class 1E power centers will be revised to 110 volt. The drawings for the subject settings will i be revised to reflect the correct bus undervoltage relay settings i and the time delay settings on the undervoltage relays of all Class

, 1E power center buses will be reviewed to ensure avoiding nuisance ,

i. relay operation under transient undervoltage conditions. The Harris l

. Plant Engineering Section will develop a design guideline and j procedures for relay protection, j

i The Startup Organization will revise Procedure 1/2-9000-E-05 to t i require the bus voltage to be recorded when measuring the motor 3 running current under both the loaded and unloaded conditions. .In ,

addition, the appropriate process will be established to require the startup organization, upon request from the Harris Plant Engineering Section, to perfom additional tests on large 460 volt and 6.6KV motors for which specific data cannot be obtained from the vendor.

5.8 CABLE DESIGN AND ANALYSIS l

c The IDI team reviewed cable sizes to ensure cables were properly sized for both nomal and overload conditions. Power and control l cables were reviewed for both ampacity and voltage drop i considerations. The IDI team identified certain cases where cables

, were not adequately sized for the worst case voltage conditions.

l Deficiency 5.8-1 stated that when the de system is at minimum

voltage, the power cable feeding the de motor operated valves is not

! adequate. Corrective action has been taken by revising the cable

. size.

The basis of Deficiency 5.8-2 was the possibility of simultaneous operation of multiple relays in the auxiliary relay panels. Under worst case conditions, several relays can operate concurrently, i 1

which might result in a high inrush current (approximately 125A) through the cable feeding the panel. Due to this high inrush

current, the voltage at the relay coils may not be adequate to i operate the relay. The Ebasco electrical group has just completed l an evaluation of all ac and de control loops. The objective of this

effort was to ensure that all associated cables were properly I sized. As a result of this evaluation, in conjunction with D5.8-2, the subject cable size has been revised.

i l 32

5.8 CABLE DESIGN AND ANALYSIS (Cont'd)

Deficiency DS.8-3 involved discrepant data for the Reactor Coolant Pump brake horsepower. Westinghouse had submitted a letter with a brake horsepower value that was inconsistent with a Westinghouse technical manual. The electrical auxiliary system computer program did not use the more conservative value. Westinghouse was contacted to clarify the discrepancy and the electrical auxiliary system programs were re-run using the corrected value. The results showed the electrical distribution system voltages to be acceptable.

5.9 ELEMENTARY / INTERCONNECTION DIAGRAM REVIEW The IDI team evaluated the process of circuit development from the i system concept through the generation of control wiring diagrams and power distribution and motor data sheets.

The IDI team found that Ebasco correctly implemented Westinghouse design for the chemical and volume control system, but cited two examples where the independence of safety related circuits was not <

! maintained.

With respect to the breaker protection for the reactor coolant pumps, a design modification has been initiated to provide two e overcurrent relays each connected to an independent source of 125V de control power. (See the response to Deficiency D5.6-1 for 3

additional details.)

The second concern dealt with non-independence of the reactor coolant pump breaker status inputs to the Reactor Vessel Level t Instrumentation System (RVLIS). We consider the application of i single failure criteria to non-safety RVLIS inputs to be

! inappropriate and beyond present design connitments. A more i detailed discussion of this concern can be found in the response to i Deficiency D5.9-1.

The IDI team has cited the two concerns discussed above as evidence of ineffective design interface between Ebasco's Electrical and

l. Instrumentation & Control groups. Since one of the concerns is not ,

considered to be a design deficiency, and in light of the large

l. number of control wiring diagrams reviewed, one example of a design

oversight is not indicative of faulty design interface.

5.10 DESIGN ACTIVITY / DESIGN CHANGE CONTROL The IDI team evaluated the adequacy and degree of conformance to project procedures governing original design, design changes, and drawing changes. The IDI team questioned the cumulative effect of several minor changes and concluded that a more conscientious

!. adherence to the project's commitments to the requirements of ANSI N45.2.11 in the areas of design and design control are required within the Harris Plant Engineering Section electrical discpline.

(DS.10-1) 1 3

l

'f l 33  :

r 5.10 DESIGN ACTIVITY / DESIGN CHANGE CONTROL (Cont'd)

Concerns relating to " minor" design changes are addressed in Chapter 7.4 of this report. Regarding design change control within the Harris Plant Engineering Section electrical unit, appropriate instructions are in place that apply to the entire Harris Plant Engineering section. To enhance application of these instructions, the electrical unit has identified the need for and will issue design guidelines for its design activities. The specific concern of incorrect cable sizing of 2 cables (D5.10-1) resulted from a

" dummy" size being used to ensure routing only and subsequently being inadvertently released for installation. The two cables have been changed and other work in this area has been reviewed and found adequate. A design guide (as mentioned above) to place additional control and verification requirements on cables sized by HPES has been issued.

5.11 ELECTRICAL SEPARATION The IDI team reviewed the design provisions for electrical separation and physical independence of Class IE circuits. Based on this review, the team expressed a concern over the use of a walkdown to identify violations of electrical separation criteria.

The IDI team acknowledged our plans to identify cable tray separation violations by walkdown, design cable tray covers to meet separation requirements, and analyze the lengths of cover that exceed the standard limit. We feel that the programs are well controlled, appropriate procedures are in place, and adequate expertise is available to assure electrical separation and physical independence of Class lE circuits.

34

6. INSTRUMENTATION AND CONTROL 6.1 DESIGN INFORMATION AND PERSONNEL The IDI team reviewed design information provided by Westinghouse and other equipment suppliers. Technical information exchanges between Carolina Power & Light and Ebasco, with emphasis on the selection of a given technical approach for several design modifications of interest to the IDI team, were examined.

Design control procedures and instructions prepared by Carolina

' Power & Light were reviewed, as well as a number of Shearon Harris Project design control procedures and instructions prepared by Ebasco. This material included the Department Manual, the Ebasco SHNPP Manual of Procedures, and an Ebasco instrumentation and control design guide.

The technical qualifications and relevant design experiences of approximately twenty Ebasco and ten Carolina Power & Light personnel involved in these design activities were reviewed by the IDI team.

Internal technical audit reports of the emergency service water and the auxiliary feedwater systems, prepared by Carolina Power & Light and by Ebasco respectively, were reviewed. Technical design verification activities performed by both Carolina Power & Light and Ebasco were also examined.

The IDI team concluded that the transfer of information between

.Westinghouse and Ebasco appears to be generally well controlled.

The discrepancies noted by the IDI team are either minor or easily corrected.

The relative absence of documentation inconsistencies, errors, and omissions.in instrumentation and control design documents provided evidence of sufficient attention to detail in the checking, design review, and design verification processes.

The Instrument List contains over 26,000 instruments and control devices each having more than twenty pieces of information. The Instrument List provides a comprehensive compilation that aids a user in quickly locating pertinent sources of information, such as applicable specification, vendor drawings, Control Wiring Diagrams and installation details. The IDI team identified approximately 25 cases of missing or incorrect inputs after reviewing approximately 10,000 entries. Although valid, these discrepancies were minor in nature and did not affect design or installation. This information was inadvertently omitted or incorrectly transferred from input data sheets.

Since the Instrument List is a controlled drawing, any incos,sistencies or errors will be noted and changed by a controlled process (Field Change Request or Design Change Notice).

Additionally, as part of the start-up procedure preparation and pre-operational check-out, the site is screening the Instrument List and issuing FCRs as required (D6.1-1).

35

6.1 DESIGN INFORMATION AND PERSONNEL (Cont'd)

The IDI team noted discrepancies between the FSAR and the Instrument List. Since the Instrument List is an integral part of the design process, it sometimes leads the FSAR with respect to detailed system design. System changes occur as part of the nonnal process of finalizing design. When necessary, the areas of the FSAR that need to be revised are identified and scheduled for revision (D6.1-2).

The IDI team noted "significant inadequacies and incompleteness of design control procedures" used by the Instrumentation and Control discipline at the project site. The IDI team concluded, however, that it is not a "significant concern" considering the limited amount of safety related design work performed at the site. We feel that adequate design control procedures as required by ANSI N45.2.ll, are in place. The procedures are supplemented by instructions and technical guidelines as needed. Based on the specific deficiency mentioned in this section, however, we have written and issued drafting room guides which contain drafting requirements and guidance for instrumentation and controls. In addition, we have issued drafting guides for the other disciplines.

We feel that these guides, in addition to our design change control procedures and our technical guidelines, provide adequate control of the design process on site (D6.1-10).

6.2 PROTECTION SYSTEM AND PROTECTIVE ACTION SYSTEM The IDI team examined the Ebasco design for automatic initiation of the main diesel generators from either a safety injection signal or a loss of offsite power signal. Inspection emphasis was placed on periodic test design provisions included in these circuits to ensure high operational availability and reliability of the Class lE power system. The IDI team also reviewed Ebasco's design review and design verification activities relative to the design of these instrumentation and control circuits.

2 Based on the documents reviewed, the IDI team concluded that a controlled and effective design process had been used by Ebasco for the balance of plant protection system and protective action system.

6.3 INSTRUMENTATION AND PANEL SEPARATION The IDI team reviewed the physical separation and electrical isolation criteria and examined the installation of instrument impulse lines, wiring and cables, both internal and external to '

instrumentation and control panels and racks.

36

6.3 ' INSTRUMENTATION AND PANEL SEPARATION (Cont'd)

The IDI team found the ' separation criteria established by Westinghouse and Ebasco .for electrical components and instrumentation tubing to be consistent with applicable industry standards and. regulatory requirements. However, they expressed a i

concern related to the use of the following terms; train, safety  !

train and nuclear safety train (06.3-1). Since the design criteria ,

-stated in drawing CAR-2166- B-060 is specific with regard to the +

installation of safety and non-safety equipment, and no violations of this criteria were identified, we judge the existing definitions to be adequate.  ;

Our position, with respect to physical separation in control panels, is consistent with its FSAR commitments to Regulatory Guide 1.75, IEEE-279 and IEEE-384. Barriers between Class IE wiring are permissible to maintain separation and equipment manufacturers are required to provide separation by means of barriers, wireways, and/or conduits. Control panels are patterned after the basic ,

design of the Main Control Board supplied by Westinghouse (the NSSS

-supplier), which utilizes _ flexible conduit to maintain separation.

An evaluation by Westinghouse considering the combination of wire sizing per approved codes,- fuse protection, and insulation type prec1: !es any possibility of a fire being generated or propagated

.between trains or between train and non-train. A numeric analysis >

was not deemed necessary for the above reasons. Nevertheless, an engineering analysis has been perfomed to provide additional assurance that the existing design fulfills the Shearon Harris Nuclear Project's FSAR commitments (D6.3-2).

Regarding the concerns relating to the Shearon Harris Nuclear Project's " minor" designation of Field Change Requests, which are  :

not subject to design verification,- (U6.3-3) the procedures used by -

the Harris Plant Engineering Section are written assuming that design verification applies unless the changes are " minor". " Minor" changes are defined as those items which do not change the safety i review, design calculations nor the design basis. We have i thoroughly reviewed the specific field change in question and concluded that the " minor change" designation applies. ' The change ,

is in compliance with regulatory criteria and our licensing 1 commitments because the change approved was consistent with the .

project's design bases. Also, numerous field change requests  !

involving tubing re-routes have been reviewed by Ebasco for '

! unacceptable consequences, and none have been found. Further l' discussion and details of our conclusion are contained in Chapter  ;

l 7.4 and Appendix A of this response, i

l l

l L

37

6.4. ENVIRONMENTAL ~AND SEISMIC QUALIFICATION The IDI team reviewed a sample of Ebasco balance of plant instrumentation, rack, and panel procurement specifications to verify implementation of applicable FSAR environmental and seismic qualification requirements. A number of environmental and seismic qualification test reports were also reviewed at Ebasco and the Shearon Harris site. Based on this review, the IDI team concluded the equipment should meet its intended function.

The IDI team found that the containment hydrogen analyzer system does not conform to its specification, that the vendor has not taken exception to the specification and that Ebasco had not noted the deviation (D6.4-2).

The specification required that the analyzer be explosion proof where the equipment could be in contact with flammable concentrations of hydrogen. The vendor has provided calculations that demonstrate the analyzer does not develop a flammable mixture of hydrogen and that the analyzer part: in contact with hydrogen do not provide an ignition source. As suci., the intent of the specification is satisfied. The applicable sections of the specification have been modified to reflect these considerations.

6.5 BALANCE OF PLANT INSTRUMENT SETPOINT CALCULATIONS Ebasco has design responsibility for balance of plant safety related instrumentation in the Class lE power systems and various engineered safety feature supporting systems. A number of these instrument loops either actuate or provide automatic control of safety related systems, while others provide Class 1E alarms in support of actions taken by the plant operator. Regulatory Guide 1.105 provides guidance regarding the documentation of assumptions and the minimum margin used in selecting the setpoints used to initiate automatic protective actions and alams in a manner similar to the IEEE Standards 279-1971 and 603-1980.

The Shearon Harris Project utilizes this explicit setpoint methodology only for Plant Protection System input setpoints (ESFAS and RPS). The Westinghouse Precautions, Limitations, and Setpoints (PLS) Document provides the values for the great majority of these items. Westinghouse has also provided in the setpoint methodology document all the calculated values that were used for detemining the ESFAS/RPS setpoints. Ebasco's responsibility is limited to the explicit setpoints for Containment Radiation and Chlorine Level for the Containment Ventilation and Control Room Ventilation Isolation l signals respectively. This is considered to be consistent with l position Cl of Reg Guide 1.105.

l 38 L

i

)

6.5 BALANCE OF PLANT INSTRUMENT SETPOINT CALCULATIONS (Cont'd) i All other setpoints (non-ESFAS/RPS) are considered implicit and as such do not require the same degree of documentation or calculation

[ basis. Therefore, the values listed in the Setpoint Documents which -

reflect fluid processes have sufficient documentation (U6.5-1).

p

Design basis documentation for the selection of safety related setpoint values and tolerances is satisfactory for the nuclear steam

. supply scope and for instrument loops in the balance of plant scope. Each of the areas noted by the IDI team will be reviewed

- and, where necessary, additional documentation to support the issued 4

setpoints will be provided. In the event there are changes to existing setpoints, a design change notice will be initiated.

! Further discussion and detafis are contained in Appendix A of this  !

J response. -

j i 6.6 CONTROL SYSTEMS

{ l j The IDI team reviewed Westinghouse and Ebasco design practices for i the chemical volume and control system, as well as balance of plant [

HVAC, containment spray, Class IE power system sequencer, and the i

i normal and emergency service water systems. This review also j included fire door monitoring and integrated leak rate test ,

1 instrumentation designs developed by Carolina Power & Light, and '

. design details for one instrument rack fabricated at the Harris plant site. Emphasis was placed on the identification of potential  :

adverse interactions between these individual control systems and  ;

the plant safety systems, i I

i Based on this review the IDI team concluded that adequate control

had been placed on the design of control systems by all three design I organizations. A minor inconsistency on one Westinghouse interlock l sheet was identified which did not affect the corresponding Ebasco

! control wiring diagram. This is discussed in the response to +

Observation 06.1-5.  :

i j 6.7 INDICATION AND ANNUNCIATION The IDI team reviewed the instrumentation provided for indication and annunciation of the chemical and volume control system. The FSAR, control wiring diagrams, control panel drawings and .

specifications were examined to detemine if all required l instrumentation and alarms had been provided. Implementation of  !

human factors considerations was also examined on the auxiliary  !

equipment and control panels.  !

l .  !

! The IDI team concluded that the design of indication and i annunciation instrumentation appeared to be controlled, that Ebasco j control wiring diagrams clearly identified the required indications I 1 and that the control panels provided indication and annunciation i alarms as required by the FSAR. The IDI team further concluded that l human factors considerations appear to have been properly  ;

{ implemented.

39

6.7 INDICATION AND ANNUNCIATION (Cont'd)

A concern was expressed with regard to the accuracy of instrument ranges given-in the FSAR. The IDI team has correctly noted the FSAR is not an original source of design information but reflects' input obtained from controlled design documents. As such, the FSAR will sometimes lag these documents as design modifications are made.

Since the FSAR is not used for implementation of detailed design this is not of concern. The FSAR, however,.will be updated to incorporate the latest design input. For further details, see the response to Deficiency D6.1-2. ,

6.8 BALANCE OF PLANT DESIGN AND FIELD CHANGES The IDI team examined Ebasco procedures for design change notifications and Carolina Power & Light procedures for field change requests used to modify the balance of plant design for instrumentation and controls.

The IDI team found the Ebasco design change procedure to be adequate but expressed concern with regard to review of design changes by Ebasco's licensing group. Previously, licensing would review design changes in the Design Change Notice (DCN) or when these changes had been incorporated into the applicable drawing. The Ebasco licensing group no longer reviews drawings since they have been transferred to the Harris Plant site. As outlined in the response to observation 06.8-3, the Design Change Notice procedure has been revised to require licensing review of Design Change Notices affecting safety related drawings prior to their final release.

The Shearon Harris Nuclear Project's field change request procedures provide definition of " minor change", with all other changes assumed ,

to require design verification. This subject is further discussed '

l in Chapter 7.4 and Appendix A of this response. Procedural changes now require that voided field change requests obtain the same level of approval as for original issue. Further details are provided in Appendix A of this response.  ;

The electrical / instrumentation and control drafting guide has been i

expanded to include guidance for the instrumentation and control discipline and has been issued formally. We feel that, although the l

guide was not previously issued, appropriate procedures were in r place to require a detailed review and approval of drawing revisions l by experienced engineers and supervisors completely familiar with the drawings.

Wooden supports have been replaced with metal on the safety battery l room service sink (D6.8-5). Material substitutions (both safety and NNS) are required by procedure to be documented by an approved 1 change document. Our review of field change requests written against the specification for the subject sink indicates that this is an isolated case of material substitution without engineering review and approval.

i 4

40 ,

- - - - . - . . . . . - _ - - - - . - - - - -_--- a

6.8 BALANCE OF PLANT DESIGN AND FIELD CHANGES (Cont'd)

Unverified calculations which supported an FSAR amendment have been completed, verified and found to be conservative. We have revised the Harris Plant Engineering Section instructions to require that FSAR changes be supported by a design document (and thus a design verification) as appropriate (D6.8-6).

The above actions, in addition to responses to the specific items in Appendix A of this response, serve to strengthen the design procedures not only in the instrumentation and control discipline but the entire Harris Plant Engineering Section.

See Appendix A for detailed responses to items 06.8-1, 06.8-2 and 06.8-4.

i l

l l

41

7. DESIGN CONTROL ASPECTS RELATED TO MORE THAN ONE DISCIPLINE 7.1 COMPUTER SOFTWARE QUALITY ASSURANCE The IDI teain evaluated the use of computer codes in the design and analysis of safety related structures, systems and components to ensure that NRC requirements for design control were met. The IDI team expressed a concern that Ebasco's Administrative Procedure No.

A-30 did not adequately establish a quality assurance program for computer codes, especially with respect to the documentation of errors / bugs.

In this regard Ebasco Procedure A-30, " Computer Based Safety Related Programs - Development and Control", has undergone a revision which

.more fully defines a system for reporting and evaluation of computer

' program errors. In addition, Ebasco has prepared Procedure A-37,

" Control of Operating Systems, Software Problems Affecting Computer Based Safety Related Programs," to define specifically a system for reporting and evaluation of computer system errors. All applicable Ebasco personnel will receive the appropriate training.

7.2 DESIGN CAPABILITY OF THE HARRIS PLANT ENGINEERING SECTION We believe the questions and concerns raised by the IDI team about the design capability of the Harris Plant Engineering Section (HPES) did not reflect the benefit of infomation from on-site or off-site CPAL design organizations or the benefit of discussions with the engineering staff and management of these organizations.

We agree that there is a need to ensure that appropriate skills and technical guidelines exist within HPES.- We have had a controlled

! program to accomplish the plant design transfer from Ebasco to HPES i and previous HPES design activities have been performed in j accordance with approved procedures compliant with ANSI requirements

and our Corporate Quality Assurance Program. These activities have i been the subject of Corporate Quality Assurance audits. As a result l of our controlled program, and the IDI team's concern, we have j augmented our program to strengthen the overall capability of our

organization with emphasis in the electrical discipline. The l following actions have been taken.

I o HPES has been focusing on these areas of needed emphasis since its inception. An effort in mid-1984 identified needed instructions and technical guides. A result was a listing of Design Technical Guidelines which were needed.

HPES is expediting this process and completing needed j guidelines on an accelerated schedule.

l o A plant design basis program was established in mid-1983

! as a joint CP&L-Ebasco effort. Ebasco is producing Design i Basis Documents (DBD's) for each plant system, structure,

! and for generic equipment. Over one hundred D80's have been identified. Nearly all have been drafted and l reviewed with CPAL comments being generated by both j

42 L

~

7.2_ DESIGN CAPABILITY OF THE-HARRIS PLANT ENGINEERING SECTION (Cont'd) o Engineering and Operations. . About half of the DBD's have been approved and, although not yet invoked, are under controlled distribution on site. We have established a schedule.for completing the DBD's.which when completed will form a substantial part of our design control process as they are invoked.

o In late 1984, CP8L requested Ebasco's assistance in mapping out a plant completion strategy for preparing HPES to take on its role as the design organization. Ebasco responded in December with a comprehensive assessment covering staffing, control, procedures, organization, and training. HPES is making maximum use of Ebasco's help and expertise. We have established a pattern of this in the work which we have transferred to the site to date. We utilize experienced Ebasco personnel on site, conduct training sessions, and establish transition periods for each type of design work transferred to ensure that product quality is maintained as the responsibility is transferred to HPES.

o An Engineering Transition Program has been established on the Harris Project to ensure that the transfer of original design engineering responsibility from Ebasco to HPES occurs in an orderly, controlled fashion. The team consists of several senior level Ebasco engineers under the direction of a full-time transition program manager.

The members were selected from each of the major engineering disciplines where original design responsibility is being or will be transferred to HPES.

They assist HPES to identify, plan and implement the actions necessary to prepare HPES to accept responsibility for each 1;ype of design work transferred.

o A program has been established to bring selected supervisory Ebasco personnel to the site to strengthen key design areas. They assist in design guide preparation and organization development. One of these personnel is supervising the HPES Electrical discipline. Chapter 5.2 of this response has provided additional information regarding our plans to strengthen the Electrical discipline.

o We reorganized our Drafting and Computer Graphics from a unit whose charter was solely drafting, (i.e.,

incorporating approved design changes into original documents) to design subunits integrated functionally and administratively into each engineering discipline. This provides more engineering expertise, direction, and control of changes to design documents and of engineering anc design of future modifications.

43

"e 7.2 DESIGN CAPABILITY OF THE HARRIS PLANT ENGINEERING SECTION (Cont'd)

In addition to the improved capabilities that are being derived from the above actions, HPES has the integrated design capabilities of CP&L available for support. CP&L has demonstrated its integrated (corporate and site specific) design capabilities and design performance at its three operating nuclear units (H.B. Robinson 2 ,

and Brunswick 1 and 2) and will apply the same program in support of '

Harris in the long term, i 7.3 DESIGN VERIFICATION PROCESS The Engineering Procedures developed by Ebasco and in use on the

, Shearon Harris Project provide the framework for the preparation,  :

i ' checking and design verification of calculations. These procedures, supplemented by discipline engineering and design guides, meet the j' requirements of ANSI N45.2 and Regulatory Guide 1.64 and have been the subject of audits by the NRC, many of Ebasco's clients, and

. third party reviewers. Ebasco considers the calculations performed on the Shearon Harris Project to be well controlled and reflective j of a sound plant design.

The Integrated Design Inspection (IDI) Team evaluated calculations i in the mechanical, electrical, civil, fire protection, stress  !

5 analysis, supports / restraints, radwaste and applied physics groups.

! Based on concerns related to calculations in four of the

disciplines, the IDI team has stated that the design verification

program was not consistently effective and indicates a lack of management attention. After a thorough review of the findings identified in this report we conclude that this assertion is not <

, justified. As previously discussed in sections 2 through 6, we do 4 not consider many of the fi dings, cited by the IDI team as evidence 1

of inadequate design verification, to be deficient. Further, the IDI team has recognized a well controlled effort in the fire protection, supports / restraints, instrumentation, electrical,and radweste groups. Nevertheless, in recognition of the concerns <

i identified during the inspection, comprehensive activities have been

, initiated on the project level. These activities are briefly i described below:

1 o During and after the inspection, appropriate actions were i initiated on all calculations in which the IDI team identified l concerns. Work completed to date has, in all cases, verified the adequacy of the existing design. ,

I o The potential for the concerns identified by the IDI team to be generic has been evaluated in our review of additional calculations. Where appropriate, a systematic review of 1 existing calculations has been conducted, supplementary .

calculations have been performed and additional documentation i of data used in calculations has been provided. ,

i 44  !

7.3 DESIGN VERIFICATION PROCESS (Cont'd) o The calculation procedures and guidelines in use on the project have been reviewed to determine if inadequacies therein could have resulted in the concerns expressed by the IDI team. These procedures and guidelines have been found to contain the elements necessary for a controlled calculation process. The conditions identified by the IDI team have resulted from differences in the application of these procedures. Project personnel have been familiarized with concerns raised during the IDI to prevent repetition of the items identified. In addition, all senior engineering managers are ccgnizant of these concerns and have instructed project level supervisors to closely review future work in light of the IDI report.

The concerns identified in each discipline are unique and do not have a commonality that would require or permit generically designed corrective programs to be effective. As such, the corrective actions . initiated have been tailored to address these problems on a discipline basis.- Below is a brief description of the concerns expressed by the IDI team and the corrective actions that have been undertaken.

MECHANICAL In the mechanical discipline approximately 22 calculations were evaluated. The IDI team questioned the methodology utilized in two calculations concerning vortex prevention (D2.3-2) and eductor modelling (D2.3-5). We concur that the eductor. calculation contained modelling errors, which have been corrected. With respect to vortex prevention, existing analytical techniques available in the industry have been found to be incapable of accurately predicting containment sump behavior. As such, reliance is placed on model testing which was considered to be an integral part of the verification process.

The IDI team stated that the design of the containment sump and the chemical addition portion of the containment spray system was deficient. We do not concur with this assessment. As detailed in the responses to 02.3-1 and D2.3-6, both designs satisfy applicable regulatory criteria and are capable of performing their intended function.

The small number of concerns raised in relation to the number of calculations reviewed, and the nature of these concerns are judwd to be indicative of a well controlled design process. Neverthe' ens, the following activities have been initiated in the mechanical discipline:

45

7.3 DESIGN VERIFICATION PROCESS (Cont'd)

MECHANICAL- (Cont'd) o All applicable calculations will be reviewed for vortex prevention considerations.

o The Na0H e'ductor calculation has been. revised. No design change has resulted from the deficiency identified, o Calculations have been performed to document minimum post-accident containment water level required to document confomance with Regulatory Guide 1.82.

APPLIED PHYSICS The IDI team evaluated five shielding calculations and identified several discrepancies of varying significance. One of the calculations cited was performed to detemine the radiation dose from the Volume Control Tank following an accident. The resultant dose was found to be too high to permit post accident use of this tank and as such this tank is isolated upon an accident. Because the calculation demonstrated a dose level that was too high to permit post accident access to the tank area, the calculation did not need to be rigorously checked or kept current.

The IDI team also questioned the validity of certain techniques and the rounded values of several variables utilized -in calculations.

Although in some cases the techniques or values used were nominally non-conservative, they would not in themselves invalidate the results and are consistent with the accuracy attainable in shielding calculations. The IDI team further identified a significant underestimation of a gamma dose that was carried through two calculations. This error resulted from a misinterpretation of data supplied by Westinghouse. Overall, this underestimation was compensated by other conservatisms in the calculation yielding results which are entirely consistent with those of similar plants.

The effect of this compensation was to hide from the verifier of the calculation an incorrect input that would otherwise have been identified. While the IDI team questioned the lack of apparent

! conservatism in several calculations, the errors identified are i isolated in nature and are not indicative of a progrannatic weakness.

Listed below are the activities that were initiated in applied physics as a result of the IDI:

o The volume control tank shielding calculation has been revised to correct deficiencies identified during the IDI. This work i has verified the adequacy of the existing design, i

o Radiation dose calculations utilized for equipment

! qualification purposes have been revised to address all IDI concerns. Use of more refined calculation techniques has confirmed the conservatism of existing radiation dose terms.

b

7.3 DESIGN VERIFICATION PROCESS (Cont'd)

APPLIED PHYSICS (Cont'd) o Other calculations in the applied physics group have been reviewed for the concerns raised during the IDI. The deficiencies noted have been found not to be generic.

o The shielding group supervisor has instituted a program to enhance the degree of checking and verification of future calculations. The deficiencies identified during the IDI have resulted in the issuance of guidelines and increased attention by engineering management designed to prevent recurrence of these concerns.

STRESS ANALYSIS Sixteen calculations were evaluated for correctness of design and for proper utilization of Westinghouse design input. The IDI team found the stress analysis program to be particularly well controlled from the generation of flow diagrams and piping drawings to the preparation of isometrics and the coding of calculations. Minor discrepancies were found in several calculations, some of which required reanalysis. The IDI team noted that these discrepancies were predominantly in the older vintage calculations and were not

-manifest in calculations performed in the last few years. A number of the discrepancies identified were very minor and by inspection determined to be insignificant. It is probable that some or all of these discrepancies were identified by the original checker but because of less rigorous documentation requirements for early calculations, the evaluations of these discrepancies were not specifically noted. Given the hundreds of inputs required in each stress calculation, the number of minor discrepancies identified are indicative of a well controlled design process.

Listed below are the activities that were initiated in stress analysis as a result of the IDI:

o Discrepancies identified during the inspection have been evaluated and corrected and have been found to have no effect on existing design, o The stress analysis group supervisor has conducted a seminar to familiarize discipline personnel with the concerns identified during the IDI. All personnel have been instructed to review carefully calculations that are revised during design finalization for these concerns. In addition, personnel involved in the stress calculation transfer program carefully review these calculations, prior to turnover to CP&L, for the concerns identified during the IDI.

47

7.3 DESIGN VERIFICATION PROCESS (Cont'd)

CIVIL-STRUCTURAL The concerns raised in the civil-structural discipline generally fall into two categories, neither of which jeopardize the validity of the calculations or the resultant design. The first category can be best described as concerns of a " housekeeping" nature. The examples cited by the IDI team include lack of/or incorrect references and format discrepancies which made the calculations difficult to follow. Problems of this nature occur in calculations that have undergone numerous revisions using different personnel.

The problems cited are not design related and are not indicative of a weakness in the project design verification process.

The second category relates to the lack of sufficient documentation of engineedng assumptions and judgements utilized in various analyse. The level of engineering judgement that requires documentation and the degree of documentation required is subjective. While several examples of insufficient documentation were cited, they are not evidence of a faulty design verification process. In all cases completed to date, where additional analyses were requested to justify engineering judgements, these judgements were confirmed and the existing design validated.

The activities initiated in the civil-structural discipline are given below:

o Civil engineering has addressed and corrected all calculation reference / format discrepancies identified by the IDI team during the audit.

o During the inspection, supplementary analyses were performed at the request of the IDI team to provide justification for analytical techniques used. In all cases, the techniques utilized have been demonstrated to be adequate.

o Programs have been initiated to provide more complete documentation of masonry block walls, HVAC duct and cable tray supports.

In summary, while we do not minimize the concerns identified by the IDI team, these concerns are not evidence of a systematic weakness in design verification. The small number of concerns identified in relation to the significant amount of documentation reviewed, coupled with the fact that in all cases subsequent work has demonstrated the adequacy of existing design, demonstrate that the design verification process on the Harris project is effective.

Nevertheless, the activities described above have been initiated to address the specific IDI concerns and further strengthen design verification.

48

]

s p3 fQ3 - J7 :[s ,f w

'a , j'y 'b- s <

cq  % - _,; ? .

,3  : :m 7.4, MINOR DESIGN CHARGC' DESIGNATION

~%  :, .-

Carolina Power & Light Company (CP&L) d5partmentat procedures and 4 section' instructions contain design verification instructions and the' conditions.uhder which they apply. The procedures are written .

dssuming that ' design verification applies unless an item is a " minor chaage". "Mindr changss" are those items which do not involve - ,

changing the safety review, design calculations or design basis. J.

j" Major change"gis not defined since design verification is requirei '

.for everything that is not a minor chanse. As such, many field -

change requests are processed as ."minot changes"; however, some +

' field change reqdsts undergo design verification. The IDI team's implicatich that thd established procedure is routinely to Idintify-changes as ," minor" is not justified. .

y y 3 ,

Af t.dr an extensive review of the specific field change request,that i was questidned with regard to " minor" vsrsus " major", both CPOand -

Ebasco agrde tha{ design verification.is not required. The design basis wasdot changed and the subject field change request 'is in accordande with regulatory critchia and our FSARicommitments. 4 Additforal specific information relative to this. concern is p'rovided ~.

in our response to Unresolved Item US3-3 in Appendix A of"this .'

response. We are also performing an engineering analysis to, fdetermine conservatively the amourit of " mini-trim" and ensure that its use, allowed by a fieTd change req'uest, has not impacted the dosigri' basi s. Consequently, th( "minar" designation is

.sppropriate. We haVe also'revisid our? Harris Plant Engineering s Section instructions to ensure the Firez Hazards Analysis islassessed

~

for imNet tiy i'leid change requests. Alditional specific informaton '

relativdhto;this concern'is prodded irtour response to Deficiency S 2.4-4:in Appendix A qf tH s reporn -

i

_ _ .w q -

The IDI team 6xpressed a^ concern that the" cumulative imptet+of minoh changes is not' assessed in ' stress analysis calculations. 'While we agree that an accumulatidnhf a r; umber of minor changes could impact existing'Sti ss' Analysis calculations, the specific example mentioned Msulted from an isolatsd error iddntifying an Ebasco Flow '

. Diagram Design Cha'nge Etice not a field initiated change - as Mving no effect on Stress Analysis. Therefore, CP&L's " minor change" desigutfori is not applicable. Furthemore, Ebasco's 9

• procedure is to assess 'the impact of chahges collectively. The, appropriate calculation.has been rerun using the revised pipe. f.

schedule with no'impat.t'ori design. , Additional / details are provided

,in u- cur, responsof to Daficiency 03.2-16 in Appendix A of this response, llarris Plant Engineering Secticn procedures require that the substitution of m'aterials'other than those specified be aedomplished via'an approved design change document. "The purchasing group -

procedures also require procurement to specifications or approved change documents. We have reviewed field change requests for.

meterials substitutions for the subject specification and found that thfi y

if an isolated case. -

\

.g g Q

• *j s E. /\ , h 5 -

A '

,s 49 - 1

,- 9 g ,

p ;e v

7.4 MINOR DESIGN CHANGE DESIGNATION (Cont'd)

We reviewed the concerns regarding " minor" and " major" changes as applicable to our design change control procedures and have confirmed their adequacy or revised them as stated above. We do not believe the deficiencies cited constitute a programmatic deficiency, and we have concluded that no design impact resulted from them.

7.5 DESIGN INTERFACE PROBLEMS We have reviewed the instances cited by the IDI team and do not in all cases agree that they are evidence of faulty design interface.

For example, a conscious decision was made by the project to defer review of cable tray overloads for fire protection concerns (see response to D2.4-2). Similarly, a conscious decision was made not to calculate end loads for non-active valves, as stipulated in Westinghouse specifications. The decision not to calculate end loads, which is not a regulatory requirement, was reviewed in several letters between Ebasco and Westinghouse. Although written confirmation of Westinghouse's concurrence with this decision was received after initiation of the IDI, it is evident that inter-organizational communication is active and effective.

All design interface concerns identified by the IDI team have been addressed in Appendix A of the response and, where necessary, corrective actions have been initiated. Our review of the items indicates that design interface is not a problem. In fact, these concerns were isolated instances. In addition, all project personnel have been familiarized with these concerns to prevent their repetition and to ensure more effective communication in the future.

To further enhance interorganizational communication, we have initiated a program to place HPES supervising discipline engineers and section managers in Ebasco's offices on a rotating basis. Other programs currently in place are:

The Ebasco transition program The FCR review team

. The monthly discipline review meetings The SHNPP team building initiative These programs, along with our continuing review of any concerns relating to interorganizational communication, should ensure adequate interface control, t

i a

i 50

, APPENDIX A RESPONSES TO DEFICIENCIES, UNRESOLVED ITEMS AND OBSERVATIONS Item Title D2.2-1 (Deficiency) Westinghouse Design Interface Drawings 02.2-2 (Deficiency) Proof-of-Design Calculation D2.3-1 (Deficiency) Containment Recirculation Sump Design D2.3-2 (Deficiency) Containment Recirculation Sump Yortexing D2.3-3 (Deficiency) Refueling Water Storage Tank Vortexing D2.3-4 (Deficiency) Refueling Water Storage Tank Capacity D2.3-5 (Deficiency) Containment Spray System Eductor Flow Rate D2.3-6 (Deficiency) Containment Spray System Single Failure U2.4-1 (Unresolved) Cable Tray Combustible Load Calculation

+

D2.4-2 (Deficiency) Cable Tray Overfill D2.4-3 (Deficiency) Combustible Load within Fire Area 1-A-BAL D2.4-4 (Deficiency) Use of Minitrim (PVC) in Areas Outside Containment D2.5-1 (Deficiency) Volume Control Tank Shielding Analysis D2.5-2 (Deficiency) Control of Design Drawings D2.5-3 (Deficiency) Post-LOCA Shielding Design Review D2.5-4 (Deficiency) Gamma Radiation Source Strength Assumptions for Equipment Qualification D2.5-5 (Deficiency) Equipment Qualification Beta Dose D2.5-6 (Deficiency) Integrated Dose Analysis for Equipment Qualification D2.5-7 (Deficiency) Radiation Dose in Equipment Qualification Zone R-6 of RAB D2.6-1 (Deficiency) Installation of Charging Pump Room Air Handling Units D2.7-1 (Deficiency) Non-Seismic Piping Interaction Damage Study D2.7-2 (Deficiency) Seismic II/I Interaction of Field Routed Piping D2.8-1 (Deficiency) Field Installation Tolerances for Hangers D2.9-1 (Deficiency) Pump Vendor Drawing Error D3.1 -1 (Deficiency) Allowable Nozzle Loads D3.1 -2 (Deficiency) Regenerative Heat Exchanger Nozzle Thermal Displacement Data D3.1-3 (Deficiency) Stainless Steel Pipe Supports

APPENDIX A RESPONSES TO CEFICIENCIES, UNRESOLVED ITEMS AND OBSERVATIONS (Cont'd)

Item Title D3.1-4 (Deficiency) CP&L Pipe Support Procedures D3.1-5 (Deficiency) Supplementary Steel D3.1-6 (Deficiency) Pipe Support Stress Check U3.1-7 (Unresolved) U-Bolt Load Interaction U3.1-8 (Unresolved) Friction Anchor Clamps D3.2-1 (Deficiency) DBE Inertia / Functional Capability 03.2-2 (Observation) LOCA Anchor Movement 03.2-3 (Observation) Evaluation of Valve Accelerations D3.2-4 (Deficiency) Westinghouse Active Valve Qualification Program D3.2-5 (Deficiency) Modeling of Yalve Center of Gravity D3.2-6 (Deficiency) Emergency Condition Stress Ratio D3.2-7 (Deficiency) Thermal Modes D3.2-8 (Deficiency) Thermal Expansion Input D3.2-9 (Deficiency) Volume Control Tank Nozzle Displacement D3.2-10 (Deficiency) Nozzle Thermal Displacements D3.2-11 (Deficiency) Design Pressure 03.2-12 (Observation) Stress Summary Checklist 03.2-13 (Observation) Flange Evaluation D3. 2-14 (Deficiency) Node Point Spacing D3.2-15 (Deficiency) Anchor Location

, D3.2-16 (Deficiency) Pipe Schedule D3.2-17 (Deficiency) Regenerative Heat Exchange Seismic Analysis l U3.3-1 (Unresolved) ITT Grinnell Air Operated Valves

D3.4-1 (Deficiency) Pipe Support Strut Design D3.4-2 (Deficiency) B-P/CP&L Pipe Support Design U3.5-1 (Unresolved) Westinghouse Supplied Non-Active Valves U4. 2-1 (Unresolved) Coupling Effects of Eccentric Structures D4.2-2 (Deficiency) Mass Moments of Inertia U4.2-3 (Unresolved) Floor Response Spectra Based on One Dimensional vs Three Dimensional Seismic Analysis j D4.2-4 (Deficiency) Preparation of Calculations l

APPENDIX A RESPONSES TO DEFICIENCIES, UNRESOLVED ITEMS AND OBSERVATIONS (Cont'd)

Item Title D4.2-5 (Deficiency) Inconsistency Between Calculation and Seismic Models 04.2-6 (Observation) Peak Vertical Acceleration Assumption D4.2-7 (Deficiency) Shear Area D4.3-1 (Deficiency) Radial Reinforcement Stress Documentation D4.4-1 (Deficiency) Loading Combinations - Seismic Load 04.4-2 (Observation) Superseded Design Calculation D4.5-1 (Deficiency) Slab Design Using Direct Design Method D4.5-2 (Deficiency) Spacing of Slab Reinforcement D4.5-3 (Deficiency) Load Combination for Slab Design D4.5-4 (Deficiency) Seismic Analysis for Masonry Walls D4.5-5 (Deficiency) Use of Floor Response Spectra D4.5-6 (Deficiency) Design of Masonry Wall Stairway A-4 D4.8-1 (Deficiency) Load Combination for Main Dam Spillway D4.8-2 (Deficiency) Main Dam Spillway Abutment Design D4.9-1 (Deficiency) Boron Recycle Holif-Up Tank Seismic Loads D4.10-1 (Deficiency) Cable Tray Support Frequency D4. ll-1 (Deficiency) Frequency of HVAC Ducts D4.11-2 (Deficiency) Loads on HVAC Duct Supports D4.11 -3 (Deficiency) Frequency of HVAC Duct Supports US .'l -1 (Unresolved) Westinghouse Guidance for Nuclear Instrument Cables US.2-1 (Unresolved) Electrical Power Design Procedures and Guidelines US.3-1 (Unresolved) Independence of Electric Systems D5.4-1 (Deficiency) Protection of Safety Related Buses D5.4-2 (Deficiency) Motor Operated Valve Thermal Overload Protection D5.4-3 (Deficiency) Design Verification of Thermal Overload Settings D5.4-4 (Deficiency) Station Service Transformer Protective Relaying US.4-5 (Unresolved) Procurement of Quality Components D5.5-1 (Deficiency) Battery Sizing Calculation D5.5-2 (Deficiency) DC Equipment Rated Maximum Voltage D5.5-3 (Deficiency) Battery Discharge Voltage Profile 05.5-4 (Deficiency) DC System Minimum Voltage

APPENDIX A R5PNI5ETTO DEFICIENC,IES, UNRESOLVED ITEMS AND OBSERVATIONS (Cont'd)

Item Title 05.5-5 (Deficiency) DC System Control Room Indication 05.5-6' (Observation) DC System Undervoltage Alarm D5.G-1 (Deficiency) Penetration Protection Qualification D5.7-1 (Deficiency) Use of Motor Data in Setting Procedure D5.7-2 (Deficiency) 480-Volt Bus Undervoltage Alarm 05.7-3 (Observation) Motor Acceptance Testing D5.8-1 (Deficiency) DC Motor Operated Valve Voltage Drop D5.8-2 (Deficiency) Control Circuit Voltage Drop D5.8-3 (Deficiency) Reactor Coolant Pump Power Cable Voltage Drop Calculations D5. 9-1 (Deficiency) Reactor Vessel Level Instrumentation System RCP Inputs D5.10-1 (Deficiency) Site Engineering Design Change Control D6.1 -1 (Deficiency) Instrument List Data Base D6.1-2 (Deficiency) FSAR/ Instrument Index Consistency D6.1-3 (Deficiency) Diesel Generator Starting Air Compressor. Pressure Alarm D6.1-4 (Deficiency) Component Identification on Control Wiring and Process Control Block Diagrams 06.1-5 (Observation) Volume Control Tank Isolation Valve Interlock 06.1-6 (Observation) CVCS Design Basis D6.1-7 (Deficiency) Process Instrumentation Cabinet Interconnections 06.1-8 (Observation) Flow Indicating Switch Design Pressure and Temperature D6.1 -9 (Deficiency) EBASCO Procurement Specification D6.1-10 (Deficiency) Incomplete and Unissued Drafting Manual 06.3-1 (Observation) Train Terminology D6.3-2 (Deficiency) Conduit Separation U6.3-3 (Unresolved) Instrument Inpulse Line Separation Distance 06.4-1 (Observation) ITT Barton Differential Pressure Switches D6.4-2 (Deficiency) Vendor Conformance to Specification

APPENDIX A RESPONSES TO DEFICIENCIES, UNRESOLVED ITEMS AND OBSERVATIONS (Cont'd)

Item Title U6.5-1 (Unresolved) Design Basis for Safety Related Instrument Setpoints U6.7-1 (Unresolved) Westinghouse Reactor Coolant Pump Instrumentation 06.8-1 (Observation) Design Change Notification Supporting Data 06.8-2 (Observation) Sump Pump Control 06.8-3 (Observation) Licensing Group Interface 06.8-4 (Observation) Voiding Field Change Requests D6.8-5 (Deficiency) Battery Room Service Sink D6.8-6 (Deficiency) Calculation Basis for Licensing Amendment l

4 TD i

(

l l-l

i 1

! D2.2-1 (DEFICIENCY) WESTINGHOUSE DESIGN INTERFACE DRAWINGS

DESCRIPTION i

In late 1980, Ebasco and Westinghouse established a Design Interface Drawing List (DIDL), which identifies the Ebasco design documents that were the bases for Westinghouse proof-of-design calculations. Ebasco was to send periodically '

! to Westinghouse the revisions of these documents and the various design change 2

documents (DCNs/FCRs/PWs) that precipitated the revisions. i

< l The IDI team compared the DIDL list with the latest transmittal from Ebasco to

• Westinghouse to' determine if the required information was being sent. . The IDI j team found three of the eighteen drawings that Westinghouse had detemined to be part of the DIDL were not included in the Ebasco transmittal. These j drawings were CAR-2165-G-137, G-139, and G-245. The IDI team noted that the drawings were included in Ebasco transmittals prior to January,1982 but were ,

dropped in subsequent transmittals to Westinghouse. *

RESPONSE

I i

I The DIDL was generated automatically by extracting from the Project Drawing Control Log (DCL) those previously identified drawings determined to be  !

< necessary for Westinghouse proof-of-design calculations. The DCL is a '

controlled document that contains the latest status of design drawings as well as the individual design change documents (DCNs/FCRs/PWs). A special 1 identifier, "WTH", is provided in the DCL against those design drawings

required for the proof-of-design calculations. In addition, these same drawings are specifically identified on the drawing reproduction / distribution l schedule to ensure that Westinghouse receives these drawings each time the l drawings are revised and distributed.

Just prior to the latest transmittal of the DIDL, a computer clerical input i

error inadvertently changed the identifier for the above noted drawings from "WTH" to "WST". Subsequently, when the DIDL was generated for transmittal to Westinghouse, these drawings were dropped from the list. However, these drawings were not dropped from the drawing reproduction / distribution schedule. The above drawings were in fact transmitted to Westinghouse through the drawing reproduction / distribution schedule.

l Following the IDI, we reviewed the DCL and OfDL to detemine if any other

- design drawings were missing from the list. It was identified that one

! additional drawing was dropped from the DIDL due to the same clerical computer ,

i- input error. The DCL was revised to correct the special identifier for these l drawings. An updated DIDL, along with the individual design change documents, l (DCNs/FCRs/PWs) were transmitted to Westinghouse for their information and use. In addition, Westinghouse confirmed that they had in fact received the latest design drawings, and that the individual design changes transmitted have no significant impact on the system design. The IDI team recognized that

. this was not systematic.

I Drawings are now maintained by the Harris Plant Engineering Section. To  !

assure continued distribution of the Westinghouse interface drawing, the DIDL is now a standard distribution list maintained by the Harris Nuclear Project 4 Document Control group. Any revision of the drawings or any change document affecting these drawings is automatically distributed to Westinghouse. .

i i

, v.-_-~.,-,-,---- .-.-_._,-.,,-m.__,-,.~-_,, ,,....._,_.--,-___m_.. _ _ , , . , ~ _ , - , _ _ ~

D2.2-2 (DEFICIENCY) WESTINGHOUSE PROOF 0F DESIGN CALCULATION DESCRIPTION Errors were noted by the IDI team in the calculation that determined NPSH available for charging pumps and in determining the effect of an Ebasco pipe reroute. Calculation CWS-CQL-025 contained two pipe lengths that were twice as long as the drawing indicated and also contained an extra 900 elbow.

Calculation SD/SS-CQL-026C used twice the correct flow in the pressure drop.

A note on the calculation coversheet indicated that the calculation was supplemented by Calculation CWS-CQL-324C.

RESPONSE

1. Calculation CWS-CQL-025 In calculating system piping losses, errors were made in scaling composite piping drawings and in tabulating fittings and pipe lengths.

Due to the available margin in the calculation and the conservative nature of the errors, the operation of the centrifugal charging pumps should not be adversely affected. To correct the error, calculation CCWS-CQL-025 will be revised.

2) Calculation SD/SS-CQL-026C A calculation error was made such that the actual increase in the pressure drop induced by the piping change would be larger than the calculated increase and would not be within the margin provided by the orifice. The actual sizing of the flow restriction orifice occurred subsequent to CWS-CQL-324C in calculation FSD/SS-CQL-642, which was reviewed during the inspection. Calculation FSD/SS-CQL-642 effectively includes and supersedes calculation CWS-CQL-324C. A note to this effect and a copy of this report will be attached to the file copy of CWS-CQL-324C. The orifice sizing has been demonstrated to be effective.

The IDI team concluded that these were minor errors and were not systematic in nature.

U

\

- D2.3-1 (DEFICIENCY) CONTAINMENT RECIRCULATION SUMP DESIGN f

DESCRIPTION

[

The IDI team concluded that the SHNPP recirculation sump screen design is not i consistent with the guidance of Regulatory Guide 1.82 and is therefore contrary to the Shearon Harris Nuclear Project's FSAR commitments. In

particular, the IDI team concluded through a review of calculations and design '

i drawings that the water level in the containment following a safety injection l does not result in the sump screen being fully submerged, and the approach i velocity of water at the face of the sump screen does not approximate the value identified in Regulatory Guide 1.82 (0.2 feet per second).

- The IDI team states in the " Background" paragraph for Item D2.3-1 that "this guide (RG 1.82) also indicates in position C.8 that the top deck of the sump should be designed to be fully submerged after a LOCA and completion of safety injection". In addition the IDI team further notes in the " Description"

. paragraph that, "first and most importantly, the water level in the

! containment following safety injection does not result in the sump being fully  :

submerged. Secondly, the approach velocity of water at the face of the sump  !

i screen does not " approximate" the value identified in Regulatory Guide 1.82 l . (0.2 feet per second).

RESPONSE 4

! We have reviewed the IDI team's finding concerning the containment i' recirculation sump design and concluded that the Shearon Harris Nuclear Project design not only meets the guidance provided in USNRC Regulatory Guide 1.82 but actually exceeds the requirements.

1. We are fully confident that the design provides the proper safety function 1 necessary to mitigate the consequences resulting from a LOCA. Our position is reinforced by the conclusion documented in the Shearon Harris Nuclear Project SER (NUREG-1038), Section 6.2.2 where it is stated that "the Applicant's sump ,

design confonns to Reg. Guide 1.82 and is therefore acceptable".

I The Shearon Harris containment recirculation sumps are designed to provide an adequate supply of water to the containment spray (CT) and residual heat removal (RHR) pumps during the recirculation mode of the emergency core i cooling system (ECCS) and the containment cooling system (CCS). There are two i

separate containment recirculation sumps located 900 apart in the annulus ,

formed by the secondary shield wall and the containment liner (see figure 1).

Each sump is divided and contains the suction intakes for one RHR pump and one i CT pump. The' pump suction intakes are located in the floor of each sump at elevation 216.66 ft. The sumps are enclosed by a concrete structure with a j solid steel roof. Six (6) 4' x 4' vertical openings in each sump facing the l secondary shield wall admit flow. Each opening houses a convoluted perforated

fine inner screen (p i

flow area of 26.5 ft} ate),These .

40% open, fine inner having 1/8" diameter screens holesby are protected with an an open outer i l trash rack with 1 1/2" square openings. Protecting the sumps from high i density debris is an 18" high curb around the structure. The floor l immediately in front of the curb is sloped gradually down and away from the

! sump (approximately 3" from the highest to the lowest point). Figures 2, 3,

,. and 4 describe the sump structure, screens and curb. 4 4

l 1

i

,-..-,-..-9n--_-.._- , , . , - . - _ __.cm-m- . . _ . _ . . + - - - , -

f D2.3-1 (DEFICIENCY) CONTAllMENT RECIRCULATION SUMP DESIGN (Cont'd)

RESPONSE.(Cont'd)  ;

The USNRC Regulatory Guide 1.82, Sumps for Emergency Core Cooling and Containment Spray Systems, provides methods acceptable to the Regulatory staff 4

for implementing requirements with regard to design, fabrication, and testing

_ of sump or suction inlet conditions for pumps in the emergency core cooling

and containment spray systems. This guide is not specific for any particular ,

sump design, but rather provides general and specific recommendations to be implemented in order to assure adequate performance of the sump.

l As noted in USNRC Reg. Guide 1.82, the fine inner screens should be of sufficient screen surface area to keep the flow velocity at the screen to approximately 0.2 ft/sec. In addition, Reg. Guide 1.82 requires that ,

consideration be given to partial screen blockage in sizing the fine screen in  :

order to assure an adequate margin of conservatism on free flow area through r the screens. For the Shearon Harris design, a convoluted fine inner screen design was selected. The convoluted design of the fine inner screens increases the effective screening surface area. In addition, the convoluted screen improves performance of the screen since the debris that does reach the 4 fine inner screens will tend to accumulate in the downstream corner; the 1 upstream corner is kept clean. This can be seen in Figure 5.

Furthermore, all piping insulation utilized inside the containment is of the

, reflective mirror type design which minimizes the potential for this 4 insulation to float and thereby prevents the accumulation of insulation at the

! screens. In addition the reactor coolant loop pi)ing is totally contained within the secondary shield, along with most of tie auxiliary branch piping.

Since the secondary shield wall has relatively small openings at the base of

. the containment floor, the potential for large floating debris to reach the screens is further reduced. Light material that is considered as potential source of floating debris such as paints and coatings have been specified and

qualified to be of special type and application such that the material will not be a source of debris.

The IDI team interprets the Regulatory Guide to require that the containment i recirculation sump must be fully submerged following a safety injection. We .

l disagree with this interpretation. The Regulatory Guide discusses sump l l submergence in two specific areas: paragraph 8 (page 1.82-2) states that "no credit should be taken in computation of the available surface area for any  !

top horizontal screen, and the top of the screen structure should preferably be a solid deck". This guidance is provided since the uncertainties about the extent of water coverage on the screen structure, the amount of floating debris that may accumulate, and the potential for early clogging do not favor the use of a horizontal top screen. Furthermore, paragraph C.8 states that, '

"a solid top deck is preferable, and the top deck should be designed to be fully submerged after a LOCA and completfon of the safety injection.

i o

9

D2.3-1 (DEFICIENCY) CONTAINENT RECIRCULATION SUMP DESIGN (Cont'd)

RESPONSE (Cont'd)

The above requirements are clearly intended to assure structural capability of the screen structure if the top deck of the structure is fully submerged. The Shearon Harris recircliTation sump is designed to accomodate the loads resulting from the highest containment water level following a LOCA and completion of the safety injection, along with other concurrent design basis loads. Therefore, the IDI team conclusion that the design of the recirculation sump does not meet the Reg. Guide " requirement" for screen submergence for the minimum containment water level is not substantiated by the language or the intent of the Reg. Guide.

The IDI team also noted that "even though the emergency sump screens are apparently not fully submerged during the minimum containment water level conditions, the IDI team found no considerations given to the effects of floating debris in the design analysis reviewed". In addition, the IDI team stated that no consideration of functional loads was found. However, the team noted that if the water velocity is maintained sufficiently low in accordance with the regulatory guide the functional load would be small. As discussed below, since the velocity of the water is approximately 0.2 feet per second and since the arrangement of the sump relative to the major plant features minimizes the potential for large floating debris to reach the sump, the functional loads on the screen structure due to floating debris can be assumed to be negligible.

Calculations performed subsequent to the IDI concluded that the minimum containment water level resulting from a LOCA and completion of the safety injection is elevation 223.96 ft, which is 2.96 feet above the containment floor. This elevation was calculated assuming the most limiting single failure, the minimum volume of usable water in the Refueling Water Storage Tank for the injection mode, and the most conservative instrument inaccuracy.

Further, no credit was taken for primary coolant system inventory.

Considering the above, the containment recirculation sump water velocity at the fine screens was calculated to be approximately 0.2 feet per second.

The velocity is calculated as follows:

l V=0 l A l

r Where:

I j V = velocity at the fine screens, ft/sec

Q = total flow rate for each containment recirculation sump, ft 3/sec A = 50% of theminimum calculated free flow area of the fine containment screens water level,gubmerged ft by the l

i I

D2.3-1 (DEFICIENCY) CONTAINMENT RECIRCULATION SUMP DESIGN (Cont'd)

RESPONSE (Cont'd)

The total flow rate for each containment recirculation sump is 5,388 GPM, which is comprised of the flow associated with: one (1) residual heat removal pump (RHR) -3,000 GPM; one (1) hi-head safety injection pump (CVCS) -500 GPM; and one (1) containment spray pump (CT) -1,888 GPM.

The flow area to be considered in calculating the maximum water velocity is equal to 50% of the free flow area of the fine scrgens submerged by the minimum containment water level, which is 61.39 ftc.

Therefore, the velocity at the fine inner screens is:

V = (5,388 gal / min)(1 min /60_sec)

(61.39 f t')(7.481 gal /f tJ)

V = 0.196 ft/sec Following a LOCA and completion of safety injection, additional water will spill from the primary coolant system (steam generators, reactor vessel, pressurizer, piping, accumulators). Under these conditions, the screens will be fully submerged and the velocity will be 0.15 ft/sec.

The IDI team evaluation of the water velocity presented is inconsistent with the guidance provided in Regulatory Guide 1.82. For example, the IDI team refers to an " approach" velocity limitation of 0.2 feet per second. The Regulatory Guide states, however, that 'ne velocity at the fine inner screen should be approximately 0.2 feet por second. We acknowledge that coolant velocities greater than 0.2 feet per second are possible u) stream of the fine inner screens. These velocities are more dependent upon t1e geometry of plant features in and around the containment recirculation sump than on the sump screen free surface flow area. Furthermore, the inherent superior performance features of the convoluted fine screens offer more resistance to screen blockage, i.e., debris that does reach the fine inner screens will tend to accumulate in the downstream corner while the upstream corner is kept clean.

Also, the IDI team assumed an " analogous 50 percent blockage of the trash racks" and concluded that a resulting flow velocity would be in excess of 0.2 feet por second. Again, there is no specific or implied requirement that the velocity at the outer trash racks be approximately 0.2 feet per second. The intent of the Regulatory Guide is to address the capabfif ty of the containment emergency sump to provide an adequate water source to sustain long term recirculation cooling following a large LOCA. This intent is satisfied by minimizing the LOCA-generated debris effects. As noted earlier, the plant design features have been selected specifically to satisfy this requirement.

In light of the fact that all design requirements were satisfied, we conclude that design verification was performed consistent with ANSI N45.2.11.

f '

+e 0 8 g

$6 4 h d

g j - -

i I

e -

i

. E e /

a *< #

L.

~

5e s. m p

i s'

l

/ ,.

. , - .. ~ s s y a_,a -

e e

x.

s.

1 g m

? , s A- I .,8 N. h y 4

. +#~ .

g - -

3 .

T y l

\, ymo E -

w r a: - -

l B a

?$ e i ...

)

, - . . . . .... - .. .... ,.. ... _ _.y _ _ , , , _ , , , , , , , , ,

o B

4.0' -*

', .,e.<? , , c

.n a

, . . "...A. r. ,,

I *

'$  ?

l'

f. h.

5'

'if 15 18"

1. y'.

J. . . ',f...'l2p 7

{1 9: c.9,h 1 n

u

/ OUTER TMNFR Rf'.R F FN l/R" Hof,E f)fA. . RACK PERFORATED Pi. ATE, 40%

OPEN, 26.5 FT2 M . ,

FREE FLOW AREA A-A TI i e n i Ie

o. . .. o .. .. o . .

. 9.. _o.

is l 's i l 's ,

ll l' ll l' l' ll ll 5/8" ora.. # d dA'4ME .

4.. ll ll i

lll

.i

ll ll is is i ll i.

6 .

ROD T i INN ER SCREEN 18 es e

i fA Jf A

.l :e I se e se il se i is 18 se s, e el 11 si i e _in it is ei in in is , INNER 9' ei e' i. j. e, e je ~,' '

SCREEN

,, ,i i ,, s, i

l ll e il je

,l l

8:

l.l ll le il ll's  : o!!!!

oo o3.83' l',l o 6, .m. _._. ---. 'i,'

Il i

ll .. - .

..-r~:: '. . I I,

11

,e l,e

i,el  : 1 i,

- = :l :l eI le e i

-::, is is is li ,

w- - ~ ~ ~ .

an in in in in e

]' b s-

~

o~

OUTER TRASH RACK 1-1/2" SQ. OPN., W4if;$.wi'dG;. 1/2" SQ. BARS ON 2" CNTRS. . FRONT VIEW 33 SilEARON llARRIS NUCLEAR POWER PIANT Figuro 2

                                                                                                                                                                                --~

__ . - . - - - . ._ .a-- --._,,a., . - , ~ - . - - . .

                                                                                                                                                                                      .,               .m:g g

li A=3

                                                                                                                                                                                                               '4 4

L - i - 4

                                                                                                                                     ;        .r                                             i 4

A , s

                                                                                                                                                                                                              -b Sot!D 800F                                                                           *    .

d I .:"4 s

                                                                                        ' ., 4                                                ,4.                                                               k 87 EL 222.9*                                            ,                                            4.
                                                                                                                                               )

EL 224.0' O. , - . . - . .

                                                                                                                                            ,,4

4. , . . .

                                                           . s. .
                                                                          . . r. .             ..

a. . .

                                                              ?.
                                                                        '~g       ..
                                                                     . ' . . ~ . '.,';)
                                                                                    .                                              .,,h:.

r; . * ' , . s . ; Q , - { 'C'.', ',' YO e M t6.31l' i i '_a,

                                                                                              . ,%. *.. I                  l . %,,'A) e l fln i               .M. -

s -f.s A-A rt 227.o  ! carcatato etatt ener I l T.. tl w s.a,.i

                                                      /       -

2' i i'

        .                                    I                                     *                            *
                                                                                                              '                                                       l st 221 &                    l             ,y                               ,

l . Et a: .o _

                                                                               ,                        C-C
                                  .                                            C

_*/e ___/ ' tt 222.s 4 re *

• bk -

                                                                                                                                                  '"33                          _i l        6                                                        ~I
                              'E       F               I                         I                            -
                              /      -                 i                       / is             ,

I :, F  ! II l t l'  ! l 1 I i i l i U t . .ie. *. -p . 0 **

                              \                        l aus                               $,*,f
                                                                                           '                         hLADott                                               /
                                                                                                              ,.?                                               'T, (lyAt*-                         %, ~e s'} } ger,na . cf. \f/-                                  i.

A l t in-  ! e SUMP PLAN I ROOF _ REMOVFI) SilEARON llARRIS N!! CLEAR POWER PLANT , rigure 3

                                                                                                                                                                                                                 .i e

  . . ., o O

L h '.g' C W M*4HeoT CTCO+1bq %AM

                  '.     % Mitt,rs g g
                                                                                                                                                                             . . . . ~

f

                ,                                                                                                                                                                    -I    . s s

N

                                                                                                                                                                      \                     ,                        n r

_e.

                                                                                                                                                                                               \                 fjk
                                                                                                                                                                                                   ;\                     . )

e m m ...

                                                                                                                                                                                                                                 \ \

e ,~ . CHtengt ug, Eb AIE

                                                                                                                                                                                                                       % 9)

({nf 71F 6) \\> ] f~ l

                     /

V

                                           /
                                                                                                                                                                                                              ,\            -
                                    'IL. 27f.0 sb"Cug f                                                                                                    .

( SilEAnon qqhRRIS NUCLEAR pwER PLANT / Figure 4 l l i m.

                                                        ,       ~

i 1

                                                        ,            ,~p.-

) q F1.a] 50 EE4

 ;                                              *) > . [ L ~                          1 I

l] sn .vc ceen5 Auov.utA*i se co F6At cc.rn,4 sq A (epuniuer iu MAmmu,A g4gcic.pco oc sc I ~- ~ V' m,ws - l I c.or\wdct JJit') ; ut 7tNA # Ct) EL =3 N- l \ _. +-

                                                        '  4cowotuise 5 creed
                                                           ,g.

r^ - n l G4o+/15 DC6t qtS AccutAut A1 44 tu t>ow45' QEASA . C0Q)ER,,, Or }ut scgstd.uPp}(fM(c(uf(,gt CLEAM

/

SilEARON llARRIS NUCLEAR POWER PLANT Figure 5 I 1 1 l l J

D2.3-2 (DEFICIENCY) CONTAINMENT RECIRCULATION SUMP VORTEXING DESCRIPTION The IDI team concluded that the methodology employed to evaluate containment recirculation sump vortexing was incorrect. In particular the IDI team concluded through a review of calculation TANK-13 perfomed by Ebasco, that the critical submergence required to prevent vortexing at the low pressure safety injection and containment spray pump suctions was incorrect since the equation used appeared to be inappropriate for appifcation to the containment emergency sumps. The IDI team further concluded that "the veriff er of calculation TANK-13 did not establish that the appropriate design method was used, and that design control has not been maintained".

RESPONSE

We believe the IDI team's concerns regarding vortex prevention in the containment recirculation sump are not substantiated. The best available methodology was used for this item and design control has been maintained. Consistent with the literature published by industry and academia on vortex prevention, the equation utilized to evaluate sump vortexing is a strong function of the Froude number. Using this equation, the minimum submergence required to prevent vortexing in the recirculation sump was detemined to be 3.0 feet. We have reviewed this value considering the guidance presented in NRC document NUREG-0869, USI A-43 Resolution Position published April,1903 for coment, and concluded that the methodology presented in this document is also a function of Froude number similar to the subject equation. The wrfters of USI A-43 and other literature on this subject, however, have concluded that it is impossible to determino all the essential facts for a given fluid flow by pure theory, and hence much dependence must be placed upon experimental investigations. Subsequently, in USI A-43, the minimum su) mergence suggested for the prevention of vortexing is 10', which was based heavily on the data collected from experimental tests. Because of the need for experimental test data, Carolina Power & Light Company contracted with Alden Research Labs to perform a test consistent with the regulatory gutdance to determine the adequacy of the containment recirculation sump. Since the results of the test demonstrated the possibility of vortex formation, recommendations to Jrevent vortexing were outlined. We are presently incorporating into t1e recirculation sump the necessary modtfication to prevent vortexing. With regard to the 10! team's concern related to adequacy of design verification, we note that one acceptable method of design verification (per ANS! H.45.2) is the use of availabic test and/or physical modeling data. This was clearly the case for the Shearon Harris design. l

02.3-3 (DEFICIENCY) REFUELING WATER STORAGE TANK YORTEXING DESCRIPTION The IDI team found an unsubstantiated assumption was used for the volume of water required to remain in the tank to prevent vortex famation. In particular the IDI team concluded through a review of calculation EQS-2 that no justification was provided for a statement that "54,430 gallons of water was required to prevent vortex formation." The IDI team further concluded that the assumptions and justifications presented in the calculation were not substantiated nor verified and that this deficiency is indicative of a weakness in the preparation and design verification of design analyses.

RESPONSE

We believe IDI team's finding regarding an unsubstantiated assumption noted in calculation EQS-2 for the volume of water required to remain in the tank to prevent vortexing is erroneous. As stated in calculation EQS-2, the purpose of this calculation was to establish the Refueling Water Storage Tank level instrumentation set points. The IDI team statement "that 54,430 gallons of water was required to prevent vortex famation" is inaccurate. The true statement in EQS-2 is that "54,430 gallons of water were available for vortex prevention and shutting off pumps". This statement was not intended to suggest or imply that the 54,430 gallons is required to prevent vortexing. The objective of calculation EQS-2 was to establish set points to satisfy previously established values for various modes of operation and function, not to determine the minimum level of water requried for vortex prevention. Tank-13 Rev 2, which is referenced in calculation EQS-2, and was reviewed by the IDI team established the minimum water level required to prevent vortexing in the tank. Although the methodology used to determine the minimum water level required torecirculation the containment prevent vortexing(was similar totothe sump see response Deffmethodology used cf ency D2.3-2), for the potential for vortex formation was further reduced by installing a 900SR elbow turned down at the CCCS inlet nozzle. This resulted in approximately 5'-0" (59,250 gallons) of effective water height to prevent vortexing in the tank. In accordance wf th Hydraulic Instituto Standards, 3.71 feet of water is required to prevent vortexing. As such, the available water " set aside" to prevent vortexing from EQS-2, satisfies the design requirements. Based upon the above discussion the IDI team's findings regarding an " unsubstantiated assumption" in EQS-2, and the failure of the preparer and verf fler to ensure that design inputs and assumptions are justifled, are erroneous. Therefore, this is not indicative of a weakness in the preparation and vertfication of design analyses.

D2.3-4 (DEFICIENCY) REFUELING WATER STORAGE TANK CAPACITY DESCRIPTION The IDI team found that the ninimum available capacity of the refueling water storage tank for injection into the containment building via the reactor core or spray headers following a LOCA is not consistently identified in the FSAR. The IDI team was also concerned because Ebasco procedure E-77 provides guidelines for generation of design inputs, which allows the FSAR as a source of design input. The IDI team states that "the use of an FSAR which has incorrect or inconsistent information does not provide a consistent basis for making design decisions, accomplishing design verification measures, and evaluating design changes".

RESPONSE

It must be recognized that Ebasco Procedure E-77 provides gutdance for identifying, selecting and documenting design inputs, and that the generation of design inputs originate from a review of applicable documents which includes the Safety Analysis Report, llowever, the information contained in the Safety Analysis Report is prepared and documented in accordance with other Corporate procedures which require design verification of these inputs. Consequently, the " design inputs" contained within the FSAR have been previously reviewed and vertfled as to their accuracy and appifcability. In this regard, the FSAR is not a detailed design tool but reflects data from controlled documents such as drawings and spectfications which do provide a consistent basis for design. When design c1anges are made on primary documents the FSAR necessarily lags. In particular, the volume of water required to allow switchover from the injection to the recirculation mode of operation is contained in calculation TANK-13. This calculation was based upon the guidance provided by Wostinghouse and supersedes the data presented in the Shearon liarris Nuclear Project FSAR. The FSAR will be revised to reflect the correct switchover time consistent with the latest design data. The Shearon liarris Nuclear Project has recently initiated an FSAR consistency review of a sample FSAR chapter. The pur)ose of this review is to demonstrate consistency between the FSAR and the "as-)uilt" design of the plant. Although this program is not intended to act as a design vertfication effort, it will demonstrate that while minor inconsistencies may be discovered, the design processes are controlled and that these inconsistencies in and of themselves do not adversely affect plant safety, i

D2.3-5 (DEFICIENCY) CONTAINMENT SPRAY SYSTEM EDUCTOR FLOW RATE DESCRIPTION The IDI team reviewed calculation CT-27 and concluded that the methodology used in the calculation to determine the eductor minimum and maximum flow rate was incorrect and the hydrodynamic conditions under which the eductor will operate was in error. Furthermore, the IDI team considered this deficiency as another example of an error which should have been detected during the design verification process and that this deficiency coupled with Deffcfencies D2.3-1, D2.3-2 and D2.3-6 appears to be indicative of a systematic problem with implementation of the verification process.

RESPONSE

We have reviewed the IDI team's finding against the calculation associated We acknowledge minor with errors theincontainment spray the calculation (system incorrect eductor piping network flow rate. model and a sign reversal). Based uson our review of the other deficiencies identified by the team and the fact t1at this deficiency resulted in only a minor change in the flow rate and had no further imp.ict on other calculations, analyses and design, we disagree with the IDI team's conclusion regarding a systematic problem with the implementation of the design verification process. As observed by the IDI team, calculation CT-27 involves a complicated piping network and is based upon a trial and error method for detemining system flow rates and hydrodynamic conditions. An erroneous aarallel branch assumption was used for the system piping that connects to t1e sodium hydroxide supply inlet of the eductor. This resulted in identical overall flow coefficients for both eductors. Since the piping is different in this area for each eductor, the flow coefficients are not identical. The pre)arer and the checker assumed the piping networks to be similar, since t11s is the design philosophy enployed throughout the plant design. Notwithstanding the above, calculation CT-27 was revised to reficct the actual piping network and to correct the sign reversal. The results of this recalculation yielded only minor changes to the overall hydrodynamic conditions in the system and eductor flow rates. This minor change in flowrate did not in itself effect the pil calculations.

D2.3-6 (DEFICIENCY) CONTAINMENT SPRAY SYSTEM SINGLE FAILURE DESCRIPTION The IDI team reviewed the flow diagram for the containment spray system to confim that the design met the single failure criterion for nuclear power plants. The IDI team found that a single, normally closed, manual valve is used to isolate the emergency sodium, hydroxide fill connection from the suction side of both containment spray eductors. The IDI team concluded that if this valve is mistakenly opened and remains open undetected, a common mode failure of both trains could occur, preventing tie addition of sodium hydroxide through the containment spray system. The IDI team also noted that "the failure to identify this common mode failure through the design verification process is systematic because of other examples found during the inspection of inadequate implementation of the verification process." l

RESPONSE

We have reviewed the IDI team's finding against the design of the containment spray system, the regulatory guidance provided on this subject, and the Itcensing bases for the plant and we believe the conclusions expressed by the i IDI team on this matter are erroneous. 1 l The subject valve is normally closed and is not required to be openod for any plant operating or testing mode. The nomal source of NaOH for the containment spray system is from the chemical addition tank. This tank is , sized to hold the entire amount of NaOH required to maintain the post accident ' L emergency recirculation sump pH within specified limits. The emergency Na0H l fill connection is provided for addf tional backup and is not expected to be used. In addition, considering that the containment spray system f s used only af ter a LOCA or MSLB (a one time event), the erroneous opening of this valve j is considered an unrealistic failure scenario.  !

It should also be noted that although the emergency sodium hydroxide addition connection is equipped with a non safety quick disconnect fitting, the  !

disconnect fitting is installed in a safety related sof smic Category I piping system. This quick disconnect would prevent air ingestion if the isolation valve is left open. l Nevertheless, in order to assure a more positive method of maintaining the f valve in its closed position, locking devices have been speciffed for the t valve. t Based upon the above we do not consider this item as indicative of a failure of the design verification )rocess and this item is not an example to support t the IDI team's conclusion t1st inadequate triplementation of the vertfication ' process exists. ' i i l l I f i

U2.4-1 (UNRESOLVED ITEM) CABLE TRAY COMBUST!BLE LOAD CALCULATION . DESCRIPTION Electrical Calculation 46-A0 calculates typical cable combustible values for power, control, and low level cables in trays throughout the plant. Typical calculations are performed in order to avoid complicated specific calculations and still provido conservative, representative values for the three categories of cable trays. These values are used as design input to Fire protection combustible load calculations. One basis for Calculation 46-A0 is an assunption that a typical cable type is an adequate re)resentation of all other cable types. This cabic type is used to establish t1e Stu value of the insulation per foot of cable in trays. For single conductor power cables, size AWG 6 was selected because it is the median size cable. For control and low level cables, the selection was based on the cable type wf th the largest length ordered and installed. The IDI team found no justification for these assumptions to show that the tables selected are representative of a typical value of Btu /f t of cable, or that the approach is conservative.

RESPONSE

We belfove the approach for selection of cables used in Electrical Calculation , 46-A0 is conservative. The justification for power, control, and low level cable selection is discussed below: Calculation 46-A0 calculates the combustible load of a tray by determining the  ! number of resresentative cables normally pennf tted in it and multiplying by the combusti31c load of the representattye cable. Since the number of cables permitted in a tray is controlled by cross-sectional area (i.e., percent fill), the relative combustible values of the cables are a function of their cross-sections. As an af d in analyzing the cross-sections, the area of the non-combustible material (copper) is subtracted from the total area and then divided by the total area. This yields a percentage of combustible material of the cable by area. Total area - Copper arca X 100 = Percentage of Combustible Haterial Tolal area power Cable The representative power cable was 0/M 025 03 (1/C No. 6 AWG) manufactured by Kerfte. This cable had a percentage of combustible material of 88%. D/M D25 03 is considered a conservative cable to use in the calculation since its combustible material ratto envelopes 83% of the total footage of power cable. l The other bulk supplier of power cable for this project is Anaconda Ericsson. . The Stu values that were obtained from Anaconda for their cable were less than those supplied by Korf te. Therefore, Korf te cable is a representative sample since the Stu values and the combustible material percentages for all Anaconda cables are less than the selected Kortte cable. l I

l U2.4-1 (UNRESOLVED ITEM) CA8LE TRAY COM8UST!BLE LOAD CALCULATION (Cont'd) RESPONSE (Cont'd) As most of the control and low level cables have a relatively constant combustible material percentage regardless of conductor size, the selection of ! one cable as opposed to another is less important. The representative cable types selected for both control and low level cables were based on the most frequently used cable type (total footage and total , number of circuits installed). 8/M D50-11 (2/C No.16) was selected as the i representative control cable as it accounts for approximately 24% of the total footaqe of control cable routed and 22% of the total number of circutts. 8/M D60-O' (Ipr No.16) was selected as the representative low level cable as it accounts for 29% of the total footage and 32% of the total number of circuits. Both cables have combustible material percentages of 97%. Calculation 46-A0 will be revised to state more clearly the criteria and assumstions delineated above. As the representative cable types selected do j not ciange, associated Fire Protection calculations do not require revision. l l l l l l l l l i l l

02.4-2 (DEFICIENCY) CABLE TRAY OVERFILL DESCRIPTION Cable trays with power cable fill in excess of the design criterion were justified from an electrical point of view; however, fire protection design calculations are not revised acuordingly.

RESPONSE

The abcVe situation was discussed prior to initation of the IDI on November 27, 1984 during a meeting between CPAL and Ebasco. The Shearon Harris Nuclear Project agreed that cable tray overfills need not be reviewed at this time by l the Fire Protection group due to the conservatism built into the combustible loading calculations (Letter E0-C-18131 dated January 14, 1985). The existing combustible loading calculations conservatively assume that all cable trays are filled to their nominal maximum capacity, except for three fire areas, where the average tray fill plus 5% addf tional margin was used (See FSAR Section 9.5.1.J for details). In reality, most cable trays are not filled to maximum capacities. Consequently, we believe that individual review of cable tray overfills for combustible load considerations is unnecessary at this time. l The calculations will be revised, however, to reflect the latest information prior to core load af ter cable routing is essentia11y complete. Since the calculations must reflect the final "as-built" conditten when the plant becomes operational, a logical engineering approach has been developed which forms a plan to achieve this goal. We do not view this as a serious concern. In fact, this is an example of the Applicant's proceeding on a course of action which is prudent both in the sense of ensuring accuracy of the calculations and cost control through the elimination of unnecessary and repetitive calculations (which would not have represented the final plant conffguration).

D2.4-3 (DEFICIENCY) COMBUSTIBLE LOAD WITHIN FIRE AREA 1-A-BAL DESCRIPTION Amendment 18 of the FSAR revised the combustible loadings in fire area 1-A-BAL  ; based upon revision of calculation FP-5-1-A-BAL. However, the FSAR amendment ' incorrectly reported the total combustible load in fire zones 1-A-3-PB and 1-A-3-COMI. Calculation FP-5-1-A-BAL was inttially >erformed assuming that all cable trays were 4 inches deep. Revision 1 of tie calculation corrected I the depth of control instrument cable trays to 5.25 inches. To account for this increased depth, the combustible load per unit area was multiplied by a  ! factor of 1.3125; however, the total combustible load was not increased in a i similar manner. As a consequence, the combustible load per unit area ' correctly reflects the potential fire severity within a fire zone, but the total combustible load is incorrect.

RESPONSE

Revision I to Combustible Loading Calculation FP-5-1-A-BAL changed values of 8tu/Sq Ft. The corresponding FSAR table was changed to reflect these new values but the total 8tu values were not revised in the calculation or FSAR. The original calculation was based on the sasumption that all electrical trays were 4 inches deep. All calculations, including F."-5-1-A-BAL, were revised where necessary to account for new input from Electrical which identifled cable trays of 4 inch and 51/4 inch depth. Eleven calculations were ori concerning cable tray depth.ginally We haverevised reviewedbased all 11 on andthe input found thatfrom Electrical the total 8TU values were correctly re calculated in all but FP-5-1-A-8AL, which ineludes ffre zones 1-A-3-COMI and 1-A-3 P8. The existing design is adequate. The revised Stu/Sq Ft values, upon which the , level of fire protection is determined, were correctly calculated in FP 5-1-A 8AL and FSAR. , Fire Protection has revised FP 5-1 A-DAL to change total 8tu values and issued  ! an FSAR Change Request. i t

02.4-4 (DEFICIENCY) USE OF MIN! TRIM (PVC) IN AREAS OUTSIDE CONTAINMENT DESCRIPTION Burning polyvinyl chloride (a type of plastic) can be hazardous to fire fighters and detrimental to the operation of safety related equipment, and it is normally excluded from use in sensitive fire areas. The use of Minitrim, made of polyvinyl chloride, was not pemitted by Shearon Harris project specifications; however FCR-E-030 pemitted limited use in a specific appifcation. Minitrim was pemitted as a trimming material for cutouts at cable exits in non-seismic instrument cable trays. The material was to be used on those trays constructed with solid bottoms and covers. In FCR-E-3499 the use of Minitrim was expanded to protect cable from edges of cable trays and equipment. FCR-E-3499 was approved without being sent to the field fire protection discipline or to Ebasco's fire protection group for review and approval.

RESPONSE

Harris' Plant Engineering Section Instructions 3.1 and 3.3 have been revised to ensure the Fire Hazard Analysis is specifically considered for impact by Field Change Requests. Training has been performed to raise the awareness of site discipifne personnel with the content of the Fire Hazard Analysis and the specific requirements of the revised procedures. An engineering analysis is presently underway to ensure that the Ifmf ted use of Minitrim will not impact design. This analysis will conservatively detemine the amount of Minitrim presently existing in cable trays and motor control centers, investigate suitability of alternate materials and ensure use has been minimized.

_ _ _g - - n . y q

                                             .         l-                       .

~

7 ' 9 , i m w U g

                                /                                                                     .,
                                                                                                                                 -                  ,     s                       ,

D2.5-1 (DEFICIENCY) VOLUM( CONTR0t. TANK SHIELDING ANALYSIS - DESCRIPTI6N w j ' N

                    ,        j &                          ,
     ' CalculatiorFVC-05-VT-51.1 used scurce strengths for the liquid and vapov phases of the volume control taiWfrom the PSAR. The so@ce.of tke PSAR source strengths appears to be a Westinghouse'UCAP. The currentlsource terms provided infestinghouse's Final Radiation Analysis' Manual revised the, energy grouping, however, and the source strengths appear;to be higher than the PSAR numbers by inspection.
                                                                                       '                                     ~                                                 ,
                           <                                          g                                                        -

7-RESPONSE h .

                                                                                                                                                                       'c    -

4 . s s - Calculation 025 was perforred during the early stages of construction to verify the adequacy of shielding around .the volume control tank. > Source terms

                                                                                                                                            ~

were obtained from the PSAR which contained the latestivalues aWilable at that time, and were neither incorrect nur inconsistent. While the Ib1 team is correct in observing that the calculation was not revised when Westinghouse issued a revised source term docuc4nt, this resulted from the engineering judgement exercised by the calculation originator that conservatisms prssent , in the original calculation.more than compensated for any increases in'sourcer strength. Thus, the shieldidg Aesign was judged to be still satisfactory. ;' In response,to the< inspection findings, a full, more sophisticated reanalysis, (calculation 040) superseding ths, original calculation 025 inspected by the ' ' IDI team," was performed utilizing: revised Westinghouse data (Final Radiation 4 Analysis Manual). The results of this new calculation clearly demonstrate that the design' remains edequate,ssubstantiating.the judgement of the s v calculation originatora :In fact, the conservatism' inherent in the original calculation may be apprdciated by noting that calculated dose rates fell from the original 8 mrem /hr to the current calculated 1.87 mrem /hr outside the 3'-3'? wall adjoining the hlve gallery. Hence, the IDI team's finding reflects a lack of documentation of engineering judgement snd not an error in design or verification. t i

                                                                                                        ',                                               b,.

DESCRIPT!0J{ [s 1

                         ~

The liquid and vapor vo7u of the v$lume control tank were based upon values obtained from a.Westingh%se WCAP. However, prior to verificatidn of the analysis these ~~ _ values were reMsed by Westinghouse's Final Radiction Analysis ' Manual. ' 3 O N , ,

                                                                        )

y a , RESPONSE , y - p  ; The original WCAP information stated that thh golvme cor.tr 1 tank vapor and; ' liquid phase volumest are apportioned.as 175 Et and,125 ft respectively. The Final bdiation Analysis Manual changeo[sthis to 150 ft ,each. )Since the: j- vapor phase f,ource terms (containing noble ' gases](are far'in; excess ~of-the  % liquid phase source terts, overestimation of the vapor pnase volume fraction resulted in a(q a'ppreciable degr9e of conservatism in the original calculation. - y\ 9 t. f , qct ) *.-<

                                                                         }                                            .

l , r ,) *

                                                                                                                         ,?      *r                           -

e' : . , g i (

                                                                                                                                                             }                      _

(

                                                                                                      /                                                 ,7
                                    \t     y]                  k[;        i
                                                                                                     't            I f

s;

                                 ?t
   ,            ,.. L...-            ,Y                    ~.-,--,-,4                  +-----$---~~----

                                       . . =                      . -                -- - .- - . . .-                           -_          - .

E i D2.5-1 (DEFICIENCY) VOLUME CONTROL TANK SHIELDING ANALYSIS (Cont'd) RESPONSE (Cont'd) The verification process applied in this case must be judged in light of the purpose of the calculation, which was to verify that the shielding design is adequate (i.e., dose rates in Zone II areas do not exceed 2.5 mrem /hr), and not to determine the actual dose rates. Thus, since the original calculation using volume _ fractions overestimating dose rates demonstrated that the design is adequate, subsequently changing volume fractions in the direction leading to substantially lower dose rates only increases the degree of confidence that the shielding design is adequate. Therefore, the verification process was appropriate for the situation. DESCRIPTION The dose rate was calculated at a point 7 feet from the floor; however, the basis for selecting this dose point _was not identified. Although the vapor

           - phase has a significant effect on the resultant dose rate, the dose point was o            not selected to maximize the contribution from the vapor phase.

RESPONSE

The _ general practice is to choose a dose point that is representative of an occupancy position which would be exposed to the highest dose rate. Usually a nominal six foot head height is chosen, but this calculation recognized that dose rates would increase with height since the vapor phase contribution was significantly higher than the liquid phase contribution. Hence, since some people are taller than six feet, a more conservative seven feet height was selected for the dose point. A higher point was not chosen because access is not-provided to these elevations; and consequently, selecting a point to maximize the contribution from the vapor phase was not necessary. Although the above rationale for dose point selection did not appear in the calculation, this represents at most a minor omission because the reason for choosing a point slightly above nominal head height (i.e., additional conservatism) is obvious from consideration of the source-shield-dose point geometry. In addition, given the substantial size of the source and thickness of the shield wall, dose rates would not be expected to change significantly by moving up or down one foot. This was realized by the verifier, who judged that an explanation was not required. p ~ -+ . +- -- ,- wwy .y- , e ,, ~ _ . - , - , -- - - = - - - - - --w-we.e,- -----c--- .--w.,-.. .S--- -m u --- .e - =-,. --e+=m

D2.5-2 (DEFICIENCY) CONTROL OF DESIGN DRAWINGS DESCRIPTION Eight of twenty General Arrangement Drawings contained in a project stick file were not the latest revision. There was no indication that the drawings were to be used for information only or that they had been superseded by a later revision.

RESPONSE

The civil engineering & design drawing stick files were in the process of being updated to ensure that only the latest applicable drawing revisions were present. However, this process was not complete at the time of the IDI with the result that several out-of-date drawings were present in the file. This would not affect design since engineering and design personnel are instructed to consult the Drawing Control Log (DCL) which gives the latest revision and lists all outstanding changes for the drawings. All unused or seldom used stick files have been discarded to prevent use of uncontrolled drawings. Each discipline will maintain a controlled stick file of its own drawings. Only 2 controlled files of project drawings will be maintained. The Project Designer controls the distribution of all drawings and reviews the stick files on a periodic basis. Ebasco Quality Assurance performs quarterly audits to ensure that the files are being maintained up to date. l l l I l l

+ D2.5-3 (DEFICIENCY) POST-LOCA SHIELDING DESIGN REVIEW i DESCRIPTION Calculation TMI-026 pertains to a shielding design review of areas outside containment. In this calculation, the source term derived for the i contribution of cesium isotopes is based upon values obtained from Combustion Engineering's System 80 standard design, scaled to the Shearon Harris power output. A formal calculation does not exist to document how the source terms were derived. These source terms are used in the dose rate calculations for long-term radiation sources and are reported in the FSAR on Table 12.2.1-28 in the form of gamma source strengths. The IDI team was informed that Combustion Engineering input was used because Westinghouse does not provide this infomation in their Radiation Analysis Manual. The IDI team was informed by Ebasco that Westinghouse was not consulted as to the derivation or use of the cesium source terms.

RESPONSE

The finding misinterprets how the source terms were created. The Westinghouse source terms, taken from their Radiation Analysis Design Manual, present energy-grouped source data for assumed release from the core of 100% noble gases, 50% halogens, and 1% others. The cesium is included in the last category. These release. percentages are in accordance with NRC requirements j as stated in NUREG-0578, NUREG-0737, etc. The appropriate Westinghouse tables , are included in the subject calculation, which clearly states the release i percentages by class of nuclide. Sir;ce there was considerable discussion in the nuclear community directly following TMI about the possibility of increasing the assumed percentage of cesium release to some amount above the required 1%, the engineer performing the calculation decided to include an extra 20% cesium above the 1% already present in the Westinghouse source terms. This was not required by regulation. The Westinghouse source term data, however, were in an energy group format, and not in an isotopic activity format. This precluded separating out the contribution from cesium. The engineer then decided to use information at hand, and supplement the 1% cesium data from Westinghouse with appropriately modified Combustion Engineering data which was by nuclide. In the judgement of the shielding engineer and verifier, obtaining the optional, additional 20% cesium source terms in this way was perfectly reasonable, and perfectly justifiable on the basis of nuclear engineering considerations. This was especially true in light of the use to which it was put; resulting in conservative dose rates. It is true, however, that the calculation does not identify how the additional 20% cesium source terms are obtained, and the verifier did not note this. This oversight, however, is one of form rather than substance since the terms were reasonably obtained and went beyond what was required in any event. Since they are not required, reference to the additional 20% cesium will be removed from the FSAR. Since the inspection, the applied physics group has instituted a program of instruction to increase engineer's awareness of QA requirements for calculations. The documentation of assumptions has been emphasized. w - __-_,.,.g..,wm .m.

D2.5-3 (DEFICIENCY) POST-LOCA SHIELDING DESIGN REVIEW (Cont'd) 1 DESCRIPTION The minimum quantity of water available from the refueling water storage tank for dilution of the primary coolant is overestimated at 50,894 cubic feet (approximately 380,700 gallons). A mechanical discipline calculation determined that 213,000 gallons was the minimum volume available for injection. Although the originator indicated in the calculation that the value was obtained from another analyst's calculation, the IDI team was informed that a calculation did not exist to support the value of 50,894 cubic feet.

RESPONSE

Calculation 026 assumed a higher volume of dilution water than the minimum amount of water available for injection. Hence, the specific activity of the diluted reactor coolant was understated, as were dose rates and doses calculated using this source. This calculation has been superseded by a series of other calculations correcting all discrepancies, and the Tt1I access results are unchanged. Even with underestimated calculated dose rates, access to areas in the RAB in the-vicinity of the CVCS system is precluded. Thus, a recalculation would not affect this conclusion since the increased dose rate increases and reinforces the original conclusion. Applied physics engineers have had further instructions concerning the documentation of input data. DESCRIPTION The free volume of the containment available for dilution of the core release i- is overestimated. The calculation states that the containment free volume is ! 7.1E(+10) cubic centimeters; however, 6.41E(+10) cubic centimeters corresponds !- to the free volume calculated by the civil discipline for use in the ( containment pressure and temperature analysis. The source for the value used in the calculation was not referenced and the IDI team was informed that a calculation did not exist which established the value. This discrepancy results in a 10% underestimation of the atmospheric source strength.

RESPONSE

An increase in the containment atmosphere source specific activity by 10%, along with the correction of other discrepancies, does not alter any of the - conclusions reached on the basis of this calculation with respect to access to various areas following a TMI-type accident. This is confirmed by calculations 041, 042 and 043. As noted previously, a departmental training program has been initiated to prevent recurrence of this type of documentation error. L I 1

ll D2.5-3 (DEFICIENCY) POST-LOCA SHIELDING DESIGN REVIEW (Cont'd) DESCRIPTION In determining the gamma source strengths at various energy ranges for undiluted primary coolant, a computational error was made such that the resulting gamma source strengths are underestimated. In the conversion process, the thermal output in watts was divided by the reactor coolant system volume in cubic centimeters. A value of 11.24 watts / cubic centimeter was used; however, the value should have been 11.5 watts / cubic centimeter. This discrepancy results in approximately a 2% error in all of the gamma source strengths for undfluted reactor coolant. , RESPONSE i This part of the calculation converted Westinghouse source term units of

MeV/(w-sec) to gammas /(cM-sec), and included a calculational inaccuracy 4

underestimating the primary coolant source terms by about 2%. However, neither the Westinghouse source terms, the containment volume, nor any of the i other inputs and assumptions are known to anywhere near 2% accuracy. Even F with perfect data, no shielding calculational method could produce results close to this accuracy suggested by the IDI team. Thus, this very slight

inaccuracy in converting units is hardly significant, and in no way affects

  . any of the conclusions reached as a result of this calculation. This does not
         ~

excuse the error, but the verification process can not be faulted for not flagging an inconsequential error of such diminutive magnitude. j DESCRIPTION

In calculating the dose rate inside and outside the cubicle containing the volume control tank of the chemical and volume control system, the thickness of the tank used was incorrect. The calculation assumed that the volume control tank has a steel thickness of 2.54 centimeters. The vendor drawing of-the tank indicates that the steel thickness of the tank is 0.250 inches or 0.635 centimeters. This discrepancy has a significant impact on the resultant dose rate for the gamma energy levels of interest.

D2.5-3 (DEFICIENCY) POST-LOCA SHIELDING DESIGN REVIEW (Cont'd)

RESPONSE

The calculation assumed a heavier tank wall thickness than actually exists. However, as previously stated the purpose of the calculation was to determine where safe access would be possible following a TMI-type accident, and not to determine absolute dose rate values. Calculated dose rates outside this cubicle using a one inch thick volume control tank shell are so high as to preclude access in the area. Even with the assumed heavier wall thickness,  ! the correct conclusion was reached. Due to the high dose rates, it was determined that the tank could not be used following an accident. This rendered the calculation moot. This was known to the verifier who properly assessed that the conclusion was correct. As previously noted, applied physics is conducting training with regards to the documentation of assumptions. DESCRIPTION The heights of the liquid and vapor phases in the volume control tank were based upon a liquid volume of 125 cubic feet and a vapor volume of 175 cubic feet. These volumes should have been 150 cubic feet based upon Westinghouse's Radiation Analysis Manual. It appears that these volumes were taken from calculation VC-RS-YT-51.1, or a superseded Westinghouse WCAP. However, this misstatement does not appear to have an effect on the calculation, because the analyst assumed that the tank was completely filled with reactor coolant. Because of the self-shielding effect of water, this assumption may not result in a conservative dose rate. The IDI team noted that the contribution from the vapor phase was shown to have a more significant effect than the liquid phase in another shielding calculation.

RESPONSE

The partition factors cited above apply to normal operating conditions only, and do not apply to post-LOCA conditions with significantly different pressures and temperatures. Under reduced pressure, any noble gases contained in the reactor coolant source may escape before reaching the volume control tank. Under elevated pressure, the gas may remain entrained in the liquid and not separate out at all. Hence, the normal operation partition factor (discussed further in the response to deficiency D2.5-1) has no bearing on this calculation, particularly since the calculator, recognizing this fact, exercised his judgement in deciding to model the tank as completely full of reactor coolant water.

L l D2.5-3 (DEFICIENCY) POST-LOCA SHIELDING DESIGN REVIEW (Cont'd) RESPONSE (Cont'd) The coment on increased self-shielding from the water only tells part of the story. The assumption of no partitioning brings the noble gases closer to the dose point (a person outside the cubicle), and results in less effective shielding by the cubicle walls than if partitioning is assumed. In the latter case, the noble gases (located at the top of the tank) would be further from the dose point (located at a lower elevation) and the slant path from the gases to the dose point through the concrete wall would include more effective shielding than would a less oblique path. Thus, due to these competing effects, it is far from clear which assumption (all water or partition) would actually give the higher dose rate. Since no detennination has ever been made concerning the issue of possible partition, the engineer's assumption of all water is reasonable. In any case, the calculated dose rates outside the volume control tank cubicle post-LOCA were found to be so high that access is precluded. Any possible increase in dose rates from a changed calculational assumpton would only reinforce this conclusion. Once again, an accurate determination of dose rates was not the object of this calculation. Also, repeating an earlier point (as noted to the IDI team), the volume control tank is not going to be used during a TMI-type accident. DESCRIPTION Instances were identified where the checker / verifier (same individual) indicated that additional information was required to complete the documentation or to add clarity; however, the information was not provided. In many instances, only the results were listed without supporting calculations to document how the results were obtained. A specific example is in the calculation of the dose rates through various thicknesses of concrete and the extrapolation for the thickness of concrete walls intervening between the control room and the containment atmosphere. The checker / verifier requested that a sketch be provided of the configuration including dimensional data. This infonnation was not provided. In the same calculation the checker / verifier questioned the value used for the radius of the containment. The calculation assumed a containment radius of 130 feet, which is approximately twice the actual radius. Other similar instances were observed by the IDI team.

RESPONSE

l Taking the last item first, the IDI team is incorrect in asserting that the calculation assumed a containment radius of 130 ft. Rather, the correct value of approximately 65 ft was used. This can be seen from the calculation l sheets. i i l l l l [

  .                   -.                                  =-      ._     .-  .         .

M D2.5-3 (DEFICIENCY) POST-LOCA SHIELDING DESIGN REVIEW (Cont'd) RESPONSE (Cont'd) The checker's comments were incorporated wherever deemed necessary, but were not subsequently initialed because of the discontinuance of the idea of considering the CVCS system accessible (FSAR Section 12.3). Most of the

checker's comments refer not to substance but to form such as providing references, clarifications, etc, and in no way impact on results. The originator had resolved the verifier's comments, resulting in the signoff of the calculation.

Worksheets included in the calculation present the calculational parameters required by the Rockwell shielding method to determine dose rates. It is not true that "in many instances, only the results were listed without supporting calculations to document how the results were obtained". The calculations are there, but often separated from the summaries of results by intervening pages. Perhaps it would have been clearer to reference the appropriate pages, but internal referencing (where one page of a calculation references another within the same calculation) is not mandatory. 4 The exaaple given by the IDI team, that the control room calculations are not documented, is likewise incorrect. Several pages of Rockwell method worksheets, plus a graph, are included in the calculation. A sketch of the geometry might have been useful, but was not necessary. DESCRIPTION The IDI team found that assumptions were not consistently justified when they , were introduced into the calculation. Likewise, the IDI team found that source documents for base data were not consistently identified in the i calculations. For example, in calculating the radiation in the vicinity of air handling unit AH-5, sources and magnitude of that radiation were identified; however, a reference was not provided. The verifier requested that a reference be provided to identify where these numbers were derived. The IDI team confinned that the information came from unchecked and unverified calculations performed with a hand calculator. This unsubstantiated

information was recorded on general arrangement drawings and called dose l maps. The IDI team observed examples of these dose maps and confinned that they were not prepared or controlled in accordance with established company i procedures. Another instance of this same type occurred in calculating the

! radiation dose from equipment located near air handling unit AH-10. In addition to these, other instances were observed by the IDI team. l l

D2.5-3 (DEFICIENCY) POST-LOCA SHIELDING DESIGN REVIEW (Cont'd)

RESPONSE

The response to the first part of the finding is the same as in the last response; i.e., that omissions of some references constituted an issue of form rather than substance. There is no indication that any of the non-referenced material is incorrect. In the cases of dose determination in the areas near air handling units AH-5 and AH-10, cited in the finding, as well as other unspecified instances, the calculation first determined dose rates (using the Rockwell method) from pipes in the vicinity containing reactor coolant fluid, then summed them to give the total dose rate at each time of interest. The arithmetic going into the summation was not presented, as it merely consisted of addition. The piping drawings showing the detailed piping arrangements are properly referenced in the calculations. The dose maps produced as a result of this process were clearly marked

 " Preliminary" and were intended for information only and not for final design work. The final maps were based on a different calculation, and appear in the FSAR. These maps are controlled in accordance with established company procedures governing FSAR modifications.

DESCRIPTION Contrary to Shearon Harris Nuclear Project requirements, the originator did not clearly state the justifications and source of assumptions and base data used in the calculation; the checker initialled and dated each page of the original calculation without ensuring that necessary corrections and additions had been made or requiring that a statement be added indicating that known errors were acceptable because there was no significant impact on the final result; and the verifier of the calculation did not verify that inputs were correctly selected and that assumptions were adequately described and reasonable. Contrary to Shearon Harris Nuclear Project FSAR conunitment, design control has not been maintained in that applicable regulatory requirements for verification of design analyses were not correctly applied to ensure that personnel can perforri necessary post-accident operations in vital areas following a LOCA. I l l l \ l l l i

D2.5-3 (DEFICIENCY) POST-LOCA SHIELDING DESIGN REVIEW (Cont'd)

RESPONSE

The possible shortcomings of the verification process in this calculation are overstated. The verifier was satisfied that the results were proper and consistent with the goals of the calculation, which were not to determine dose rates but to determine whether safe access would be possible to various areas of the plant following a TMI-type accident. The cited discrepancies in the sump water and containment atmosphere source dilution volumes both reduced calculated dose rates. However, dose rates were still so high that access would be precluded in the vicinity of the CYCS system. Increasing the dose rates would only strengthen this conclusion. This calculation was superseded by later calculations 041, 042, and 043 which corrected all of the noted possible shortcomings. These calculations have confirmed the original results. SUPEARY The IDI team was concerned that the errors described in this deficiency may affect the integrated doses used throughout the plant because the dose caused by recirculating fluids is to be included in the total dose to equipment when establishing radiation environment. In addition, the IDI team stated that these errors were found in other radiation protection calculations perfomed by the applied physics discipline. It should be noted that TMI access, not equipment dose, was the subject of calculation 026. This calculation has been superseded and the conclusions confirmed by others which corrected all of the noted, possible shortcomings. Many of the comments pertain to omitted references, etc, rather than to the correctness of the calculation for its intended use. The " errors" cited are not as extensive or significant as alleged and have been evaluated with no design impact. I l

D2.5-4 (DEFICIENCY) GAMMA RADIATION SOURCE STRENGTH ASSUMPTIONS FOR EQUIPMENT QUALIFICATION

 ,           DESCRIPTION Although calculation 027 initially states that the containment source terms
;            for a design basis accident are to be based upon a core release of 100 percent of the noble gases, 50 percent of the halogens, and 1 percent of the solids, a halogen release of only 25 percent was assumed and used within the calculation.

RESPONSE

The inspection finding is correct. The calculation originator mistakenly assumed that half of the 50% halogen release imediately plated out in the

 +

reactor coolant system and was, therefore, not part of the containment atmosphere source. term. This calculation has been superseded by calculations 041, 042, 045, 046 and 049 which use the correct halogen data. ^ DESCRIPTION

           - Although Westinghouse provided source terms in the Radiatbn Analysis Manual which were consistent with the core release required, the analyst misinterpreted the basis of the halogen source term. The analyst appears to have assumed that the source terms released were 100 percent of the noble gases and halogens instead of 100 percent of the noble gases and 50 percent of j           the ' halogens. Because of this error, the analyst divided the Westinghouse

! halogen inventory by _4 to obtain what he thought was a 25 percent halogen inventory; instead he obtained only a 12.5 percent halogen inventory.

RESPONSE

4 The IDI team is correct. Both the analyst and the verifier misinterpreted the

          . Westinghouse data, thinking that is was for 100% noble gases and 100% halogens rather than just 50% of the latter. Combined with the earlier factor of 2 mistake noted in the previous description, this resulted in consideration of only 12.5% of the halogens in the containment atmosphere rather than the l            proper 50% required by regulation.

l -Correcting of this mistake would have resulted in the calculation of higher-I- equipment qualification doses in the containment. However, a new analysis was performed based on a more sophisticated shielding technique, including all significant shield walls in the containment, as well as using a mechanistic model for radionuclide transport which will be described in the FSAR. The results of this new calculation show that the equipment qualification doses in j the containment are actually lower than previously calculated. Thus, there < will be no effect on equipment already qualified in containment. The increase in the halogen source terms did not have a larger offsetting effect on dose rates since the main contributor to containment dose rates is the noble gas source, which remained unchanged. Neither the analyst nor the verifier suspected an error in the halogen source tems in the original calculation since the final calculated doses were well within the range of values determined for similar sized plants on other projects. Thus, the many conservatisms present in the original calculation served to hide from the verifier an incorrect input that would otherwise have been identified. . 1

D2.5-4 (DEFICIENCY) GAMMA RADIATION SOURCE STRENGTH ASSUMPTICNS FOR EQUIPMENT QUALIFICATION (Cont'd) RESPONSE (Cont'd) Although the error had no consequence, the applied physics Chief Engineer and radiation protection group supervisor have instituted a program to minimize the chance of recurrance. The program has two parts: intensified training in procedure for analysts and verifiers; and, systematic review of a sample of past calculations to determine the extent (if any) of required corrective actions. This should both increase the degree of confidence in the results of future calculations and identify any shortcomings in past verifications, if they exist. 4 V

                                                 . - . . . - . , n,,     ,,.- - -

l l D2.5-5 (DEFICIENCY) EQUIPMENT QUALIFICATION BETA DOSE DESCRIPTION f Calculation EQ-027 contained an incorrect assumption which reduced the post-LOCA integrated beta dose by 30 percent. Specifically, the calculation  ! determined the integrated beta dose directly from the beta integrated energy fluxes provided by Westinghouse. However, the analyst concluded that - ,

                 -Westinghouse derived the beta energy fluxes while assuming no credit for the                                                                               !

, washing effect of containment spray. By comparing the washed and unwashed beta source terms calculated in Appendix D of NUREG 0588, the analyst

concluded that the beta dose could be reduced by at least 30 percent.

4

RESPONSE

The assumption of 30% washout is not incorrect, but rather is unsubstantiated for the Shearon Harris Nuclear Project. Calculations done for other similar plants, as well as the NRC's NUREG-0588 calculations for a generic plant, show that the 30% figure is quite reasonable. The shortcoming, then, is not in fact, but in failure to apply a plant specific model. The analyst and the

verifier accepted the 30% reduction as a reasonable engineering judgement.

Reanalysis of the entire equipment qualification calculation has shown that when a mechanistic model is applied to the transport of radionuclides in the containment following a TMI type accident, the beta dose is reduced substantially below what was originally determined. For example, the one year beta air dose at the containment centerline decreased from 3.7 x 108 to less than 2.0 x 108 rads. Thus, the original calculation was quite conservative. SUP9lARY The IDI team stated that this deficiency will require reanalysis and documentation changes. Although the IDI team believes the increase in the beta dose will not result in hardware changes, the Shearon Harris Nuclear Project has committed to qualify electrical equipment to an integrated dose [ which corresponds to the sum of the gama and beta doses. As mentioned in the last response, a revised calculation was performed, reducing the doses below those originally determined; hence there is no effect on equipment qualification in containment. However, even merely eliminating . credit for washout in the original calculation would have no significant effect on equipment qualification.

  ~ ~ m            ,-   --
                           ,.o.. - ,, - , . - , .   --   --
                                                            , - - , . -    ,..,-.o.,     ,, - - - ,.,,--~.---3--,--.,-------.,--v.re------a                          er-w- -

02.5-5 (DEFICIENCY) EQUIPMENT QUALIFICATION BETA DOSE (Cont'd)

SUMMARY

(Cont'd) In addition, the qualification dose is not merely a simple sum of the gamma and beta doses as determined by this calculation, since the latter applies only to a free-air situation with no attenuation of beta radiation. The proper beta dose to add to the gamma dose considers attenuation provided by coverings (i.e., cases, boxes, insulation, sheaths, etc) present on all safety-related equipment in the containment. Since only approximately 10 mils of steel or 70 mils of elastomer is sufficient to reduce beta dose by a factor of 100, the beta qualification dose is typically less than 10% of the gamma dose. Hence, a 30% increase in free air beta dose would result in only about, at most, a 3% increase in total dose. The FSAR will be amended to reflect the use of a mechanistic washout model in the revised calculations. i 4 I

, I D2.5-6 (DEFICIENCY) INTEGRATED DOSE ANALYSIS FOR EQUIPMENT QUALIFICATION DESCRIPTION t In calculating source strengths, a value was used for the containment free volume which was not conservative. Specifically, a value of 6.51E(+10) cubic i centimeters was used instead of 6.41E(+10) cubic centimeters. The latter value corresponds to the free volume calculated by the civil discipline for i use in the containment pressure and. temperature analysis. The source of the , larger containment volume was not identified. This discrepancy results in a

1.5 percent underestimation of the radiation source strengths.

RESPONSE

The conta3 ment free volume is improperly cited as a deficiency. The value of 6.51 x 10 ucm 3 used in calculation 027 represents the most probable value, and is used in several places in Chapter 6 of the FSAR for subcompartment 1 analysis. The value of 6.41 x 1010cm3 represents a minimum value (as ' stated in the Civil Department's Calculation Procedure E-54 for the Shearon Harris Nuclear maximum of 6.64 xProjegt),3, 10iucm whereas 6.51 x 110 cm3 uwith represents an uncertainty an range of 3.427%, yi average of these two values. Hence, the radiation spect fic source strengths . were not underestimated. Applied physics engineers have received further 1 instruction concerning the documentation of input data.

DESCRIPTION 3'

The decay constant for Kr-89 was incorrectly calculated. Specifically, a value of 3.659E(-2)/sec was calculated and used in the analysis when a value of 3.629E(-3)/sec should have been used based upon a half-life of 3.182 minutes.

RESPONSE

The inspection finding is correct. This resulted from a transcription error, and has no consequence with respect to equipment qualification dose since the isotope decays so rapidly. As previously stated, a recalculation was perfomed using corrected values and produced significantly lower containment

doses than the original calculations. We have initiated additional training l

to prevent recurrence of this type of discrepancy.

  ~

DESCRIPTION The decay constant for Xe-135m was incorrectly calculated. Specifically, a value of 1.8E(-1)/sec was calculated and used in the analysis when a value of 7.38E(-4)/sec should have been used based upon a half-life of 15.651 minutes.

RESPONSE

The finding is correct. However, the half-life is quite short, so Xe-135m contributes little to the overall equipment qualification dose, and a 4 recalculation produced significantly lower doses throughout the containment.  : We have initiated additional training to prevent recurrence of this type of discrepancy. 1

    - - . -           , , , - - - - -             - . - . _ , -        , , - , . , . , - , - . . , , - - - - . . , .                         , , - . - . - - - . , , . . - - - - .                 - , - , . ~ - - - -

D2.5-6 (DEFICIENCY) INTEGRATED DOSE ANALYSIS FOR EQUIPMENT QUALIFICATION (Cont'd) DESCRIPTION For isotopes Kr-89 and Xe-138, the gamma radiation energy levels emitted and . their corresponding intensities were not complete. Although the analyst had available to him an Ebasco-prepared computer library of energies and probabilities by isotope, undocumented values were used.

RESPONSE

The originator of the calculation took the data from the Radiological Handbook, although it was not referenced. The possible omission of some levels would have a negligible effect on dose since these isotopes have such short half-lives (Kr-89, 3.182 min; Xe-138,17.5 min). Further, a recalculation demonstrated a reduction of all the containment doses. More recent vintage calculations generally rely on computer programs to convert from isotopic activity to a gamma energy level source, eliminating the potential for this sort of error. As previously stated, applied physics is conducting training regarding documentation of data sources and has provided adequate references in the revised calculations. DESCRIPTION A computational error was found in the source strength associated with a 0.305 MeV gamma for Kr-85m. The gamma source strength should have been 1.625E(+6) gammas /cc-sec instead of 1.25E(+6) calculated.

RESPONSE

The inspection finding is correct. This error is slight, the gamma energy is relatively low, and the half-life is short (4.4 hr). Hence, no significant effect is expected, and a total recalculation using corrected values resulted in significantly lowered containment doses. In all the preceding errors found in source term calculations, the effects are inconsequential, as these isotopes are unimportant contributors to the total dose. Almost universal use of computer programs to do this sort of source term generation has precluded this type of error from other, later calculations. Thus, it is not expected that source term errors are systematic. - DESCRIPTION A rigorous analysis was not performed for the contribution from daughter products to the source terms. Instead of performing an analysis, an assumption was used that the contributions from the daughter products will increase the integrated dose by 30 percent. Since the contribution is significant and since the emphasis in NUREG-0588 is on mechanistic and analytical treatment in such areas as activity redistribution and spray removal, explicit treatment of daughter products should have been included.

i l l I i D2.5-6 (DEFICIENCY) INTEGRATED DOSE ANALYSIS FOR EQUIPMENT QUALIFICATION I

!.     (Cont'd) i     RESPONSE I

A complete reanalysis of the equipment qualification doses using a mechanistic nodel to account for radiation from daughter product decay has been performed, and has established that the contribution in containment is less than the 30% 4 previously assumed (and cited as typical in NUREG-0588). Thus, the original calculation was conservative in its assumption of 30% daughter product radiation contribution, and the judgements of the analyst and verifier have been confimed. As previously noted, applied physics is conducting training with regard to the documentation of judgements. DESCRIPTION , The integrated gamma dose within various defined volumes of the containment

was calculated based upon the contribution from the containment atmosphere and i the containment sump. For these volumes, the contribution from immersion in the containment atmosphere was assumed to be equal to the product of the integrated dose calculated at the center of point of the containment and a volume fraction corresponding to the ratio of the defined volume to the total containment volume. This assumption is not justified by the equations used.

RESPONSE

l The methodology was not correct. The revised calculation models the containment in three-dimensional detail, using the Span-4 code. This model , explicitly represents sub-compartment structures, and replaces the original calculational method. It yields lower doses in containment than the original calculation; thus no impact on qualification of equipment results. DESCRIPTION i In calculating the airborne dose rate within the reactor auxiliary building, a , . gamma source strength was used that did not include the contribution from decay of daughter products.

RESPONSE

The reanalysis replacing calculation 027 includes calculation of the contribution of decay daughter radiation to the airborne dose within the RAB. See above. 4 DESCRIPTION Instances were identified where the checker of the calculation identified

computational errors. These errors were not consistently corrected in the calculation nor were statements provided to indicate that the errors were not significant with respect to the calculation. Numerous examples were observed by the IDI team.

I

         ,,--y e--,-.,,- e -, , - n.m - .n.,-,m-  7,,     . -y eem,_g, am--- ..e,, m-,.mmn,,,,. -u,-,c,g,,----,-+-.,+ g ,n.-- - ~

l D2.5-6 (DEFICIENCY) INTEGRATED DOSE ANALYSIS FOR EQUIPMENT QUALIFICATION (Cont'd)

RESPONSE

j The original calculation, 027, has been superseded by a reanalysis which corrected all identified errors, and produced lower doses in the containment than before. In addition, the new calculation conservatively assumed that all of the containment leakage enters the RAB. l The numerical problems found in the original calculation, consisting of nearly 130 pages of hand computations, are not typical of other calculations, which ,

make much greater use of computer codes. In addition, an intensified training '

4 program was instituted in the applied physics department to heighten awareness of the calculation / verification procedure, and should greatly reduce the

chances of recurrance of verifier comments not being addressed.

DESCRIPTION - The IDI team found that assumptions were not consistently justified when they were introduced into the calculation. Likewise, the IDI team found that source documents for base data were not consistently identified in the ! calculation. For example, in calculating the gamma dose in the Reactor Auxiliary Building information was used from " dose maps prepared for post-TMI

;            shielding review". Although the verifier of the calculation requested that a reference be provided to identify the source of this information, a specific                                                                            !

reference was not provided. The IDI team confirmed that the information came from unchecked and unverified calculations performed with a hand calculator. This unsubstantiated information was recorded on general arrangement drawings and called dose maps. The IDI team observed examples of these dose maps and i confirmed that they were not prepared or controlled in accordance with ' established company procedures. An exariple of an assumption that was not justified occurred in calculating the post-LOCA activity in the Reactor Auxiliary Building. Specifically, the calculation assumed that the leakage from the containment is unifomly distributed in four quadrants around the  ; containment structure and seven elevations. These assumptions reduce the resulting source strengths significantly (i.e., by a factor of 3.57E-2). The IDI team was informed that no consideration was given to potential leakage i paths (such as containment penetrations) and the presence of those paths within a given area.

RESPONSE

;-                                                                                                                                                                   +

The dose map coment above is identical to the one made in Deficiency D2.5-3.

  .          The response is the same; the doses recorded on the dose maps were constructed i            from tables of calculated doses from indvidual pipes, and piping arrangements taken from piping drawings. The only part not appearing explicitly in the
!            calculation is the arithmetic in the sumation for each location.

l The new calculation replacing 027 corrects all the errors, provides proper

            - references, etc. Further, it conservatively assumes that all the containment                                                                           i 4

leakage enters the RAB. m-- . - - - ,_e,m, .--em-,,,-,.,-- - e ,. . . ~ ~ - ,--n -r , ,. -~. .r--.,..,--------, r..n.~ ----,-----,--,.-,-,,,n. -

                                                                                 }

D2.5-7 (DEFICIENCY) RADIATION DOSE IN EQUIPMENT QUALIFICATION ZONE R-6 0F RAB DESCRIPTION In calculating source strengths (in calculation 039), a value was used for the containment free volume which was not conservative. Specifically, a value of 6.51E(+10) cubic centimeters was used instead of 6.41E(+10) cubic centimeters.. The latter value corresponds to the free volume calculated by the civil discipline for use in the containment pressure and temperature analysis. This discrepancy results in a 1.5 percent underestimation of the radiation source strengths. This error was also found in another calculation and is described in Deficiency D2.5-6.

RESPONSE

This has been fully addressed in the response to Deficiency D2.5-6. The value used for the containment free volume is the most probable value, which appears in the FSAR. Hence, the radiation specific source strength was not under-estimated. DESCRIPTION An assumption was made concerning the leakage rate from the containment to adjoining buildings and to the atmosphere without justification. The analysis assumed that half of leakage occurs to atmosphere and that the res., leaks to adjoining buildings. In addition, the analysis assumes that the leakage to adjoining buildings occurs equally (i.e., the atmosphere within four quadrants of containment leak equally to adjoining buildings and equally to seven elevations within those buildings). These assumptions reduce the resulting source strengths significantly (i.e., by a factor of 1.78E-2). The IDI team was informed that no consideration was given to potential leakage paths and the presence of those paths within a given area. R_ESPONSE There is no known methodology available concerning the apportionment of containment atmosphere leakage through the griad potential pathways to surrounding buildings and the environment. This necessitates the adoption of simplifying assumptions concerning leakage to the RAB. One such assumption is equipartition of leakage, which assumes that leakage is uniformly distributed in the absence of RAB HVAC operability. The attribution of the leakage factor of 1.78 x 10-2 constitutes the leakage fraction into the R6 area. This is a conservative assumption, since, in effect, it is tantamount to saying that the R6 volume was 1.78% of the total RAB volume, when it is in fact substantially smaller. The analyst made a reasonable engineering judgement in this matter, and the verifier concurred. The replacement calculations assume 100% leakage into the RAB.

02.5-7 (DEFICIENCY) RADIATION DOSE IN EQUIPMENT QUALIFICATION ZONE R-6 0F RAB (Cont'd) DESCRIPTION An incorrect source term was used for an instantaneous release from the fuel to the atmosphere. Specifically, a halogen release of 12.5 percent was mistakenly used instead of the 50 percent required. This same error was also found in another calculation and is described in Deficiency D2.5-4

RESPONSE

This is discussed in the response to Deficiency D2.5-4. The calculation has been corrected using the proper value for halogen release from the core following a TMI-type accident. DESCRIPTION For isotopes Kr-89 and Xe-138, the gamma radiation energy levels emitted and their corresponding intensities were not complete. Although the analyst had available to him an Ebasco-prepared computer library of energies and probabilities by isotope, undocumented values were used. This same error was also found in another calculation and is described in Deficiency D2.5-6.

RESPONSE

This is discussed in the response to Deficiency D2.5-6. The effect on the conclusions of the calculation is negligible. This, and other cited numerical or nuclear data errors, are largely precluded from most other calculations, which -aly on computer codes to manipulate data and perform calculations. DESCRlr u0N In calculating the airborne dose rate within the reactor auxiliary building, a gamma source strength was used that did not include the contribution from the decay of daughter products. This same error was also found in another calculation and is described in Deficiency D2.5-6.

RESPONSE

This has been addressed in the response to Deficiency D2.5-6. A reanalysis explicitly includes determination of the contribution of decay daughter radiation to the airborne dose. DESCRIPTION An incorrect time-dependent equation was used to calculate activity of various isotopes inside the reactor auxiliary building. The originator of the calculation derived a first order ordinary differential equation to describe the production and removal of activity in zone 6 of the reactor auxiliary

D2.5-7 (DEFICIENCY) RADIATION DOSE IN EQUIPMENT QUALIFICATION ZONE R-6 0F RAB (Cont'd) DESCRIPTION (Cont'd) building. The originator of the calculation did not record the steps performed to arrive at the solution and simply identified the solution; therefore, the IDI team is unable to identify the specific error (s) made.- However, the IDI team did note that the originator stated that the solution was based upon the initial condition that the containment activity is zero at time zero. This is an inconsistent assumption because the release to the containment atmosphere should have been assumed to be instantaneous based upon Shearon Harris Nuclear Project's FSAR commitments.

RESPONSE

The form of. the equation is correct, although an error was made in its coefficients. The text of the calculation inadvertently stated that the containment activity is zero at time zero, intending instead to state that the auxiliary buf1 ding activity is zero at time zero. However, the equation in the text clearly shows that the containment activity is assumed non-zero at time zero. The calculation was redone using the proper coefficients, and outlines the derivation and assumptions in greater detail. The previous dose rates have not been adversely affected. DESCRIPTION An unstated assumption was made that the integrated dose to equipment for the first 24 hours can be calculated by assuming a linear relationship between time zero and 24 hours later.. However, this assumption underestimates the contribution of isotopes with a short half-life. Justification was not provided for assuming the longer time interval although smaller time intervals would have resulted in a higher integrated dose.

RESPONSE

This assumption was rechecked for times up to one day after the start of an accident, and found to be a good approximation if the dose at 0.1 day is used i as the first point and time is incremented in 0.1 day steps. Thus, the

original engineering judgement was upheld by a more detailed calculation.

Further, this assumption does not underestimate the dose contribution of . short-lived isotopes since these short-lived isotopes contribute little to the l total dose. The superseding calculation does not rely on this methodology. i k i i l l

D2.6-1 (DEFICIENCY) INSTALLATION OF CHARGING PUf!P ROOM AIR HANDLING UNITS DESCRIPTION The installation of Air Handling Units AH-9 (1 A-SA) and AH-9 (18-S8) was not in accordance with seismic qualification report CCL-A-198-79-02. Design drawings did not indicate a specified torque value for A-325 anchor bolts supporting the equi,pment. Site procedure WP-105 specifies bolts to be torqued to a " Snug Tight condition in. lieu of a specified torque value. In this case, neither the design drawing nor the vendor manual indicated any values required. The seismic analysis indicated torques to be in accordance with AISC (7th ED), Table 1.23.5. RESPONSE-The Seismic Qualification Report was reviewed and we found that none of the anchor bolts experience uplift in a seismic event. Since shear is the primary force considered in the analysis, bolt preload is only critical to assure contact between nut and bolted surface as well as to assure the nut remains in place. This was. confirmed by the consultant who performed the analysis. The use of " snug tight" as defined by AISC, 7th edition is adequate in this application. Seismic equipment in general is installed per requirements listed on vendor drawings. If a specific torque is not specified, site Work Procedure 105 states " snug tight" conditions must be used. This is adequate if no specific torque is used in qualification by analysis or testing. The use of WP-105 for site installation of seismic equipment is considered acceptable. The item identified by the IDI team is considered isolated since vendor drawings as well as qualification reports undergo Ebasco review prior to issuance. Furthermore, additional reviews are done in preparation for Seismic Qualification Review Team evaluation. 1 We have reviewed the installation documentation packages for the subject equipment and determined them to be in compliance with procedural requirements. The package in which the discrepancy was found was not in final form at the time of the inspection because the actual construction work on AH-9 was incomplete. Upon completion of construction, the package will become " final" and contain the required documentation.

.D2.7-1     (DEFICIENCY) NON-SEISMIC PIPING INTERACTION DAMAGE STUDY DESCRIPTION The IDI team reviewed the work performed by Ebasco's mechanical design engineering department in support of the requirements outlined in Regulatory Guide 1.29. In particular, the IDI team reviewed two preliminary calculations SSIS-01 and 02. The IDI team expressed some concern regarding the status of these calculations since they were not checked but were apparently used to formulate the FSAR position on Regulatory Guide 1.29. The IDI team also reviewed Ebasco Procedure E-30 and noted that this procedure requires that preliminary calculations be checked when the calculations are the basis of data included in submittals to public agencies.

Furthermore, since the IDI team was not able to find similar analyses in other disciplines, the IDI team was unable to conclude that the overall plant design was reviewed with respect to the requirements of Regulatory guide 1.29. The IDI team was informed that a multi-discipline seismic II/I walkdown would be performed to verify compliance with commitments to Regulatory Guide 1.29. The IDI team is concerned that Carolina Power & Light has placed too much confidence in the work performed by Ebasco in this area and too much dependence on the seismic II/I walkdown. The IDI team further stated that

  " previous experience at other facilities has indicated that walkdowns are difficult to perform and should not be relied upon to identify and resolve problems late in the construction.

RESPONSE

We have reviewed the IDI team's findings in regard to the adequacy of design activities performed to demonstrate that postulated failures of non-seismic components will not damage safety related equipment durin earthquake and do not completely agree with the IDI team'g a safe shutdown s findings. We do acknowledge that the preliminary calculations SSIS-01 and 02 performed by mechanical engineering should have been completely checked and documented prior to their use in fomulating the basis of data included in the Shearon Harris Nuclear Project FSAR. The IDI team did conclude however that the procedural error associated with these calculations does not appear to be systematic because the IDI team did not find other examples of this error. Subsequent to the IDI team's investigation, an effort to review, check, and document the subject calculations was initiated within the mechanical design engineering department. ' Although the calculations were not checked the individual design changes resulting from the preliminary piping interaction study were checked and verified in accordance with established Company Procedures. The purpose of the piping interaction study was to address previously approved piping layout and classifications so that compliance with Regulatory Guide 1.29 was assured. Furthermore, the criteria utilized to identify and evaluate seismic /non-seismic interactions were conservatively selected (large sphere of influence) so that the preliminary study would yield the worst possible effects during the design phase of the project thereby minimizing the construction impact.

D2.7-1 (DEFICIENCY) NON-SEISMIC PIPING INTERACTION DAMAGE STUDY (Cont'd) RESPONSE (Cont'd) With regard to the IDI team's concern that they were unable to find similar analyses (piping non-seismic interaction stu@) in other disciplines, we wish to point out that the IDI team members did not review the criteria established by those disciplines to address this issue. In addition, the IDI team did not recognize that the in-process design control and design verification methods include consideration of seismic /non-seismic interactions such that these interactions are avoided by physical separation, seismic supports, etc. to the degree practical. Therefore, we see no basis for the IDI team to conclude that a multi-discipline seismic interaction stu@ based upon existing design documents should be perfomed. We also have reviewed the IDI team's concern that CP&L has placed too much dependence on the seismic II/I walkdowns. The Shearon Harris Nuclear Project's commitment to perform a seismic II/I walkdown was first identified during the ACRS sub-committee hearings and later reiterated during the full-committee hearing. This commitment was based upon industry and licensing activities associated in general with systems interactions concerns as outlined in Generic Task Action Plan A-17 (NUREG 0606 Rev 2) and in particular the seismic interaction portion of this concern. Due to the typical activities associated with the construction of nuclear power plants and in particular the field routing and locating of various equipment, it is impossible to assure during the design phase of the plant that all seismic interactions have been evaluated. Therefore, it is only logical that field in-situ walkdowns should be relied upon to verify the design process work perfomed in this area and to assure that field designed systems, structures, and components have not challenged the safety of the plant. Moreover, an integral part of the Seismic II/I Walkdown effort is the development of explicit engineering guidance for construction in a condensed i fomat. This compilation provides additional assurance that adverse interactions will be avoided throughout the remainder of the construction process. 1 bis walkdown is in progress and will be accomplished in two phases. The first is a " preliminary" phase with resolution of all problems identified. The second phase follows final system turnover to ensure that-final constcuction activities have not comprised plant safety. Guidelines have been incorporated into the HPES Instruction Manual. The Seismic II/I Walkdown program, including procedures and schedule, has been reviewed with , the resident NRC inspectors by the Shearon Harris Nuclear Project engineering L management. l l

02.7-2 (DEFICIENCY) SEISMIC II/I INTERACTIONS OF FIELD ROUTED PIPING DESCRIPTION The IDI team observed a non-safety related 1.5-inch instrument air line located in the reactor auxiliary building at elevation 247 feet which was not seismically supported and which could affect the function of safety related equipment located in the immediate vicinity. The IDI team recommended a plant walkdown to identify and correct this and similar items.

RESPONSE

" Regulatory Guide 1.29 Verification Walkdown" has been approved and added to the HPES Section Instruction as Design Guide 7.6.8. This program is in progress. See response to Deficiency 02.7-1.

D2.8-1 (DEFICIENCY) FIELD INSTALLATION TOLERANCES FOR HANGERS i DESCRIPTION .f ( The IDI team observed an ASME III, seismically supported, charging and volume [ control pipe in physical contact with two heating and ventilating dampers on elevation 247 feet of the reactor auxiliary building. This item had not been identified as a nonconformance.

RESPONSE

         "Interdisciplines Clearances Verification Walkdown" has been approved and added to HPES section instructions as Design Guide 7.6.C. to address this and similar items. This program is being conducted concurrently with the Seismic                           '

II/I walkdown which is in progress. See response to Deficiency D2.7-1. I t

                                                              '                                            ._ s

02.9-1 (DEFICIENCY) PUMP VENDOR DRAWING ERROR DESCRIPTION The IDI team reviewed the minimum submergence requirements for the emergency service water pump. The IDI team identified an apparent error on the pump vendor drawing indicating the minimum submergence. The IDI team noted that the pump specification stated that the minimum submergence was to be 6.8 feet above the be11 mouth of the pump, while the pump vendor drawing stated the minimum submergence to be 4.8 feet. Furthermore, the IDI team concluded that Ebasco should have detected this error during the review of the pump vendor's certified drawing in accordance with Ebasco Procedure E-7.

RESPONSE

We have reviewed the IDI team's findings in regard to this item and do not agree completely with its conclusion. Our position is based upon the fact that the "mp vendor drawing correctly indicated the pump minimum submergence to be "6.0 feet above the suction bell lip" of the pump. However, the vendor incorrectly indicated the minimum submergence elevation to be 196.46 feet. The correct elevation corresponding to 6.8 feet about the suction lip of the pump should have been 198.47 feet (the bottom of the pump pit is at elevation 190 feet and the bottom of the suction bellmouth is 20 inches above the bottom of the pump pit). We have corrected the pump vendor drawing indicating the correct elevation corresponding to the minimum submergence for the pump. Also the pump instruction manuals have been revised to reflect the latest pump vendor drawing. With regard to the IDI team's concern that this error should have been detected during the review of the pump vendor's drawing we point out that the correct minimum submergence required was noted on the drawing. Furthermore since the actual minimum water level expected in the pump pit is 204.4 feet as noted on the pump vendor drawing, the reviewer (s) concluded that the minimum submergence requirement was met and therefore did not notice the elevation inconsistency. 1

3 iV y 5 .

^

{ t ,

                                                                                                                                          .d t                                       ';n                              ~
                                                                                                                                                   +

D3.1-f (DEFICIENCY) ALLOWABLE N0?ZLE L0EDSa 3 s

          -DESCRIPTION,           U                                                <d               $

The fo loYrig-two sdress procedures specify inconsistent tabular. data to be used for tha interim qualification of equipment chzzle loads:~ t ~ 1)- EbascoWH Calculation Procedural i -Guideline 4< Procedure SH No. 020, 10/26/85.6 q" e N ( c 'N ,s

2) Ebasco S-H Calculation ProcedurakGuidelines,~ Procedure SH No. 032, 3/8/85/ 3 3

                                                                                                                          ^
                                                                         '                          N RESPONSE                                                        r                              c          ,

We do not concur with the IDI team's interpr etation'. There exists only one procedure which tabulates' data to be used for interim qualification of ~

          . equipment nozzle loads, i.e., Procedure No. 032, ," Allowable Nottle Loads".

The portion of Procedure #20, Revision 0, which is being questioned discusses review of ' equipment specificatier:c,bycthe Stress Analysis Department. It ,

         . notes that vendors typically require nozzles to be subject to zero loads or                                                       a very low loads from the piping system. As part of the review o_f an equipment specification, Stress Analysis reFleWS =the loads in the specification, 'and, if i            they are very low,' attempts to increase these allowable nozzle loads to a i

reasone.ble leiel. The loads in question are ae,rttachment to the, procedure

           -(clea.cly' titled " Review of Pipe Connected Equipts.t Specifications and Vendor                                                         -

Stress' Analysis Reports") and represent loads which are.censidsred to.be i reasonable.; It was never the intent to use these loadsjfor nozzle '

• qualification nor is there any reason to believe they were ever used for that '

purpose. /

                    !=                                                                                             s

. For evaluating calculated nozzle loads, the dialyst codipares the calculated - ! ~ values with the Ebasco criteria (P.e., when vendor allowables arc not i available) on the Nozzle Evaluation Sheet, Q/A Forr.;#622-4_. This criteria , i confoms to that of Guideline #32, the governing pr;ocedure for qualification of equipment nozzle loads. .

          - As a practical matter, the loads of Procedur'e #20, Sev 0, are greater than c            those that vendors have generally found acceptab7e. Therefore, Revision 1 of

Procedure 20 (2/20/85) incorporates the lower load criteria of Procedure #32.

However, there is no generic requirement that the criteria in the two procedures be identical, n

                 ,o          r'                                                    ,
                                                                                                   '/

[. e ) j . m f 4 b m 4 -

03.'l-2 (DEFICIENCY) REGENERATIVE HEAT EXCHANGER N0ZZLE THERMAL DISPLACEMENT DATA DESCRIPTION The IDI team identified a discrepancy between the nozzle thermal data for the CVCS regenerative heat exchanger specified in the Ebasco design specification Appendix A, and the data transmitted by Westinghouse. The IDI team did however confirm that the stress analysis for the piping system considered the latest data provided by Westinghouse.

RESPONSE

We have reviewed the IDI team's finding and acknowledge that the nozzle thermal displacement data contained in Appendix A to the specification was not the latest data received from Westinghouse. However, the design specification was undergoing a major revision while the inspection was in process. Subsequent to the inspection, it was decided that Appendix A to the design specification would be deleted. This decision was based upon the fact that the information presented in the appendix to the specification was redundant to data presented in other documented outputs, e.g., vendor drawings, specifications, etc. Notwithstanding the above, it was pointed out to the IDI team that the regenerative heat exchanger is modeled into the pipe stress analysis so that the nozzle displacements are derived by actual calculation. The subject specification has since been revised to reflect these changes and, therefore, the duplication of information and cited discrepant condition have been elimated.

D3.1-3 (DEFICIENCY) STAINLESS STEEL PIPE SUPPORTS DESCRIPTION Specification CAR-SH-M30B, part two, paragraph 22.29, requires stainless steel pipe to be separated from carbon steel supports by stainless or plastic shims. Field change request (FCR-H-980) eliminated this requirement, but the specification was not revised.

RESPONSE

CAR-SH-M-308 was taken from CAR-SH-M-30 in the fall of 1982. The referenced paragraph concerning stainicss or plastic shims had been deleted from M-30, but was inadvertently left in the first issue of M-308. CAR-SH-M-30B has now been revised, (Revision 18) and issued with the referenced paragraph deleted.

D3.1-4 (DEFICIENCY) HARRIS PROJECT PIPE SdPPORT PROCEDURES DESCRIPTION Pipe support design guidelines used by personnel performing new design and by personnel resolving field problems had different deflection criteria. Problem resolution guidelines limit deflections for all loadings, while the guidelines for new design require the deflection limit only for OBE.

RESPONSE

These two guidelines were developed separately, with problem resolution criteria developed first. The two have now been integrated into a single document, using a single deflection criteria (Design Guide 7.2.A, Rev 1). This criterion is consistent with our pending FSAR amendment to table 3.9.3-7. I l 1 f l l l l l 1

D3.1-5 (DEFICIENCY) SUPPLEMENTARY STEEL DESCRIPTION The design guidelines for support / restraint design provide data for the use of Bergen-Paterson part E102. This is a non-seismic support steel assembly, but the design guide did not clearly limit its use to non-seismic applications.

RESPONSE

This information was taken from the Bergen-Paterson project instructions, in which Part E102 was clearly non-seismic. The Bergen-Paterson Project Instructions were used by Bergen-Paterson as design criteria for all new design within their scope. However, we have obtained from Bergen-Paterson confirmation of the seismic qualification of the standard and will incorporate this into the HPES Design Guidelines. l l l I l t

D3.1-6 (DEFICIENCY) PIPE SUPPORT STRESS CHECK DESCRIPTION Instructions for stress checks in the support / restraint design guidelines provide for combining direct shears with torsional shears in tubular members using the square root sum-of-the squares ~ (SRSS) method. This is unconservative. Direct and torsional shears in tubular members should be combined directly. In addition, other equations may be incorrect.

RESPONSE

The design guide was incorrectly worded. However, since supports are usually deflection limited not stress limited, there is little potential for significant errors. The design guidelines have been corrected. However, to provide additional assurance that no incorrect analyses were performed, we are reviewing the final calculations which were done while the erroneous equation was ,in effect. Our assessment of the pipe support design guidelines indicates there is no evidence of other incorrect equations. i l [ i l l

U3.1-7 (UNRESOLVED ITEM) U-BOLT LOAD INTERACTION DESCRIPTION Piping support / restraint guidelines provide for combining tension and shear loads on U-bolts using a linear interaction. This was not specified by Bergen-Paterson on the load capacity data sheets for the U-bolts, thus there was no formal documentation that this method is acceptable.

RESPONSE

Bergen-Paterson has confirmed that the linear interaction was appropriate, and further have stated that the qualification of these U-bolts was by linear analysis. We are also working with Bergen-Paterson to have this information incorporated in load capacity data sheets.

References:

Bergen-Paterson letter dated January 25, 1985 M Conroy to R Nilan Bergen-Paterson letter dated February 26, 1985 N Conroy to R Nilan O l i i

U3.1-8 (UNRESOLVED ITEM) FRICTION ANCHOR CLAMPS DESCRIPTION Friction anchor clamps for small bore piping depend on clamping force of two pairs of bolts to hold pipe against reaction forces, rather than welding. The stresses in the pipe due to the anchor bolt pretension are not considered in the piping stress analysis.

RESPONSE

The ASME III Codes in effect for Shearon Harris require that non-integral pipe attachments be designed such that excessive localized bending stresses in the piping system do not result. In this regard, Bergen-Paterson has limited by design the local stress resulting from a friction anchor clamp (a non-integral attachment) to 12,000 psi. This is considered acceptable and consistent with industry practice. Based on Ebasco's experience, the stress induced by the clamp is primarily local bending through the thickness of the pipe. Using Code Cases N-392 and N-318 as guides, the bending stress through the thickness of the pipe need not be evaluated except under cases where expansion stresses and peak stresses are considered. For combined mechanical and thermal loads in such cases, the allowable is SH + SA. This is quite high: for example, where SH = 15000 psi, SA = 22,500 and the total allowable limit is 37,500 psi. The maximum local stress due to a properly installed clamp is not more than 32% of the limit. l l l l l l m

D3.2-1 (DEFICIENCY) DBE INERTIA / FUNCTIONAL CAPABILITY DESCRIPTION According to Specification M-71, CVCS piping must meet emergency condition allowable stress levels for DBE loading. No DBE analysis nor emergency condition evaluation was conducted in calculations 142-1 (5/78), 141-2 (5/78), 157-2 (8/78) and 3000 (8/78). Ebasco has assumed that since DBE stresses vary between 1.0 and 1.5 times OBE stresses and the emergency allowable stress level is 1.5 that of the upset condition, the stress criteria is always met for emergency when the OBE stress analysis has satisfied Upset criteria. The IDI team does not consider the above assumption fail-safe and therefore functional capability as required in FSAR Section 3.9.3-11 is still to be demonstrated.

RESPONSE

Prior to early 1980, no calculations included the DBE case in accordance with procedure SH-006, which required evaluation of OBE only. Calculations performed after that date do include the DBE case. Omission of the DBE case is based upon our experience that, when actual DBE response spectra are considered, the DBE stress will typically vary, depending upon the pipe size, between one and one-and-a-half times the OBE stress. Since the stress limit for the DBE event is not less than 1-1/2 times the OBE limit, functional capability is demonstrated. As such, DBE stresses need not be calculated. To confirm the reasonableness of this assumption, an evaluation of the maximum stress ratios for the upset and emergency conditions was performad for ten calculations. These calculations were randomly selected from those in which the DBE condition was analyzed and includes a sample of large bore and small bore piping. For these calculations, the emergency level stresses varied from .99 to 1.24 times the upset level stresses, thus demonstrating the adequacy of our engineering judgement. For restraint design purposes, DBE loads on restraints are taken as double those for the OBE case, when DBE was not specifically analyzed. Procedure SH-006 was revised on February 21, 1985 to require running a DBE case as part of the piping analysis. The procedure specifically states that previously analyzed calculations need not be rerun for this reason alone. 1 l l

D3.2-2 (OBSERVATION) LOCA ANCHOR MOVEMENT DESCRIPTION LOCA loads were not considered originally in qualifying containment penetrations. Although this was corrected by a special program, some calculation packages do not contain penetration evaluation sheets reflecting the inclusion of LOCA loads in the evaluation. It is recomended that a review of all stress packages with penetrations be made to update the documentation.

RESPONSE

The recommended review is in progress as part of the Calculation Transfer Program. Item #3 of the check list calls for a review of penetration documentation. The check list will be completed by either Ebasco or the Harris Plant Engineering Section. In the case of Calculation 3006, Penetration M8 was found to have been analyzed without consideration of LOCA loads. A note was, therefore, entered on the Calculation Status Form indicating that this evaluation is outstanding. In the case of Calculation 141-3, LOCA evaluation has been conducted. The penetration was found to meet criteria, as noted on the revised penetration evaluation sheet.

i a 03.2-3 (OBSERVATION) EVALUATION OF VALVE ACCELERATIONS DESCRIPTION In several calculations, it was noted that valve g-loading was not evaluated.

,                     It is recommended, for previously released calculations where g-loading had

not been performed, that the evaluation be conducted and documented in the

stress package.

F RESPONSE I Experience has shown that vendor g-load criteria are extremely conservative. ! In fact, even valves that fail to meet these allowables (using the loads ! calculated by the piping stress program) generally can be shown by more j detailed analysis or by test to be acceptable for the calculated loading. Based upon this experience, no review of valve g-loading was conducted during early stages of plant design. Final detailed evaluation of g-loading was

postponed for consideration until piping layouts were more firmly established.

Subsbquently, a program initiated in 1982 was undertaken to review all valves for meeting seismic criteria. To date, this program has demonstrated the

                   .inititial premise to be reliable.

A 1 The' subject program requires that, for all new or revised calculations, an

                    ~ appropriate valve evaluation sheet be prepared and included in the calculation

- package.- Stress Analysis Procedure SH-059 gives the form for Westinghouse

! active valves and Procedure SH-063 gives the forms for Westinghouse non-active ! valves and for Ebasco-procured active and non-active valves. ! Stress Analysis Procedure SH-023, which prescribes the documentation which is mandatory for inclusion in stress packages, includes the above forms. l For a valve which cannot be demonstrated to meet g-load limits, the analyst i enters a note under " Data Required to Complete Analysis" in the calculation package. The analyst also enters the open item in the Isometric Drawing Schedule versus the subject calculation. The calculated loads are transmitted to the Mechanical Engineering Department, which maintains a list of overloaded val ves. The list is updated and published monthly. Mechanical Engineering writes to the appropriate valve vendor to request acceptance of the calculated g-l oading. When ~the acceptance is received, S/A is notified and the overload is removed from the list. S/A then revised the valve evaluation sheet and the calculation.- As a final check that valves have been evaluated per proper criteria, Items 4, 5, 6 and 7 of the Calculation Transfer Program check list assure that either '

the required evaluation sheets, showing compliance with criteria are in the package or a note in- the " Data Required"' form is present.

The above check list is completed by either Ebasco or the Harris Plant Engineering Section (HFES). i

D3.2-4 (DEFICIENCY) WESTINGHOUSE ACTIVE VALVE QUALIFICATION PROGRAM DESCRIPTION' Three Westinghouse active valves (2CS-V600-SB-1, -Y601-SB-1 and -Y602-SB-1) of Calculation 141-2 were reanalyzed as part of the valve qualification program. The valve acceleration and end load evaluation sheets were not found in the calculation package. The subject sheets were found with the documentation of the valve qualification program. It was found that the z-direction acceleration was miscalculated for valve 2CS-V602-SB-1. No reference was made on the calculation sheets to the date or revision of computer analysis.

RESPONSE

The omission of the valve evaluation sheets for the subject valves is an isolated incident. In general, the latest evaluation sheets for in-line - equipment items are filed with the calculation package. The valve accelerations for the subject valves were calculated by an acceptable procedure, as described below:

 .       The valve weight was modeled at two points along the valve stem. At each point, the internal force along a particular direction was printed out.

The sum of these two forces was divided by the sum of the weights (equal to the total valve assenbly weight). These valves were entered in the Active Valve Nozzle Load Evaluation Sheet. The valve evaluation sheets as well as a sample acceleration calculation illustrating the above method of analysis are now part of the calculation package. A reference has been added on each sheet defining the computer run date. i l l l I

D3.2-5 (DEFICIENCY) MODELING OF VALVE CENTER OF GRAVITY DESCRIPTION Three of the sixteen stress calculations reviewed contained valves with improperly _ input offset dimensions for the center of gravity. The particular cases are given below: o In Calculation 3125 the offset was omitted (i.e., the weight was modeled on the valve center line). o In Calculation 141-1 the offset was modeled in the positive-X - (horizontal) rather than vertical-upward direction. o In Calculation 3001 the valve weight was entered incorrectly and divided between two node points which would give an incorrect c.g. location. The above errors were not detected by the analyst in preparing the checklist nor by the checker in verifying the analysis.

RESPONSE

Calculation 141-1: This is an isolated coding error where the correct dimension was entered in the wrong (adjacent) field - an error easily overlooked in checking. The output shows the valve y-acceleration far exceeding the z-direction acceleration, indicating that the results were conservative in that region. However, Calculation 141-1 was rerun on 12/31/84 incorporating several design changes and correcting this error. The criteria were satisfied. Calculation 3125: The omission of the valve c.g. offset dimension is an isolated error. The isometric drawing and calculation have been revised and no impact to the existing design resulted. Calculation 3001: The analyst used the correct valve weight as indicated on the latest , applicable vendor drawing. The succeeding revision, containing different-valve weight information, did not become effective until four days after the . calculation was run (10/2/84 and 9/28/84, respectively). The specific valve drawing identified (EMDRAC 1364-4485) is used in six locations in two stress analysis calculations. In all instances, the effects are minor and this is stated in a note placed in the Isometric Drawing Schedule (IDS). No further action is required. The following process is used to assure reconciliation of the piping stress 4 analysis and hanger designs with the final valve weight.

1. If the calculation used an estimated weight, this fact is noted and addressed during the calculation transfer program.

D3.2-5 (DEFICIENCY) MODELING OF VALVE CENTER OF GRAVITY (Cont'd) RESPONSE (Cont'd) Calculation 3001: (Cont'd)

2. If a major valve change is identified (for example, a valve is replaced by a significantly different model) the system engineer notifies the stress analysis department.

3. The 79-14 program requires a final documentation of valve weights and design reconciliation, if necessary.

It should be noted that the most recent valve drawing is required to be used as the basis for valve weight in stress analyses.

D3.2-6 (DEFICIENCY) EMERGENCY CONDITION STRESS RATIO DESCRIPTION

1) For. ASME Code Equation 9, the emergency condition, the highest stress ratio tabulated by the analyst was 0.530. However, the output gives a stress ratio of 1.069 at another location (node #207). This was an error in reporting by the analyst, as well as a verification error by the checker.

2) Ebasco re-ran the subject calculation (#3125) to correct the c.g.

location of valve 3CS-Y730-SN. Subsequently, the max stress ratio for the emergancy condition - equation 9 was reported as .236 (compared to 1.069, pr viously). This drastic reduction is questionable.

RESPONSE

Point #207 is the first piping node point listed after the tank model elements of the Boric Acid Tank. The analyst and checker mistakenly identified it as part of the tank model and ignored it in evaluating the pipe. This is an isolated error. The reanalysis incorporated several changes which are important to the resul ts. They are:

1) A 2% of critical damping DBE Floor Response Spectra was used for Emergency loading vs the 2X OBE loading used previously.

2) A support (unrelated to the IDI item) was added to reduce stress.

3) Some supports were relocated by the field.

4) The vaive center of gravity was altered 'slightly.

The stress reduction, in light of the above changes, is reasonable. Note that the Equation 9 (Emergency) maximum stress ratio for the reanalysis issued is 0.259, not the 0.236 referenced by the IDI team, i I

i

!                     D3.2-7 (DEFICIENCY) THERMAL MODES DESCRIPTION For stress calculation 3084, two thermal cases were considered: (1) nomal operating at 150 degrees F, and (2) abnormal at 50 degrees F. At valve 3CS-D51-SN-1, the analyst interchanged the normal (150 degrees F) and abnormal (50 degrees F) temperature cases for the continuation of analyzed piping.

1 Thus, the themal case (1) did not have a continuous temperature of 150 degrees F throughout the mathematical model, while themal case (2) did not have a consistent temperature of 50 degrees F throughout.

RESPONSE

Although not in accordance with the line list, the temperature distribution utilized for the Ifnes being analyzed is adequate. The switch in operating temperatures occurs for only two piping elements out of a piping model of sev(ral hundred elements and is, therefore, not considered significant enough to require reanalysis. A notation will be added to the Isometric Drawing Schedule to correct this deficiency if the calculation is reanalyzed. t 3 1 1 i i i

D3.2-8 (DEFICIENCY) THERMAL EXPANSION INPUT DESCRIPTION For calculation 141-1, the analyst did not account for the thermal expansion of the pipe during the 1300F thermal mode case. Only equipment displacements were included. Thus, the calculated pipe stress levels, support loads, and equipment nozzle loads are incorrect.

RESPONSE

Due to an input error, thermal stress and loading were not accounted for in calculation 141-1. However, for a temperature as low as 1300F, the contribution of thermal loading is negligible. Calculation 141-1 was rerun to incorporate changes on documents PW-P-1457, FCR-P-2388, MN/ML 9971, MN/ML 9564 and CQL-8222. Thermal loading was also added in that run with no impact to existing design. i

D3.2-9 (DEFICIENCY) VOLUME CONTROL TANK N0ZZLE DISPLACEMENT DESCRIPTION In Calculation 142-3, the themal nozzle displacements calculated at 2500F for the VCT (nodes #360 and #508) were included for the 1150F thermal mode but omitted for the 2500F mode.

RESPONSE

This is an isolated input error where the thermal displacements were applied to loading case 5 instead of loading case 7. Since the calculation output for points in the vicinities of both nozzles give stress ratios below 0.20, it is clear that the displacements would not raise the stresses beyond the allowables. The calculation was rerun correcting the above omission and incorporating the riser relocation of FCR-P-2628, the valve relocation of PW-P-1915 and the restraint relocation of SH-MNE-8031. The existing design has been demonstrated to be adequate.

D3.2-10 (DEFICIENCY) N0ZZLE THERMAL DISPLACEMENT DESCRIPTION In calculation 750-19, the axial thermal displacement of the case drain nozzle of the charging pump was entered with the wrong sign.

RESPONSE

The sign of a nozzle thermal displacement was inadvertently reversed. Given the minor nature of the displacement (0.004"), detection of this error would not of itself have warranted rerunning of the computer analysis. It is probable that this error was identified by the original checker but because of - less rigorous documentation requirements for early calculations the evaluation of this discrepancy was not specifically noted. The subject calculation (750-19) was rerun for other reasons at which time this change was incorporated. There was no impact to existing design.

i I D3.2-11 (DEFICIENCY) DESIGN PRESSURE DESCRIPTION For calculation 141-2, the analyst calculated the stress for the sustained load equation using an operating pressure of 740 psig instead of the design ' pressure of 2735 psig. For the occasional load equation, an operating

                  -pressure of 2350 psig was used instead of the design pressure of 2735 psig.                                                                                               t
                  ' RESPONSE i

In general, the stress calculation checker, in accordance with review sheet 622-5, must verify that pressure and temperature data are correctly obtained from the piping line list or design specification. In this instance, incorrect values of design pressure were utilized in equations 8 and 9. The impact of this error was recognized as negligible by inspection. However, this calculation was revised for other reasons at which time this correction was incorporated. t i ( 4 l l I i l k

03.2-12 (0BSERVATION) STRESS

SUMMARY

CHECKLIST DESCRIPTION Stress Calculation 157-2 had no accompanying detailed checklist with the computer output.

RESPONSE

For each calculation, a checklist is completed in accordance with Procedure SH-003 and filed with the calculation package. Although this procedure was probably followed for calculation 157-2, the checklist sheets were - subsequently misplaced. This is an isolated incident. It should be noted that calculation 157-2 has been re-checked and a complete checklist was prepared. This checklist has been inserted in the calculation package. To prevent recurrence of this problem, a filing system has recently been initiated to maintain all Q/A documentation for a given calculation in a unique, labeled binder.

03.2-13 (OBSERVATION) FLANGE EVALUATION DESCRIPTION Flange evaluation sheets for two welding neck raised-face flanges were omitted from the calculation package of calculation 141-3.

RESPONSE

Where applicable, for each calculation flanges are evaluated per Procedure SH-055. A " Flange Evaluation Sheet", as shown on sheet 22 of that procedure, is prepared and inserted in the calculation package. The subject omission is an isolated incident which would have been found and corrected as part of the calculation transfer program. The flanges of calculation 141-3 were evaluated and the completed Flange Evaluation Sheets were inserted in the calculation package. They demonstrate compliance with the stress criteria, i I

} I D3.2-14 (DEFICIENCY) N00E P0 INT SPACING DESCRIPTION In calculation 750-19, three element lengths for 1-inch diameter piping ' exceeded the maximum allowable span lengths given in SH Procedure Guidelines

   #16.

It was also noted that, in Appendix R of PIPESTRESS 2010 Manual, the equation for the maximum span length indicates that the variable W has units of 1 lbm/ft. The correct unit is 1b/ft. i RESPONSE In accordance with general Ebasco practice, analysts do not make major changes  ; to isometric drawings. In this isolated case, however, the analyst revised  ; the location of several node points and inadvertently violated the span length criteria of Guideline #16. To correct this condition, the isometric (IA-236-CS-48) was revised and the calculation was rerun. The revisions had negligible effect upon the , { calculation results. The noted error in the PIPESTRESS 2010 Manual has been corrected. This i discrepancy did not affect design work on the Shearon Harris Nuclear Project since the subject equation was not utilized. , t I i r s i i  ; i 3 i 1 f s 1 l i 1 I 4 I l

D3.2-15 (DEFICIENCY) ANCHOR LOCATION DESCRIPTION Isometric drawing 1 A-236-CS-48, Rev. I shows an anchor point locating dimension (north) incorrectly. There is a 2-foot discresancy with the piping drawing and the computer run input, both of which give t1e correct dimension.

RESPONSE

This is an isolated drafting error which was found by the analyst who coded i the geometry correctly in the computer analysis. The isometric was not . . updated at that time. However, the isometric has now been revised. It should be noted that this error has no design or analytical consequences. i 1

l D3.2-16 (DEFICIENCY) PIPE SCHEDULE DESCRIPTION In calculation 750-19, Rev.1, line number 2CSI-478 was modeled in accordance , with the current Line List as schedule 40 piping. However, the calculation l was not updated to reflect the change to schedule 160 given in Design Change Notice FD-1007 of April 12, 1984 and incorporated in Line List revision 20.

RESPONSE

l As stated in Ebc:co Procedure E-69SH, design documents are changed only via signed-out Design Change Notices. Line List changes are initiated Sia Flow Diagram DCNs. Each DCN-FD is reviewed by the departments which are affected by the change (s). Upon approval, the DCN-FD is distributed to the affected departments for incorporation in impacted analyses and documents. In the subject case, this pipe schedule change as well as several other design changes are given in DCN-FD #1007. The DCN-FD, however, was inadvertently identified as having no effect upon the piping analysis. As a result, the calculation remained unchanged. The subject deficiency is not programmatic in nature since the flow path of information does provide for adequate review and distribution. This was an isolated judgement error. Calculation 750-19 was rerun with the subject schedule change incorporated with no impact to existing design.

D_3.2-17 (DEFICIENCY) REGENERATIVE HEAT EXCHANGER SEISMIC ANALYSIS DESCRIPTION The Westinghouse model of the Regenerative Heat Exchanger has a vertical restraint at point 1018, adjacent to the tube side inlet nozzle. However, the model used in Ebasco calculation 3006 omits this restraint.

RESPONSE

In letter CQL-8590, Westinghouse indicated that the subject restraint is . designed with a 3/32 inch gap and that the thermal movement in the vicinity of the restraint is insufficient to close this gap. Ebasco anticipated that this restraint would have negligible effect on the seismic analysis since it is located vertically above an anchor and adjacent to a riser. As such, it was not modeled previously to facilitate the calculation. Although Ebasco is confident in the accuracy of this calculation, upon its next revision, calculation 3006 will be rerun with the support included in the model (as a snubber). This inclusion has been documented in the Isometric Drawing Schedule (IDS). A copy of the IDS page is included in the calculation package.

U3.3-1 (UNRESOLVED ITEM) ITT GRINNELL AIR OPERATED VALVES T DESCRIPTION There was no procedure available to Ebasco stress analysts which provides instructions on the modeling of flexible valves.

RESPONSE

At the time of the performance of the subject calculations, Grinnell diaphragm-operated vs1ves were considered rigid. This assumption was consistent with information provided in the Westinghouse specification. It was not untti 9/4/84 that Westinghouse informed Ebasco of Grinnell's instructions to consider the valves flexible (letter CQL-8174). At that time, it became necessary for Ebasco to develop a procedure for modeling these valves for inclusion in piping system analytical models. Ebasco in conjunction with Westinghouse has developed a procedure for modeling flexible diaphragm ' operated valves. Affected calculations were placed on hold at the time the non-rfgid valves were identified. This activity was initiated prior to the start of IDI. I W

D3.4-1 (DEFICIENCY) PIPE SUPPORT STRUT DESIGN _ DESCRIPTION Pipe support number YD-2-261-1-F0-H-3 in particular and other similar - supports, were designed without regard for the natural frequency of the support structure. In the case of unbraced structural columns, the out-of-plane loading from self weight excitation can be significant in comparison with normal loading of the column.

RESPONSE

Shearon Harris Nuclear Project cornitments as outlined in the FSAR, do not require vibrational analysis of pipe support structures. To control the vibrational response of the structures, displacement units are specified in the FSAR. In accordance with both Harris Plant Engineering Section design guidelines and Bergen-Paterson's Project Instructions, out-of-plane loads are considered in structural design. The calculations for support YD-2-261-FO-H-3 have been revised to incorporate as-constructed data, and to consider out-of-plane loadings with no hardware change required. In addition, Ebasco analyzed a sample of slender column supports which were judged to be limiting and for the worst cases found both deflections and stresses to be within allowables, even under vibrational loading. However, to ensure Support / Restraint structural integrity, we are revising the design guidelines to incorporate additional limits on unbraced cantilevers. All safety related supports undergo a final calculation review after installation. Those reviewed prior to incorporation of the cantilever limits will be re-checked to assure compliance.

n 1

                             \

k D3.4-2 (DEFICIENCY) B-P/CP&L PIPE SUPFORT DESIGN t DESCRIPTION s - A pipe support (A-6-236-1-CS-H-2027) was designed by Bergen-Paterson and revised on site by CP&L design personnel. Design changes made wera not correctly'documnted and' justified by design calculation since calc'ulations done contain,ed errors and incorrect loads. RESPONSE. The calcu?atio.1s for this hanger have been revised, in accordance with Harris. Plant Engineering Section design guidelines, after the hanger was completed and inspected. The calculations now reflect the "as-constructed" conditien 'of the hanger, and no hardware change was necessary. This review of calculations for completed safety related hangers is continuing, and it was in effect prior to the IDI. Since' this occurrence was a random error, no action other that that described abova is required. t l'

                                            /

3

                    /
                                                                 '(

j $ 4 1

                              +

f l I f f

                                      /

4 Y

                                        .                             s l                                    <  l                                   .
                                                                              ~i l                                                                        .

4

D3.5-1 (UNRESOLVED ITEM) WESTINGHOUSE SUPPLIED NON-ACTIVE VALVES DESCRIPTION Ebasco does not evaluate Westinghouse non-active valves in their pipe systems (Class 2 and 3) for compliance with end-load criteria given in the Westinghouse specification. i RESPONSE In Ebasco letter EB-W-2593 to Westinghouse (dated 3/11/85), it was proposed that, "For non-active valves, end loads need not be considered since the

             - valves are stronger than the piping and can not fail before piping fails."

Westinghouse has formally replied concurring with this position which is fully consistent with regulatory requirements. i f t t 1 i e - - +- . -. , , , , ,. - - . - - , ,.m , m--_.. , , , , . ...--.mm.- _ ,,,.m.-, .v+e v --._.y. .mw,,

U4.2-1 (UNRESOLVED ' ITEM) COUPLING EFFECTS OF ECCENTRIC STRUCTURES DESCRIPTION Ebasco performed one-dimensional seismic analyses using its in-house computer code DYNAMIC 2037, which does not include coupling effects. Ebasco did perform a three-dimensional seismic analysis to justify that the one-dimensional approach was adequate. Their justification was based on the comparison of accelerations at each mass point. However, only one direction's acceleration values were used from the three-dimensional analysis for the comparison. This approach did not include the coupling effects.

RESPONSE

The seismic responses of Category I buildings were determined by dynamic seismic analysis of the structural systems utilizing the computer program DYNAMIC 2037. Structural system dynamic characteristics and the seismic responses were calculated by applying seismic input functions in North-South, East-West and Vertical directions individually, as described in the FSAR Section 3.7.2.1 A. For nonsymmetrical structures with significant eccentricities between mass and rigidity centers, a torsional analysis was required as described in FSAR Section 3.7.2.11 A, by modeling the structures in a three dimensional representation to account for out of plane behavior and responses. The STARDYNE program was used for the dynamic analysis with forcing functions applied individually in North-South and East-West directions. The Reactor Auxiliary Building, which has significant eccentricities in the North-South direction was one of the buildings which was modeled and analyzed for torsional effects. The analyses results in North-South direction (worst eccentricity case) were compared with the results obtained from the DYNAMIC 2037 results, and the differences in structural responses even for the worst case were insignificant, precluding the need for torsional analyses in the East-West direction, or other structures with less eccentric configurations. At the request of the IDI team, a complete three directional dynamic seismic analysis was performed on the Reactor Auxiliary Building, to evaluate the extent of the coupling in all three directions. In this new "evaluatory" analysis, existing N-S and E-W horizontals models were used. A vertical model corresponding to the coupled analysis requirements was developed using the actual eccentricities in the vertical directions. Building response accelerations in any one direction due to the individually applied seismic input functions in all three directions were combined by SRSS. The results were tabulated, and compared with the accelerations used in the design (See Table 4.2-1 Seismic Analyses for Major Structures). The differences are insignificant and the designs are not affected. The results of the new analyses were provided to, and discussed with, the IDI team. Based on the above, the IDI team has determined that the FSAR l commitments are justified and this item has been resolved. I i i E_

TABLE 4.2-1 SEISMIC ANALYSES FOR MAJOR STRUCTURES

      .                                        SEISMIC ANALYSIS                    COUPLING EFFECTS   MASS MOMENT OF BUILDING          SHAPE                PROGRAM                        CONSIDERED     INERTIA CONSIDERED Containment      Symmetrical          DYNAMIC 2037                    No (Notes 1, 2)       No (Note 3)
   , Reactor Aux      Nonsymmetrical     . DYNAMIC 2037                    No (Note 4)           Yes (Unit #1)                         STARDYNE (For Torsional )

Reactor Aux Symmetrical DYNAMIC 2037 No (Note 1) No (Note 3) N (Common) - Diesel Gen Symmetrical DYNAMIC 2037 No (Note 1) No (Note 3) ESW Intake Nonsymmetrical STARDYNE No (Note 4) Yes Structure . Tank Bldg Nonsymmetrical STARDYNE No (Note 4) Yes Notes

1. .The coupling effects are insignificant since the building is symmetrical and the eccentricity between the mass and the rigidity center is small.

2.' A comparison of seismic responses for Containment Building using one-dimensional DYNAMIC 2037 and three-dimensional STARDYNE models was made and showed that the differences in resopnses were insignificant. . 3. The exclusion of mass moments of inertia for mass points above the foundation mat

in symmetrical structures is_ considered to be insignificant. At the request of the IDI team, the effect of exclusion of local mass moments of inertia was evaluated for. Containment Building and found to be insignificant (see response to D4.2-2).

, 4. At the request of the IDI team, the coupling effects on the final seismic responses

l. was evaluated for Reactor Auxiliary Building (typical nonsymmetrical Bldg). See response to 04.2-1.

l ('

D4.2-2 (DEFICIENCY) MASS MOMENTS OF INERTIA DESCRIPTION Ebasco omitted the mass moments of inertia in its one-dimensional seismic analyses for the Reactor Containment Building, Tank Building, Diesel Generator Buildings and.the Emergency Service Water Intake Structure. This property was only considered for the foundation mats of these buildings. Ebasco used its in-house computer code DYNAMIC 2037 for the one-dimensional analysis. In Ebasco's three-dimensional torsional analyses using STARDYNE, only the torsional mass moments of inertia were included. This approach is unconservative because omitting the mass moments of inertia neglects the responses from rotation which are potentially significant. Note: The IDI report incorrectly lists Tank Building and Emergency Service Water Intake Structure for one-dimensional analysis; these buildings were analyzed using a three dimensional torsional analyses, which included mass moments of inertia in all three directions.

RESPONSE

1 The structural dynamic analysis computer program DYNAMIC 2037 has an option feature for inclusion or exclusion of mass moments of inertia to account for local rotational response effects of idealized mass' points depending on the-structural symmetry of the structures. The mass moment of inertia of the base , mat is always included to properly represent building rocking mode effects. Since the Containment Building and Diesel Generator Building dynamic models were judged to be symmetric, the exclusion option of the program was excercised.. Models for unsymmetric structures such as the Reactor Auxiliary Building an'd Emergency Service Water Intake Structure and Tank Building were reviewed and it was confirmed that mass moments of inertia were included in the~ one-dimensional analysis (A-116) as well as the three-dimensional analysis (TK-101, A-133 and EI-108). Upon request of the IDI team, additional analyses were performed to assess the effect of mass moments of inertia on response spectra. The Reactor Auxiliary a Building model was revised to exclude mass moments of inertia while the Containment Building model was revised by including mass moments of inertia. The results (i.e. structural responses) were compared and confirmed that the effect of the exclusion of local mass moments of inertia is insignificant for , a symmetric structure such as the Containment Building. This was discussed with the IDI team. This condition is limited to symmetric seismic Category I buildings which

                   -utilized the option of excluding local mass moments of inertia in the analsis.

Corrective action is not required since the additional analyses described above have provided the technical justification for the analysis techniques - utilized. i l b

     -+ . +.- --.--,,                -,-t , , , - . - -       3 -% ..-,m-     - , , wn   -.--~~-,---m-e----.e-,-----.--..       , , . . ,    .--,..---,---.w--.c--e-.-           -.m

if U4.2-3 (UNRESOLVED ITEM) FLOOR RESPONSE SPECTRA BASED ON ONE-DIMENSIONSL vs THREE-DIMENSIONAL SEISMIC ANALYSIS DESCRIPTION 4 Ebasco performed seismic analyses using one-dimensional approach as indicated in the FSAR Section 3.7.2.1 A. The floor response spectra were developed based on the results from the one-dimensional analyses. A three-dimensional analysis was performed to confirm the adequacy of the one-dimensional approach. By comparing computer runs, the IDI team found that the frequencies from the three-dimensional run, except the first frequency (corresponding to the fundamental frequency), differed significantly from those of the one-dimensional analyses. Ebasco did not generate floor response spectra with the three-dimensional approach to justify the adequacy of the floor response spectra results of the one-dimensional analysis. The justification Ebasco performed was an acceleration comparison which the IDI team determined was inadequate.

RESPONSE

Floor response spectra curves were generated for three directions of earthquake motion using the computer program DYNAMIC 2037 as stated in FSAR Section 3.7.2.5A. At that time, this calculation technique was considered the acceptable state of the art. As requested by the IDI team, floor response spectra curves were generated for the Reactor Auxiliary Building (RAB) floor at El 305 (DBE 4% damping). A three-dimensional mathematical model was input into the computer program STARDYNE to evaluate the coupling effect on the floor response spectra. The effect was not significant. The revised response spectra curves were provided to, and discussed with, the IDI team. Based on the above, the IDI team has determined that the FSAR commitments are justified and this item has been resolved . 4 y -- n- , .-- sv.---e.-e.w w- e v -, . , -------wr---c - - -- -- , _- - +.--.--- e .m--.--,,- - - - - *-

1 D4.2-4 (DEFICIENCY) PREPARATION OF CALCULATIONS-DESCRIPTION Several examples of inadequate preparation of calculations were noted by the IDI team: Calculations No. C131 and No. C132 do not contain criteria, assumptions, etc. They also contain incorrect reference pages and, in many instances, they leave reference page numbers blank. Calculation No. G248S03 has either wrong reference page numbers or references cannot be located. Calculation Nos. C103 and Cl32 have references which were prepared several months later than the calculations. Calculation No. C101 has inconsistent references in its loading combinations. Calculation No. EIl08. Certain calculations were deleted; however, documentation of the preparer and checker was not identified.

RESPONSE

These deficiencies are in part due to the longevity of the project, during which several changes in the layout and design of structures were made. Further, the schedule was such that in many instances different engineers ' performed calculations on various portions of an individual structure or building in parallel. 'These individual calculations were .then compiled into a ! unified calculation which sometimes resulted in non-uniform format and inconsistent page numbering. As noted by the IDI team, this condition is a documentation problem. It has no effect on analyses or design. The existing design is adequate. A program to review the cited Civil design calculations has been initiated to

    . correct page numbers and discrepancies in references and to generate references, if missing. In general, calculations will be reviewed upon turnover to CP&L to ascertain acceptability for use in future modifications.

At that time, inconsistent references or similar deficiencies will be l- resolved. 1 Design calculations are essentially complete. For new or revised f calculations, special attention will be given to format and documentation considerations. i l l I F

             ..-.,-,   -,e, _, , - . , - + w<-me--t- - - -  -=4c-    + n   *- weee----ve-*-     -t-e+-e++w ve---- e--r-eeme r  mtd+v++--r-       r *- -- +-'w=v,ni

D4.2-5 (DEFICIENCY) INCONSISTENCY BETWEEN CALCULATION AND SEISMIC MODELS DESCRIPTION The IDI team found three inconsistencies between seismic models and their associated calculations. Ebasco Calculation Ho. A116 computed the mass and mass moment of inertig for a support point at elevation 216.0 ft to be 590 kips and 0.984 E04 kip-ft , respectively, but the seismic model showg that the mass and mass moment of inertia are 350 kips and 0.79 E04 kip-ft , respectively. At elevation 286 ft, the calculated vertical mass is 20704 kips, while the model indicates the mass would be 20551 kips. Another calculation, No. C131, computed the mass moment of inertia at the founda kip-ftgion

            ; thelevel (elevation model          190 aft)value showed that     in the of North-South  to be 5.75 E07 kip-ft4 was6.25 used.E07 Ebasco showed the IDI team another computer run using correct numerical values for mass moment of inertia which demonstrated that this error had no effect on the final results, but this was never documented.

RESPONSE

Calculation Book A-ll6, " Dynamic Analysis input of the Reactor Auxiliary Building Unit 1" Elevations 206 and 226 feet - Two dummy points, 6 and 8, were utilized in the mathematical model to represent boundary conditions. For use in the computer calculations, low but compatible numbers for the weight and weight moment of inertia were to be assigned to these dummy points. Weights and weight moments of inertia for the two points were computed by an empirical equation and for conservatism the lower of the two were used in the model for both the dunny points 6 and 8. No statement of this effect was made in the calculation book. I Elevation 286.0 feet - In the original model, the weight of the vertical I mass point 3 was reduced by an amount of 153 kips (20704k - 20551k) and this difference in weight was assigned to a branch mass point shown in the vertical model only. Subsequently when the vertical model was revised to incorporate changes in the branch points, the main mass points were not adjusted. Since the new branch weights were relatively small, the resulting effect was judged to be insignificant. No statement to this effect was made in the calculation book. Calculation Book C-131, " Containment Building Dynamic Analysis Input Data" Elevation 190.0 feet - The inconsistency in the value of the mass moment of inertia used in the input data for dynamic analysis was due to an inadvertent error. Subsequent analysis using the corrected values was performed in 1983 which confirmed that this error had no effect on dynamic response (Book C-231-2). However, no statement to this effect l was made in the calculation book. { i I l

D4.2-5 (DEFICIENCY) INCONSISTENCY BETWEEN CALCULATION AND SEISMIC MODELS

  - (Cont'd)

RESPONSE (Cont'd) A review of similar calculations indicates that this is an isolated condition limited to the three instances cited above. Explar.atory statements have been added to the Calculation Books A-116 (Rev.1) and C-131 (Rev.1) to make the model and input computations consistent. The design is adequate; no additional analysis is necessary. I

04.2-6 (OBSERVATION) PEAK VERTICAL ACCELERATION ASSUMPTION DESCRIPTION Ebasco Calculation Nos. All6 and A316 indicate in their assumptions that the vertical earthquake should be 2/3 of the horizontal earthquake. However, by reviewing the actual computer input, the team noted that the horizontal and vertical earthquakes used were of the same magnitude. Therefore, the actual input was in accordance with the provisions of Regulatory Guide 1.60.

RESPONSE

The statement concerning the vertical ground acceleration was in accordance with the earlier design criteria. The relationship between the vertical and horizontal ground accelerations was later revised to be consistent with the Regulatory Guides. The design criteria and the calculations were revised accordingly. However, the sentence in the introduction part of the Calculation Book A-ll6 was not revised due to an oversight. This is an isolated case, limited to RAB, Calculation Books A-116 and 316. The introductory statements in Calculation Books A-116 (Rev 1) and A-316 (Rev

1) have been corrected. The calculations, based on the revised criteria for vertical acceleration are not affected.

l l l l l l

D4.2-7 (DEFICIENCY) SHEAR AREA DESCRIPTION Ebasco assumed tho shear areas to be the same and equal to the cross-sectional areas for several members in the seismic models of the tank building and main dam spillway. In the intake structure model, the shear areas were taken as 85% of the cross-sectional area in both directions. This approach overestimates the shear areas and could underestimate the seismic responses. Ebasco used the conventional approach to calculate the shear areas in other seismic analyses. CESPONSE In the 3 dimensional seismic analyses for the Tank Building (Calculation Book TK-101) and the ESWS Intake Structure (Calculation Book EI-108), the shear areas for individual mass points were based on shape factors available in current literature. These shear areas may not have been applicable in all cases. The Main Dam Spillway which was analyzed as a 2 dimensional model using the DYNAMIC 2037 program for seismic analysis was also questioned by the IDI team. This condition is limited to the Tank Building, ESWS Intake Structure and Main Dam Spillway. Tank Building - In response to the IDI team's concern, new response spectra based on revised shear areas were generated at representative locations. A comparison of the new and the original spectra demonstrates that the original broadened spectra envelopes the new spectra. These results, which have been documented and discussed with the IDI inspector, confinn that the existing design is adequate (TK101, Rev 3). ESWS Intake Structure - In the 3 dimensional seismic analyses for the ESWS Intake Structure, the shear areas for individual mass points were based on a shape factor of 0.85. This shape factor was applied to the gross area. j Although the shear shape factor selected was not chosen strictly in accordance t with the recommended procedure (the structure in question is open on the south side from El.190 to El. 219 feet) it was judged that the existing conservatism in the mathematical model would still yield an adequate structural design. l The mathematical model contains the following conservatisms:

1. The structure is supported on rock and surrounded by rock / soil on three l sides to varying depths from El. 225 ft. to El. 260 ft (the plant grade).

Therefore, a major portion of the structure will move with the ground leaving only a small portion to behave as a cantilever. The model l however, assumes cantilever action from the mat up. l

2. Shear friction between the structure and rock / soil interface along the sides (E-W) and the rear face (North side) has not been included in the shear springs.

3. Bearing in the N-S direction above the foundation 0 El.190.0 ft has not been considered in the bearing spring because the structure is not l surrounded by soil on the south side, though the rock / soil on the north side will have a significant damping effect on the dynamic motion of the structure in the N-S direction.

D4.2-7 (DEFICIENCY) SHEAR AREA (Cont'd) RESPONSE (Cont'd)

4. Bearing between the ' structure and soil interface along the East side of the structure above E1. 240 + feet has been ignored.

5. The foundation damping is assumed to be 5% for DBE as stated in the FSAR whereas the actual damping will be much higher (two to three times due to radiation effects).

Notwithstanding the above conservatism in the original analysis, an evaluatory analysis has been performed using the original mathematical model but including the actual shear areas. N-S earthquake. conditions were selected for this analysis since this was the governing case for the stability of the structure in the original design. The actual effective shear areas at various elevations were calculated and used in the analysis. The building response accelerations have been compared and found to be within an acceptable range of I the original values. The design of the structure is adequate (Calculation t Book EI-108, Rev 1). Main Dam Spillway - the spillway structure has been analyzed by a 2 dimensional 4 model using.the computer program DYNAMIC 2037. The model has three independent c cantilevers on a common mat which represent the two abutments and the central , pier (the abutments and the pier are structurally connected only at the-base). The cantilevers have solid rectangular cross-sections. The central pier is 78 feet long (N-S) and 8 feet thick (E-W). Since the N-S dimension is very long compared to the E-W dimension, the shape factor for N-S motion will approach unity. Therefore,100 percent shear area is considered in the analysis. For the E-W motion the cantilevers are not rigid and the deformation is predominantly due to bending.- The contribution of shear area for deformation and frequency is negligible. The computations show that the change in the frequency of the cantilever with 85 percent shear area is not'significantly different from the frequency with 100 percent shear area (Calculation Book R-104, Rev 2). Therefore,-the design of the spillway structure is considered to be adequate. l

D4.3-1 (DEFICIENCY) RADIAL REINFORCEMENT STRESS DOCUMENTATION DESCRIPTION The IDI team review of the containment basemat design calculation identified that there was no detailed calculation for the listed radial reinforcement stress of 44.6 ksi that was found in the summary of stresses for Calculation Book No. C101. However, the detailed calculation was found in an uncontrolled Ebasco Calculation No. 2-2G.

RESPONSE

Calculation Book C-101 " Containment Building Base Mat Analysis & Design", gives a comparison summary of stresses in the mat reinforcement based upon the old design and the new design initiated after issuance of NRC Regulatory Guides 1.60 and 1.61. Calculations for one case of the old design for the summary table were not documented in the Project Files. After issuance of Regulatory Guides 1.60 and 1.61 the Reactor Containment Building (RCB) mat was redesigned and the new calculations were documented in the Project Files. The comparison of stresses in the mat reinforcement was made as an evaluation of Regulatory Guides 1.60 and 1.61. The calculations for one case of the old design for the summary table were not sent to the Project Files. These calculations were, however, available in Civil Design in the calculation book titled " Containment Building NRC Studies Vol. I", and support the numbers in the summary table in Book C-101. The condition described above is isolated to the RCB basemat. Since in fact the design calculation did exist, though not in Project files, the deficiency has no impact on the design. The relevant calculation pages for the old design from the Book " Containment Building NRC Studies Vol. I" have been added to Calculation Book C-101 (Rev 3), and sent to the Project Files. These pages support the numbers used in the comparison summary table of the old and new designs. The design of the basemat is adequate. l l t

D4.4-1 (DEFICIENCY) LOADING COMBINATIONS - SEISMIC LOAD DESCRIPTION The IDI team review of the design calculations of the containment building polar crane girders and the internal structural steel platforms found that the structural members were preliminarily designed based on estimated vertical seismic loads. However, loading combinations used to assess the final structural analysis did not appear to consider the vertical structural seismic load. Ebasco indicated that there was sufficient conservatism in the preliminary design. Nevertheless, the effects of seismic load should be properly evaluated in the final analysis to ascertain that the design is in fact conservative.

RESPONSE

The Containment Building structural steel platform members and polar crane runway girders are designed for a conservative seismic acceleration of 1.0g in three directions. When the actual response spectra curves were developed the platform steel members were analyzed by the computer program for a conservative acceleration of 1.09 in three directions. The 1.0g criteria enveloped all actual response accelerations in any of the three directions. The comparison showing that 1.0g accelerations enveloped the actual response spectra accelerations was not documented. In the design of the polar crane runway girders, vertical seismic stress due to the girder's own weight which amounts to a very insignificant portion of the total stress, was inadvertently left out of the combination. Our review of steel structures indicates that this condition is limited to Containment building structural. steel platform members and polar crane runway girders. The comparison of the actual response accelerations with the design accelerations used for the design of above noted structures and the loading combination have been documented in the Calculation Books (CAS-1 and 2168-G-253 S02). The accelerations used for design are conservative. The existing design is adequate. 1 0

                                                                    - ~ , , ,,.,

04.4-2 (0BSERVATION) SUPERSEDED DESIGN CALCULATION DESCRIPTION The IDI team found two examples of voided calculations where the superseded calculations were still the basis of the design: Calculation Books G248 S02 and TK105.

RESPONSE

Steam generator support anchor plate - The initial calculation (Calculation Book G-248 502) was stamped superseded and a reference was made to the "as-built" check calculations performed later. Since the drawing is based upon the initial calculation, the initial calculations should not have been superseded. They should have been cross-referenced with the later calculation to a; count for the "as-built" condition. Anchor bolt enbedment desicn for Refueling Water Storage Tank, Reactor Make-up Water Storage Tank and Concensate Storage Tank - In Revision 1 of Calculation Book TK-101, the new tank loads were included, however, no new calculations were performed since these loads were less than the original loads. Subsequently, the original calculations were deleted in Revision 2 of the calculation book. The extent of this documentation problem is limited and has no impact on the design. The applicable calculations for the Steam Generator and Tank anchorages have been reinstated and/or a cross-reference to the "as-built" calculations has < been provided (Calculation Books G-248 S02, Rev 8 TK-101, Rev 3). r

                               - , = . , . . - .   - . . - . - , - - - . . , , . - -

D4.5-1 (DEFICIENCY) SLAB DESIGN USING DIRECT DESIGN METHOD DESCRIPTION In the Reactor Auxiliary Building slab design, ACI-318-71 Code, Section 13.3.1.3, was violated. This section states that "the successive span lengths in each direction shall not differ by more than one-third of the longer span." The portion of the sleb at El. 324.00 ft, bounded by column lines B, D and 42, 44, has a dimension of 55 ft in the east-west direction. The successive span in the same direction, bounded by column lines D, E and 42, 44, has a dimension of 30 ft. The difference between these two spans is 46%. In the Fuel Handling Building slab design, Section 13.3.1.1 of the same code was violated. This section states that "there shall be a minimum of three continuous spans in each direction." This slab contains two continuous spans that do not adhere to the code restriction (FH-540 page 12).

RESPONSE

The initial design methodology utilized for the Reactor Auxiliary Building (RAB) slab at E1. 324.0 feet (Calculation Book A-351) and Fuel Handling Building (FHB) slab at E1. 236.0 feet (Calculation Book FH-540) was based on

  " Direct Design Method" as described in Section 13.3 of Building Code Requirements for Reinforced Concrete ( ACI 318-71). For certain slab panels, the limitations of the method given in Section 13.3.1 of the ACI Code were not satisfied.

Based upon knowledge and experience with other ACI methods, the results obtained from the Direct Method were adjusted to compensate for the limitations of the method. No clarification was provided. Subsequent calculations performed in accordance with the applicable alternate design method confirm that the existing design of the slab panels is adequate. Based on a review of design calculation, this deficiency is not systematic and j is limited to certain slab panels in RAB and FHB only. Confirmatory calculations for the RAB and FHB slab panels using alternate methods have been performed and documented (Calculation Book A-351) and L FH-540). The calculations confirm that the existing design is adequate. 1 t l

04.5-2 (DEFICIENCY) SPACING OF SLAB REINFORCEMENT DESCRIPTION The spacing of the top flexural reinforcing does not match the calculations. Calculations A345 shows that No. 8 rebars at 5-1/2 inches spacing are to be placed at the top of the slab. However, the applicable drawing of the . as-built condition shows the spacing to be less conservative at 6 inches. RESPONSE. The design calculations for the Reactor Auxiliary Building (RAB) slab at El. 261 calls for 51/2" spacing of reinforcement whereas the design drawings call

    . for 6" spacing. The actual reinforcement required by calculation is 1.48 square inches per foot width of the slab, which is less than #8 rebars at 6" spacing (1.58 square inches per foot) as shown on the drawings.

Due to practical construction considerations, after review of the design calculations the rebar spacing of 6" was used for the slab at E1. 261 and 5" for the roof slab at E1. 324. The 6" spacing of rebars in the drawing provides greater reinforcement than that required by the calculations. A statement to this effect should have been made in the calculation for consistency of design and drawing. As noted by the IDI team, the inconsistency between the design calculations and drawings described above is an isolated documentation problem and limited to RAB slabs at El 236 and El 261 only. In both instances, the design reinforcement actually provided is adequate. An explanatory statement has been added in design Calculations (A-153 and A-345) for consistency with the applicable drawings. t i i

D4.5-3 (DEFICIENCY) LOAD COMBINATION FOR SLAB DESIGN DESCRIPTION The Reactor Auxiliary Building (RAB) slab at El 236.00 ft was not analyzed by using the load combinations as described in Section 3.8.4.3.2 of the FSAR. A load combination 1.5D + 1.8L was used. Instead,1.4D +1.7D +1.9E should have been used (where D = dead load, L = live load, E = OBE earthquake). The latter load combination would give higher loads on the slab, thus making the present analysis less conserative.

RESPONSE

The RAB slab at El 236.0 in the scalloped area was designed using conservative boundary conditions, for the normal operating condition in accordance with the original design criteria. When the load combinations were revised to comply with the subsequent Standard Review Plan in 1978, new calculations were not performed since the original design was based on a conservative approach. However, no statement to that effect was provided. Calculations for the revised load combinations have been performed and confirm that the slab design is adequate. As noted by the IDI team, this condition is not systematic but is limited to the Reactor Auxiliary Building (RAB) scalloped area. Calculations for design of the RAB slab in the scalloped area based on the latest load combination have been added to Calculation Book A-153. The existing design is adequate. i

D4.5-4 (DEFICIENCY) SEISMIC ANALYSIS FOR MASONRY WALLS DESCRIPTION

 . Solid masonry walls in the vicinity of the volume control tank in the reactor auxiliary building were not analyzed seismically. Ebasco calculation CAR /C91 does not include the seismic analysis of these walls. Ebasco stated that;
 . these walls were similar to another solid masonry wall that was analyzed
 ' seismically. However, the IDI team noted that such similarity did not exist due to the fact that ~the walls around the volume control tank were much higher than the wall Ebasco claimed to be similar.

RESPONSE

Seismically designed masonry block walls were designed either by individual analyses or by similarity to other seismically analyzed wall panels. The masonry block walls of the Volume Control Tank enclosure were qualified by similarity; however, due to height differences in the walls the applicability of comparison was questioned. Typical wall panels selected for complete seismic analysis / design may not have been fully similar to envelope all masonry walls. All seismically designed masonry walls analyzed by similarity are being reviewed for applicability of the typical wall panels. Any wall identified as being not covered by a typical wall panel will be analyzed and documented in the calculation book. A list of all seismically designed masonry walls will also be included in the calculation book which will relate each wall to the typical wall panel. The specific walls in question have been analyzed and demonstrated to be adequate. L

D4.5-5 (DEFICIENCY) USE OF FLOOR RESPONSE SPECTRA DESCRIPTION In the Reactor Auxiliary Building, the hollow masonry walls around stairway A-4 were designed using unbroadened floor response spectra curves (CAR /C91). This could lead to an unconservative design; if the broadened curves were used, higher acceleration and hence higher seismic loads would have been obtained.

RESPONSE

Non-broadened response spectra were inadvertently used in the seismic analysis of the wall for stairwell A-4. As the IDI . team noted, this is an isolated case, limited to this specific concrete block wall design. The wall for stairwell A-4 was subsequently redesigned due to a Permanent Waiver (PW-AS-1045) at which point the broadened response spectra was utilized. - The original design is no longer applicable and the calculations have been so noted.

D4.5-6 (DEFICIENCY) DESIGN OF MASONRY WALL AROUND STAIRWAY A-4 DESCRIPTION The as-built condition, as described in Permanent Waiver AS-1045, 4-#4 bars in each column are used instead of 4-#6 rebars. Also, the north and south ends 1 off the stairway contain only one W6x25 instead of two. Both items change the physical properties of the masonry wall, making the as-built condition different and less conservative than the requirements of design. During the walkdown, the IDI team noted that the upper portion of the masonry wall did have 4-#4 rebars in place; however, there was no W6x25 beam installed.

RESPONSE

The initial calculation for this seismic Category I block wall considered 2-W6x25 support beams and 4-#6 rebars. The final design actually utilized only 1-W6x25 support and 8-#4 rebars. Both of these changes were reviewed prior to release of the final wall design and found to be acceptable. The original calculation, however, was not revised.

   -The design was prepared for resolution of a specific Permanent Waiver (PW-AS-1045)' addressing specifically masonry walls around stairway A-4,

. therefore, this is an isolated case. 1Rua calculation book has been updated to reflect the as-built condition. i F t f i 4 e --m- --, ,-- , ,, - - , - , - , - - - , , , - - - , , , -- ,---~---e --

                                                                                                 --,q--sm,,- --g -,m  .  , -
                                                                                                                                ,~s

D4.8-1 (DEFICIENCY) LOAD COMBINATIONS FOR MAIN DAM SPILLWAY DESCRIPTION In the design of the main dam spillway, only the load combination which includes Design Basis Earthquake (DBE) was used. The load combination with Operating Basis Earthquake (OBE), which has higher load factors for dead load, soil and hydrostatic pressure, was not checked. This approach might lead to an unconservative design.

RESPONSE

The side abutment walls and central pier of the main dam spillway (Calculation Book R-104, Rev 1) were not analyzed for the OBE condition. The abutmei' walls and pier of the main dam spillway w re designed for the DBE condition ut"lizing a simplified conservative design procedure. Based upon engineering j;dgement the design was considered adequate for the OBE condition. As such, no additional calculations were provided. Our review of the design of all Category I structures indicates that this condition is limited to the main dam spillway. , The additional analyses for the abutment and the pier of the main dam spillway have been documented (Calculation Book R-104, Rev 2). The existing design is adequate.

D4.8-2 (DEFICIENCY) MAIN DAM SPILLWAY ABUTMENT DESIGN DESCRIPTION The IDI team discovered that Ebasco Calculation R104 (Rev.1) contained an error in the magnitude of the abutment's vertical compressive force. The vertical force should have been 1/2 of the value calculated. This mistake apparently arose because Ebasco incorrectly used the entire mass of both abutments to calculate the compressive force. Since the calculations show that a higher vertical force would decrease the amount of reinforcing steel provided, the error would lead to an unconservative design.

RESPONSE

In the analysis of the abutment wall under DBE conditions (Calculation Book R-104), the weight of the abutment was inadvertently over-estimated. The abutment wall has been reanalyzed with the correct weight and was found to be adequate. As noted by the IDI team, this is an isolated case. The revised calculations for the design of the abutment walls have been added to Calculation Book R-104, Rev. 2. The existing design of the abutment is adequate. 4 I

                       . . - . . - - - - . . . - -- -. --    ,      . . - - -- ..,-- , ,,          ,e-, , , , , , ,,

D4.9-1 (DEFICIENCY) BORON RECYCLE HOLD-UP TANK SEISMIC LOADS DESCRIPTION The design of the slab and beams under the boron recycle hold-up tank (FH-540) includes the vertical static and seismic loads, but does not consider the effects of the horizontal seismic loads from the tank. This represents an unconservative design approach since these horizontal loads increase the loading on the slab and the beams underneath the tank.

RESPONSE

The slab panel at El. 236.0 feet in the Fuel Handling Building, which supports the Boron Recycle Hold-up Tank, is designed for 1.25 times the applicable dead and vertical seismic loads of the tank. This 25 percent increase in the design loads was judged to be adequate to account for the actual local effect of the horizontal seismic load of the tank. However, no documentation of this judgement was provided in the calculation. The above condition is limited to Fuel Handling Building, Reactor Auxiliary Building and Waste Processing Building. Calculations for the design of the slab beneath Boron Recycle Hold-up Tank in the FHB have been performed with actual loads including horizontal seismic loads and documented (FH-540, Rev. 2). A review of slabs supporting large tanks in other buildings has also been performed to verify the adcquacy of existing design (AID-5 and W-611). These calculations demonstrate that the existing design is adequate. 1 l l

D4.10-1 (DEFICIENCY) CABLE TRAY SUPPORT FREQUENCY DESCRIPTION In the Reactor Auxiliary Building, two types of longitudinal bracing with . , either one or two rows of struts were used for cable tray supports. Ebasco  ! calculation G170S02 shows that frequency analysis for the bracing for two row i of struts was performed. However, Ebasco calculation G170S01, which applies to the design of one row of struts, failed to examine the frequency of longitudinal bracing. A frequency of 16 Hz was assumed in design. Ebasco j drawings (2166-G699S02, 2168-G170S01) show that the longitudinal bracing , provided for supports of type E2-1 are for one row of struts and should be checked for frequency.

RESPONSE

I j Cable tray supports were designed by analysis, similarity or by inspection /

;                      judgement. Specific references defining the design method utilized were not l                        in all cases provided.                                                                                                        ,

i - ! In general, the same group of experienced design engineers have been involved in the production effort of cable tray restraint analysis and design. This , continuous participation in establishing and evaluating frequency

characteristics and the state of stress of large numbers of supports, have

provided the lead design enginects with the ability to predict expected

! characteristics of the support, without performing an analysis. Similarity to ! . supports which were previously analyzed and found to be acceptable was used as a design basis. The proper references to similar and previously qualified ,

,                      supports or clarifying statements in support of engineering judgements used i

were not always provided. i l The approach described above was used in all cable tray support design. j The adequacy of the present design for those supports that may riot have a

specific qualification analysis is not in question. The inherent conservatism j in the approach and the experience and judgement of the design engineers insure a safe design. Nevertheless, a program to evaluate the calculations

for all supports shown on sample drawings has been initiated. This program will identify supports which have not been specifically qualified by stated similarity and either identify and document the similar cases, or perform and document the actual analysis. The selected sample drawings include approximately twenty-five percent of all cable tray restraints on the project. The sample is random and unbiased in that all buildings and both congested and uncongested areas are represented. If, as expected, all supports requiring qualification are shown to be adequate, no further l

corrective action will be initiated. 4 ' ! Frequency analysis for longitudinal bracing with one row of struts has been ! performed and the frequency has been shown to be above 16 Hz and thus is i acceptable. ' L 4 Future supports will either be designed by specific analysis or will be l clearly referenced to a similar support which has been analyzed. ! i

   .m - - _ . _ - . _ -.          _ _ . - , , . _ _ . _ , _ _ . _ _ _ _ . . ,                          -

04.11-1 (DEFICIENCY) FREQUENCY OF HVAC DUCTS DESCRIPTION Ebasco computer run shows that for a 30-in x 30-in duct, the maximum span is 21.5 ft; and for 44-in x 44-in duct, this span should not exceed 24.5 ft in order to have a frequency.of 33 Hz. Drawing CAR-2168-G-813 shows that spans of 25.0 ft and 30 ft 10 in were used for 30-in x 30-in and 44-in x 44-in ducts, respectively. Using greater spans decreases the frequency and makes the as-built condition less conservative. Note: The duct sizes stated above are in error. They should read 30 x 52 and 50 x 44 in lieu of 30 x 30 and 44 x 44 as referenced above.

RESPONSE

The maximum duct span between seismic supports was in some cases exceeded. This resulted from a change in the method of calculating duct moments of inertia or relocation of seismic supports in order to match embedment locations and/or to clear interferences with other equipment or subsystems. All HVAC seismic support drawings have been reviewed and there are eighty-five such cases identified. With the exception of five cases, the overspan conditions are minor and the support designs are acceptable due to the conservatism in the generic criteria for HVAC subsystem design. Confirmation of duct adequacy is in progress. Further evaluation has been initiated to confirm the support and duct adequacy for the remaining five cases. HVAC duct design is essentially complete. All future work will be reviewed to insure conformance with applicable span criteria.

__ - _ .-. . .. _ = _ _ . _ . ._ -.__ _ _ _ _ - - ~_ _.--.__ _.._ D4.11-2 (DEFICIENCY) LOADS ON HVAC. DUCT SUPPORTS

    ~

DESCRIPTION The HVAC duct along column line B in the Unit 1 Reactor Auxiliary Building El 216.00 ft shows that the two anchors F-ll52 and F-1161 are approximately 169.0 ft. apart (Dwg 2168 G-815). Between these two anchors, there are guides which

,          are spaced at shorter intervals. The static loads transferred to the anchors in the two horizontal directions should be different. The load in the
          -direction of the duct should be much greater than the transverse direction.

Calculations (Calculation Book G-7003, G-7004,'G-7005 and G-7006) show that the static loads in the two horizontal directions are the same for anchors F-ll52 and F-ll61.

RESPONSE

4 The mass loads in the axial direction for HVAC duct seismic supports (anchors) were in some cases not in agreement with actual conditions. All mass load calculations have been reviewed. There are one hundred forty , one cases where actual loads are higher than those originally used in the support design. ) Calculations for the specific cases identified by the IDI team (F-1152, F-1161) have been completed and have confirmed the adequacy of the existing 4 supports. Confirmatory analyses to establish the design adequacy of the . remaining supports is in progress. r I L i I. t i i i vr -yw w.- y,-~gy -+-pp9 - ww- ww s -w- e -.rmy. -- 7m+m-g=-,.wy-7 p -y-.g9 4.- .s--- 9y,-t em ' gm mF-- '+- T---E'--9-+' efN*-*fmw,M'-N7-

D4.ll-3 (DEFICIENCY) FREQUENCY OF HVAC DUCT SUPPORTS DESCRIPTION The seismic design of HVAC duct supports in the Reactor Auxiliary Building is performed in accordance with Ebasco specification CAR-6418-AS02. This specification requires that the frequency of the supports must be 16 Hz or higher. This requirement applies to motion in two horizontal and one vertical directions. Ebasco calculated the frequency of the duct supports in only two directions: one horizontal and one vertical. The frequency in the horizontal direction along the axis of the HVAC duct was never calculated (Calculation Book G-171 S02). If the frequency of the duct supports in this direction is lower than 16 Hz, the design would be unconservative.

           ' RESPONSE HVAC duct supports were designed by analyses, similarity or by inspection /                                                           ,

judgement. Specific references defining the design method utilized were not in all cases provided. In general, the same group of experienced design engineers have been involved in the production effort of HVAC restraint analysis and design. This continuous participation in establishing and evaluating frequency characteristics and the state of stress of large numbers of supports, have provided the lead design engineers with the ability to predict expected characteristics of the support, without performing an analysis. Similarity to supports which were previously analyzed and found to be acceptable was used as a design basis. The proper references to similar and previously qualified supports or clarifying statements in support of engineering judgements used were not always provided. The approach described above was used in all HVAC duct support design. The adequacy of the present design for those supports that may not have a specific qualification analysis is not in question. The inherent conservatism in the approach and the experience and judgement of the design engineers insure a safe design. Nevertheless, a program to evaluate the calculations for all supports shown on sample drawings has been initiated. This program will identify supports which have not been specifically qualified by stated , similarity and either identify and document the similar cases, or perform and ! document the actual analysis. The selected sample drawings include approximately twenty-five percent of all HVAC duct restraints on the project.

j. The sample is random and unbiased in that all buildings and both congested and i uncongested areas are represented. If, as expected, all supports requiring qualification are shown to be adequate, no further corrective action will be initiated, i

Future supports will either be designed by specific analysis or will be

          - clearly referenced to a similar support which has been analyzed.

1 i L 1

    .-,m,        _ - _ - , - - . - , , .           _ _ _ _ , , , ___ ., _ ,_- ,, ,_,,. , ._. _   . _ _ , - _ _ . - _ _ _ _ _ _ . _ _ . - . _ . , .

7 i - ,

                                                                               )

US.1-1 (UNRESOLVED ITEM) WESTINGHOUSE GUIDANCE FOR NUCLEAR INSTRUMENT CABLES DESCRIPTION Westinghouse provides specific guidelines for the purchase and installation requirements for equipment required to monitor the nuclear reaction in -Westinghouse plants. The IDI team discovered that Ebasco considered the Westinghouse guidelines too conservative and therefore designed and installed the nuclear instrumentation conduit system differently than the Westinghouse guidelines for maximum conduit fill and restriction on the length and total angle between cable pull po,f nts. Paragraph 4.3 of Westinghouse Standard Section 4.1. specifically- requires that deviations from that standard be reviewed by Westinghouse and be considered acceptable only after approval. Ebasco informed Westinghouse of their intent not to follow two sections in their guidelines for (a) separation and (b) pull box size and location. Ebasco also told Westinghouse that they would use their standard procedures for cable pulling tensions and lengths. However, Ebasco could not produce evidence of Westinghouse approval. - The IDI team considered this a failure to follow Ebasco Procedure E-SH-52 ' which requires that interface be maintained between Ebasco and the nuclear steam system supplier (Westinghouse). -

RESPONSE

Ebasco had specified certain exceptions to the NIS Installation Standard 4.1 via letter ~to Westinghouse EB-W-1430 dated December 12, 1978. The final paragraph of that letter read: "Ebasco design is proceeding on the above basis. If you have any concerns regarding these comments, please advise hy Janua ry ' 31, 1979. " No response was received and design continued. We do not consider the above a failure to follow procedure E-SH-52. Although letter EB-W-1430 was written such that- Ebasco was to assume approval without requiring Westinghouse's written confirmation, we believe this process to be , i adequate and consistent with the intent of E-SH-52. Note: Ebasco contacted Westinghouse on the subject during the IDI. ' Subsequently, Ebasco received Westinghouse letter CQL-8608 dated February 8, 1985 accepting the previous comments. , 4

                                                         /
                /                                                                          f t

s 4 1

US.2-1 (UNRESOLVED ITEM) ELECTRICAL POWER DESIGN PROCEDURES & GUIDELINES DESCRIPTION HPES Electrical has been undertaking several design efforts based on informal guidelines. Although formal guidelines have been planned, none have been issued nor have been worked into the overall training of Electrical Engineering.

RESPONSE

Thirteen guidelines for use in the design effort of the Harris Plant Engineering Section (HPES) Electrical Unit are currently being prepared.

These guidelines, as listed below, have been identified as being required to provide guidance to HPES electrical personnel and envelope the present HPES electrical scope. Guidelines will be generated for Cable Sizing, Cable and Conduit List, FCR Resolution, Containment Electrical Penetrations, Communications, Conduit Seals, Protective Relay Setting and Coordination, Heat Tracing, Tray Covers, Load Monitoring (SLAP), Voltage and Short Circuit System Considerations (ASD0P), Containment Building Seismic Support of Conduit, and Radiation Monitoring Isolated Ground Bus. Additional procedures will also be developed as necessary and as the HPES Electrical scope expands. Once a guideline is issued, HPES electrical personnel will be trained accordingly.

U5.3-1 (UNRESOLVED ITEM) INDEPENDENCE OF ELECTRIC SYSTEMS DESCRIPTION Regulatory Guide 1.75, which endorses IEEE Standard 384-1974, indicates that the use of interrupting devices actuated only by fault current should not be used as devices for isolating non-Class 1E circuits from Class IE circuits. The FSAR, as modified by FSAR change E-162, commits to tripping the 6900-volt emergency buses 1 A-SA and 1B-SB feeder breakers (1 A1 A-SA and 181 A-SB) to non-safety motor control centers 1 A24 and 1824 in case of a LOCA or loss of offsite power consistent with Regulatory Guide 1.75. Review.of the'FSAR change request both at Ebasco and at the site did not result in a Design Change Request or Field Change Request to modify the circuits to comply with the requirements of the FSAR change. Additionally, no load analysis has been performed to ensure that the loads which are isolated do not jeopardize plant safety.

RESPONSE

Safety Analysis Report.(SAR) change notice F-162 was generated to modify the condition under which the 6.9KV non-safety busses would be isolated from their associated 6.9KV safety busses. Originally, breakers 1 A1 A-SA and 1Bl A-SB were tripped upon a Loss of Offsite Power (LOOP) or on loss of power followed by q LOCA only, since the loads on the non-safety busses were considered necessary ' only after a Loss of Coolant Accident (LOCA) for long term cooldown. Typical loads connected to this bus are RCP Motor Oil Pumps, Charging Pump Lube Oil Pump, Turbine Generator Bearing Oil Pump, Instrument Air Compressors and Battery Charger. Since the licensing position for the Shearon Harris Nuclear Plant is that the safe shutdown condition is " Hot Standby Mode", these loads are not required for this condition. The SAR change committed tripping these breakers upon a LOOP and LOCA (Safety Injection Signal). The need for this design modification had been identified prior to initiation of the IDI

   .although it had not been completed.

The control wiring diagram for the-6.9KV breaker which feeds the Non-Class lE 6900/480V station transformer has been modified so that it will be tripped on Loss of Offsite Power or initiation of an accident signal. Design Change Notices 251-532 and 251-533 have been issued to revise the existing design. The above discrepancy between the FSAR and the current design is considered an isolated case in light of the fact that the FSAR change notice was fonnally incorporated recently to comply with Reg Guide 1.75 position C1. L l 4' L__

D5.4-1 (DEFICIENCY) PROTECTION OF SAFETY RELATED BUSES DESCRIPTION The 480V Class lE power center buses l A3-SA and 183-SB are divided by a bus reactor into two sections: 3200A (50KA sym) and 1600A (22 KA sym). At present, each 1600A section is supplying four MCC's rated 600A, each through a circuit breaker rated 800A with an overcurrent device set at 720A (long-time trip). No procedure / control exists to monitor or control the loading on this bus section. This could allow the 1600A bus section to be loaded to 2880A on the basis of breaker settings rather than load monitoring. Additional MCC's can be supplied from spare breaker positions and further increase the bus overloads. The pick-up settings of the MCC breaker overcurrent devices (long-time trip) are considered high. Reduction of these settings can restrict bus loads to 1600A.

RESPONSE

The two power center buses cited in this deficiency are the only ones affected. To prevent any possibility of overloading the 1600A bus sections as a result of the pick-up settings of the MCC breaker overcurrent trip devices, the existing CP&L " System Load Analysis Program" (SLAP) will be revised to monitor the loading on these bus sections. The computer will " flag" when the total connected MCC loads reach 1600A. Approved procedures will require that any further proposed bus loading be analyzed by the Harris Plant Engineering section to ensure that the total actual load on this bus section does not exceed 1600A. In a similar manner, the SLAP program and respective procedures mentioned above will monitor the loading of all Class 1E power distribution buses. A Harris Plant Engineering Section guideline for relay protection is under development and all MCC feeder settings will be analyzed accordingly to the guideline. f l l l I

i

         .DS.4-2 (DEFICIENCY) MOTOR OPERATED VALVE THERMAL OVERLOAD SETTINGS
 !         DESCRIPTION
 ^
          -The start-up organization procedure 1/2-9000-E-01 Rev 5 " Motor Overload Relay Heater Selection and Molded Case Breaker Instantaneous Trip Setting for
;         Continuous Duty Motors" applies the method of heater selection for continuous duty motors supplied from 480V MCC's to both centinuous duty and to interinittent duty (MOV's) motors. The heater elements for M0V's which have been selected and installed so far have used an inappropriate procedure.

RESPONSE

The heater elements of only a small portion of the total number of M0V's have

         - been selected and installed using the procedure described above. The heater elements for the remaining M0V's have not been selected.

The Harris Plant Engineering Section will issue design criteria (Dwg CAR-2166-B-041 SH4AB)for heater selection for intermittent duty motors (M0V's)

which will incorporate the vendor (Limitorque Co.) recommendations and '

i engineering practices. The start-up organization will select the heater elements of all MOV's by following the method outlined in the above design j . cri teria. I The start-up organization will also revise procedure 1/2-9000-E-06 " Motor Driven Valve Operators". This procedure will follow the method outlined in the Harris Plant Engineering Section design criteria. { I I l . - . - -

i 05.4-3 (DEFICIENCY)' DESIGN VERIFICATION OF THERMAL OVERLOAD SETTINGS DESCRIPTION In the data sheets of the start-up organization's procedure 1/2-9000-E-01 Rev 5 (see IDI Item D5.4-2) which record the selection of heater elements for continuous duty motors, both the " performed by" and " reviewed by" spaces were

     . signed by the same person. The selection of the heater elements has not been verified.

RESPONSE

Only five continuous duty motors are cited for this deficiency. However, all ! continuous duty motors supplied from 480V MCC's have their heater elements already selected and are therefore affected. The start-up organization will revise procedure 1/2-9000-E-01 " Motor Overload Heater Selection and Molded Case Circuit Breaker Instantaneous Trip Setting Guidelines for Continuous Duty Motors". The data sheet in this procedure will require an independent reviewer of the selected heaters. All continuous duty motors supplied from 480V MCC's will have their heater elements selected and ! verified independently. . In a similar manner, the data sheets in the start-up organization's new

procedure 1/2-9000-E-06 " Motor Driven Yalva Operators" (see IDI Item D5.4-2) will require an independent verification of the selected heater elements.

l i

DS.4-4 (DEFICIENCY) STATION SERVICE TRANSFORMER PROTECTIVE RELAYING DESCRIPTION The Class 1E secondary unit substation transformers (rated 6.9KV/480Y, 2000 KVA AA; self-cooled rated current of 167A at 6.9KY or 2405A at 480V) are provided each with a low side and a high side circuit breaker and protective The pick-up of the low side overcurrent trip device overcurrent trip (devices. is set at 3200A 133% of rated current) and that of the high side at 300A (180% of rated current). No procedure / program exists to monitor or control the loading of the transformers (or corresponding PC bus). The pick-up settings of the protective overcurrent trip devices are considered to be unacceptably high since they may allow the transformers to be overloaded continuously or for long periods of time and thus be detrimental to transformer's life.

RESPONSE

There are four Class 1E Secondary Unit Substation Transformers affected. To prevent any possibility of overloading the transformer allowed by the pick-up setting of the low side overcurrent trip device, the existing CP&L

 " System Load Analysis Program" (SLAP) will be revised to monitor the loading on every 480V Class 1E Power Center (PC) bus. The computer will " flag" when      _~

the total connected PC bus load reaches the transformer qualified rating of 1915 KVA at 300C average ambient (See Environmental Qualification Report 33-52994-QT). Appropriate procedures will require that any further proposed bus loading be analyzed by the Harris Plant Engineering Section to ensure that the total actual Power Center bus load does not exceed the transformer qualified rating. L The existing settings on the transformer breakers phase overcurrent trip devices are in accordance with industry standards (See ANSI C37.91-1972 "IEEE Guide for Protective Relay Application to Power Transformers" and ANSI Appendix C57.96 (1959) " Guide for loading Dry-Type Distribution and Power Transformers). However, in order to further restrict the possibility of overloading the transformer, the pick-up settings will be modified to closer match the transformer qualified rated current. The pick-up setting of the low i side solid state trip device (long-time) will be lowered to 2880A (125% of ! qualified rated amps) and that of the high side overcurrent relay to 240VA (150% of the qualified rated amps). r

U5.4-5 (UNRESOLVED ITEM) PROCUREMENT OF QUALITY COMPONENTS DESCRIPTION Terminal boxes purchased by CP&L for use in Class 1E applications were purchased as non-safety related corr.ponents with no apparent certification of quality required.

RESPONSE

Boxes purchased by the site were ordered either by a site specification or, where a manufacturer was specified, by catalog number (no specification). Those ordered with the site specification required a certificate of compliance to be provided. No certificate of compliance was required by the purchase orders for boxes ordered by catalog number. A certificate of compliance should have been required to assure the boxes meet the steel thickness requirements upon which the design criteria were based. To resolve this item, a study will be undertaken to determine all types of boxes that have been purchased without a certificate of compliance and which may have been used in Class 1E applications. A sampling of these boxes will be conducted to verify that the boxes meet design requirements.

D5.5-1 (DEFICIENCY) BATTERY SIZING CALCULATION DESCRIPTION The inputs to the battery sizing load tabulation (Ebasco Calculation No. 52-APM) consisted of eleven separate load categories. The IDI team noted errors in eight of the eleven inputs. Three of these errors resulted in

        ~1ower-than-required currents being used in the tabulation.

1. Inserter load used a 1976 preliminary input from Westinghouse instead of a later 1980 inverter load calculation.

2. Diesel generator de auxiliary loads did not agree with the attachment to the calculation.

3. All five dc motor currents were calculated in error through misinterpretation of the Engineering Guide. In addition, the auxiliary -

feedwater-isolation valve motor load was applied in the first minute period, instead of as a less (sic.) conservative random load at the end i of discharge, without justification. The feedwater isolation signal is not depandent upon loss of offsite power.

RESPONSE

As a result of recent design changes to various plant systems, a general update of de system load studies had been initiated to confirm de system adequacy. Part of this update includes a review of electrical parameters that are-used as input to the de system calculations. This de system update was initiated before the start of this Inspection and has just recently been , completed confirming existing battery size. Calculation No. 52-AMM has been superceded by Calculations 56-JRG and 57-JRG. These calculations address and incorporate the concerns specified in 1) and 2) , above as well as address the misinterpretation of the Engineering Guide regarding the detemination of a current value for de motors. Applying the auxiliary feedwater isolation valve motor load during the first minute of the battery profile is actually the most convervative approach. In the battery load profile, the design basis of loss of offsite power concurrent with an accident'is followed by assuming that the auxiliary feedwater isolation valve will operate in the first minute. l l . - . . . - . -. - . . - - - - - . - - .- _ - -. _. - - -

05.5-2 (DEFICIENCY) DC EQUIPMENT RATED MAXIMUM YOLTAGE DESCRIPTION Equipment must be designed to operate over the expected operating voltage range. FSAR Section 8.3.2.12 states that all loads can withstand 140V dc during battery equalization. This subject was also the topic of I&E Information Notice 83-08. No maximum voltage was specified or provided by the vendor for the Potter Brumfield MDR relays and the Electroswitch LSR relays in response to specification for the auxiliary relay cabinet and the transfer panel. The Harris Nuclear Plant utilizes lead calcium cells, which require periodic equalizing charges. This type of charge should only be required following the periodic discharge tests of the entire battery. High voltage during the system equalize period has not been addressed, nor could it adequately be addressed, because no attempt was made at the site to determine the Class lE de equipment maximum voltage ratings. At the close of the Integrated Design Inspection, the site operations technical support staff was attempting to address this problem by proposing to equalize cells on an individual cell basis.

RESPONSE

The above concern addresses the lack of vendor documentation to insure that equipment nominally rated for 125V de can tolerate 140V dc (the voltage reached when equalizing the battery subsequent to discharge). Below is a list of alternatives (listed by priority) that may be taken in order to insure that the upper voltage limits of the de system is within the tolerances of the equipment:

1) Verify (in writing from the equipment maufacturers) that all de equipment will meet the requirements of the specification.

Equipment that cannot withstand the maximum applied de voltage will be replaced.

2) Review the alternatives to equalizing the battery at a value of 140V dc.

3) Develop plant procedures to equalize cells on an individual cell basis.

Alternative 1 has been initiated as part of the de system update referenced in D5.5-1. While this alternate is the preferred method, the other options, although less desirable, may need to be considered if the Alternative 1 solution is not viable. Although the Potter Brumfield MDR and Electroswitch LSR relays are specifically mentioned, all de equipment will be reviewed to insure the above concern is properly addressed. This effort will also close out concerns associated with I&E Information Notice 83-08.

D5.5-3 (DEFICIENCY) BATTERY DISCHARGE VOLTAGE PROFILE DESCRIPTION. During the calculation of permissible voltage drop in the de power cables between the battery and the de switchgear, Ebasco calculated the battery discharge voltage profile based upon the battery load currents available at that time. The battery sizing calculation is based upon these currents and is corrected-for minimum electrolyte temperature, design margin and aging margin. In calculating the battery discharge voltage, Ebasco correctly derated the available battery capacity to that which would be remaining at the end of battery life, defined by IEEE 450 as 80% of battery capacity. Although Ebasco calculated the battery size using derating factors for a minimum electrolyte temperature of 70 degrees F and design margin of 15%, they failed to include these factors in the battery voltage profile calculation. By not including the correction factor for minimum battery electrolyte temperature in the voltage profile calculation, an error results because the nominal battery capacity is 4% highcr than what would actually be available at 70 degrees F. , Also, by not including the correction for design margin, the battery discharge voltage calculation (and all resulting voltage drop calculations based upon actual battery voltage during discharge) must now be repeated every time the battery discharge profile changes.

RESPONSE

As stated in D5.5-1, the entire dc system is presently being reviewed to reflect the latest design changes. As part of this effort, Calculation No. 43-SKD has been superseded by Calculation No. 20-WRE. The latter calculation includes aging, temperature correction, and a design margin when calculating the battery discharge voltage. The results of this calculation were then used to insure that sufficient voltage is available to properly operate all de safety related equipment. l l i ( l l l I I

   .. . - . . _ _ , , , _ _ - . - , _ _ _ ~ _ _ _ _ _ _ _ _ _ _ .               - . . . . , - _ - . _ _ , _ _ _ _ = _ _ . _ _ , , , , _ _ _ - - . _ _ _ . _ . . . , - - - _ . . .

D5.5-4 (DEFICIENCY) DC SYSTEM MINIMUM YOLTAGE DESCRIPTION In performing the calculation to establish. the maximum resistance permitted in the switchgear close and trip circuits, Ebasco project personnel failed to follow the Electrical Engineering Guide No. S18-62 which states that the battery minimum discharge voltage (105 Ydc) should be used in this type calculation. Instead, Ebasco project personnel calculated the trip circuit maximum resistance based upon the battery voltage at the end of the first minute of battery discharge (113 Vdc) and the close' circuit maximum resistance based upon the nominal battery voltage (125 Vdc). This resulted in values of circuit resistance which are not conservative. Safety related loads are added to the diesel generator during a battery discharge, therefore, the close circuits for those safety related loads must be designed to operate with the reduced voltage of 113 volts during the first minute battery discharge. Also, the battery sizing calculation identifies a number of breakers closing in the last minute in an attempt to re-establish the power supply. The battery discharge voltage profile developed as part of the calculation for the de panel /switchgear feeder voltage drop determined that the voltage in the last minute of the design discharge was 116 volts.

RESPONSE

One of the objectives of Electrical calculation 44-SKD is to determine the allowable loop length resistance for closing and tripping circuits for both the 480V and 6.9kV switchgear. An assumption was made that the battery charger was available when closing all . breakers. As part of the de system review effort discc. sed in D5.5-1, Electrical calculation 20-WRE has superseded calculation 44-SKD. This calculation considers the battery terminal voltage for the worst case condition and calculates the de voltage at each switchgear. This worst case voltage value is then used as an input in determining the allowable loop resistance for both closing and tripping circuits. The results of calculation 20-WRE affect the voltage drop calculations for all safety related switchgear close and trip circuits. Each circuit has been reviewed against the new calculation to ensure sufficient voltage is available at the close and trip circuits of. the respective circuit breakers. l i i e L

4 i

D5.5-5 (DEFICIENCY) DC SYSTEM CONTROL ROOM INDICATION DESCRIPTION 4

IEEE Standard 308 describes the principal design criteria for the Class 1E , electric system. As part of this document, basic criteria are given for the t surveillance of the direct current system including the battery chargers. In , response to these criteria, the Harris design for the direct current i instrumentation is documented in FSAR Section 8. 3. 2. 2.1. 4. This section states that the availability of the Class 1E battery is monitored in the control room by board mounted indication of (a) bus voltage and (b) battery . and battery charger current. i The IDI team determined that, contrary to the commitment in the FSAR, no indication is provided on the control board for battery charger current. Although the main control board front view of Panel Area D1 calls for indication of battery charger output current, the team detemined that the dc i system, including the circuitry provided in the battery charger as presently 4

                     - installed, does not have the provisions for remote indication of battery charger current. The IDI team also determined that this parameter was not included in the proposed dc system monitoring for the Emergency Response

Facility Information System (ERFIS).

< RESPONSE IEEE 308-1974 provides general requirements for de system monitoring. More i specifically Section 5.3.4(5) of IEEE 308-1974 requires that indication shall be provided for the battery charger output current to monitor the status of the charger. Remote, as opposed to local, indication is not specified. l The dc system monitoring for Shearon Harris Nuclear Project includes an i ammeter in the charger output. In addition, a common de system trouble alarm is provided in the control room which includes an alarm condition for abnormal

battery charger output current. This annunciation alerts a plant operator in the Main Control Room of a malfunction involving the battery charger which can

then be further investigated.

, A Safety Analysis Report (SAR) Change Notice (F-93) was prepared with the , l intention to revise the FSAR to reflect only local indication for the battery (- charger output current. F-93 was used as input to respond to NRC question ! 430.116. However, the wording of F-93 was ambiguous and the official response

to 430.116 did not reflect the intended revision.

l ' In the original design stage of the Emergency Response Facility Infomation L _ System (ERFIS), this parameter was considered as an input. Since no licensing F commitment was ever made regarding this input to the ERFIS, the final design , elected not to include this parameter. The appropriate cection (8.3.2) of the FSAR will be revised for consistency i with the intended design stated above. i l i l-i

   - , - . - - , - - , ,    ,-,---nm,-n-a.~._-                n,,-,-n,m_nn   -,-.,n,                ,     , , - . w n,-  -,n-

05.5-6 (0BSERVATION) DC SYSTEM UNDERVOLTAGE ALARM DESCRIPTION A common de system trouble alarm exists in the control room which sounds for either bus undervoltage alarm or battery charger trouble. The battery charger undervoltage alarm is set at 130 volts. The de bus undervoltage alarm is set at 121 volts. There is no ringback capability on this alarm window which would sound if a second input went off-normal. During a battery discharge, the alarm will' sound simultaneously from both sources as the de system voltage drops from the float voltage of 135 volts to below 120 volts at the start of the discharge. The operator must then periodically consult the bus voltmeter located on the control board to monitor the bus voltage. If the bus undervoltage relay was set lower (at 110 volts) and able to ring back the de trouble alarm, the operator would be aware that the battery is nearing the design discharge voltage and remind him to shed additional load. This load shedding would result in restoration of the battery voltage to permit

     . continued operation of the remaining de equipment.

RESPONSE

In addition to having the dc bus undervoltage relay annunciate at the MCB, via the common de trouble alarm, this relay is also wired to a separate annunciator at the Emergency Safety Feature (ESF) light box. This allows the operator to identify an undervoltage condition at the dc bus. We agree that the intent of the design should not be to alarm de bus undervoltage during the first minute of battery discharge. The relay settings will be reviewed to ensure that they achieve the desired objectives. l l { r + L L

D5.6-1 (DEFICIENCY) PENETRATION PROTECTION QUALIFICATION DESCRIPTION This concern stems from the inability to trip the Reactor Coolant Pump (RCP) backup breaker when the pump motor breaker fails to trip upon an overcurrent condition, coincident with a loss of the one train 125V de control power to the motor feeder breaker. Each of the three reactor coolant pump motor circuits, which are the only 6.9KV circuits penetrating the containment, is provided with overcurrent protection designed to trip the motor feeder breaker. If current still flows in the feeder after a preset time (i.e., the feeder breaker fails to trip), a transfer trip signal is given to the upstream breakers (bus main breakers) feeding the bus. As stated in FSAR Section 8.3.1 the feeder breaker to the Reactor Coolant Pump and the breakers feeding the bus are supplied with separate sources of de control power. Since the RCP feeder breaker tripping circuit and the breaker failure transfer circuit are both connected to the same safety related 125V de control source via a common lockout relay, its failure will prevent tripping the bus nain breakers.

RESPONSE

The design philosophy of the overall primary and backup scheme was intended to meet both Class 1E and Non Class 1E isolation and penetration protection requirements. In accordance with FSAR Section 8.3.1, the need to trip the feeder breaker and the associated bus main breakers from different 125V de control sources has been implemented. As part of the general circuit design philosophy, the individual circuits were designed to alarm upon loss of control power in lieu of tripping in order to enchance overall system reliability and minimize potential plant upsets. In the case of the reactor coolant pump breaker control, an " energized" to operate in lieu of a "de-energize" to operate lockout relay was selected. In order to fully comply with the design commitments, the following scheme will be implemented. 1- Two sets of overcurrent (0/C) relays for primary protection, set for proper time and instantaneous operation will independently trip each of the redundant breaker trip coils. These 0/C relays will activate time delay relays through the breaker failure relays which are set to operate at current values below the motor full load current.

1 D5.6-1 (DEFICIENCY) PENETRATION PROTECTION QUALIFICATION (Cont'd) 4

RESPONSE (Cont'd) 2- The primary 0/C relays, the breaker failure relays and the associated time delay relays will be powered on a two train basis, such that, loss of one 125V de control source (Train A or Train B) will not prevent the

, operation of the breaker failure transfer scheme and therefore will energize the bus main breaker trip circuit if required. Therefore, the 6.9KV containment electrical penetration primary and backup protection circuits will be designed to provide the independent penetration protection required by Regulatory Guide 1.63. I e i 1

D5.7-1 (DEFICIENCY) USE OF MOTOR DATA IN SETTING PROCEDURE DESCRIPTION The Harris Plant Engineering Section did not use specific vendor data for motor starting times and safe stall times when it calculated the relay settings and coordination curve of the large 460V Class 1E motors (supplied from 480V Class 1E Power Centers). Instead all calculations were based on assumed motor starting time. Safe stall times were not considered in the calculations. Furthermore, no procedure existed for relay settings and calculations.

RESPONSE

There are a total of four large 460V Class 1E motors per safety load train and all are similarly affected. The corresponding motor vendors have provided starting times and the safe stall time for all of the four motors. The relay setting calculations and coordination curves for the motor circuits have been updated and will be verified. Each motor's relay calculations are incorporated in a standard relay calculation and data sheet as required by the Harris Plant Engineering Section guideline No. 7.5.G for relay protection, which is under development. Accordingly, any missing assumptions will be identified and verified.

D5.7-2 (DEFICIENCY) 480V BUS UNDERVOLTAGE ALARM DESCRIPTION According to the project load study of 1982, the minimum allowable steady state voltage of the 480V Class 1E Power Center Bus 183-SB (supplying MCC loads only) was 428V (i.e. 90% of motor rated voltage or 414V plus 3% for cable voltage drop of MCC motors to the PC bus). The calculated bus ur.dervoltage relay settings are: drop out of 100V (correspondong to 400V bus voltage) and a time dial position 4. However the relay settings for all 480V Power Center (PC) buses are incorrectly entered in the relay setting drawings as 80V and time dias position 1 instead of the above calculated settings.

RESPONSE

There are total of four 480V Class lE Power Center Buses and all are affected. The latest project voltage study, " Adequacy of Station Electric Distribution System Voltages" for compliance with BTP PSB-1 of February,1985 which was not available at the time of the IDI inspection, revises the minimum allowable steady state voltage for PC buses 1 A3-SA and 183-SB (both supplying MCC loads only) to 444V and that for PC buses lA2-SA and 182-SB (supplying motor loads only) to 424V. Considering that the preferred drop out setting of the bus undervoltage relay is just below the bus minimum allowable steady state voltage, bus PT ratio (480/120V), and the relay available taps (60,70,80,1008110V - not continuously adjustable) the relay dropout setting of 100V was acceptable. However, in view of the revised values for the minimum allowable steady state voltages for the 480V Class 1E PC buses per the voltage study of 1985, the dropout settings of the undervoltage relays for buses 1 A3-SA and 183-SB will be revised to 110V. The relay dropout settings of 100V for PC buses 1 A2-SA and 183-SB are considered acceptable. The time delay settings on the undervoltage relays of all Class 1E PC buses will be reviewed to ensure avoiding nuisance relay operation under any transient undervoltage conditions. l The relay setting drawings for all 480V PC buses will be revised to indicate the correct bus undervoltage relay settings. The Harris Plant Engineering Section is developing design guidelines for relay I protection including that for all 480V PC buses undervoltage relays. i 1

05.7-3 (0BSERVATION) MOTOR ACCEPTANCE TESTING DESCRIPTION At the time of the IDI the vendors had not yet furnished the specific values for starting times and safe stall times for all large 460V Class 1E motors which are required for the protective relay calculations. Harris Plant Engineering Section had performed the relay calculations by using assumed data (See IDI Item D5.7-1). The start-up organization's existing procedure 1/2-9000-E-05 Rev 4 " Initial Checkout of Electric Motors" does not require tests to determine the starting currents and starting times of motors for use, if necessary, by the Harris Plant Engineering Section in relay calculations. Also, this procedure does not call for recording the voltage when measuring the motor running current under both the unloaded and loaded conditions.

RESPONSE

Since the IDI, all large 460V Class lE motors have been provided with specific vendor data (See item D5.7-1). Motors provided with documented vendor starting currents and starting times do not require an additional verification of these data by means of motor testing. However, the Start-up Organization will issue a new Procedure 1-9000-E-19 " Motor Starting Time and Starting Current Testing", and only large 460V and 6.6kV motors for which specific starting times and/or starting currents can not be obtained from the vendor will require additional tests per above procedure. Future motors requiring these tests will be identified and the tests requested by the Harris Plant Engineering Section. The Start-up organization will also revise Procedure 1/2-9000-E-05 to require the bus voltage to be recorded when measuring the motor running current under both the unloaded conditions.

D5.8-1 (DEFICIENCY) DC MOTOR OPERATED VALVE VOLTAGE DROP . DESCRIPTION

                             .The cable size for de valve 2AFVll8 was selected without regard to voltage
                           . drop in the circuit cables. The IDI team assumed a battery voltage of 115 volts and independently calculated that the starting voltage drop in this circuit would be greater than 65 volts, resulting in a voltage at the motor of less than 50 volts. No minimum acceptable starting voltage was obtained from the vendor. This low voltage may prevent the valve from performing its safety function. Ebasco Criterion No.18 presents a typical method to determine the maximum circuit resistance permitted for different types of circuits. Ebasco i                              failed to extend the philosophy of Criterion No.18 to other loads not specifically detailed in the criterion, such as de motor operated valves.

RESPONSE

_ Originally,- the source of power to the motor operated valve 2AF-V118-SA-1 was 480 volt 3 phase ac. Based on the motor rating, the associated power cable '

(#10 AWG) was sized in accordance with the project cable sizing criteria.

Subsequent to this, the motor for the valve was changed from ac to dc. With the above change, the cable size that was used for the ac motor was i inadvertently translated to the dc motor circuit and was not verified against the appropriate criteria.

                          - As a result of recent design changes within various plant systems for this

! project, it became necessary to update both the ac and de system load studies to verify system adequacy. Part of this update includes review of all 1 electrical equipment parameters that are used as input to the load studies and i associated calculations. Specifically, for the valve in question, its power i supply cable selection and other pertinent electrical parameters are included ! in this effort. At the time of the inspection, this effort had not been ! completed. Although only MOV-2AF-Vil8-SA-1 has been specifically cited for

insufficient voltage, other de M0V's are involved. Therefore, subsequent to this Inspection the cable size for all .seven dc MOV's was verified against the i cable criteria to ensure that sufficient voltage was available at the motor.

l The cable size has since been revised to meet the criteria. l l i s l

D5.8-2 (DEFICIENCY) CONTROL CIRCUIT VOLTAGE DROP 1 DESCRIPTION Calculation No. 44SKD was developed based upon current drawn by typical components such as Westinghouse relays and ASCO solenoids. Electroswitch relays are used in the design of the transfer panels and auxiliary transfer

                                       . control panels to transfer control from the control room to the auxiliary control panel. These relays, which draw up to 6.6 amperes during the transfer, were not included in the design basis calculation. The control wiring diagram identifies a minimum of 19 such relays which transfer at the same time. This circuit was wired with AWG No.10 conductors according to the Ebasco cable report. This cable size was specified by the Harris Plant Engineering Section electrical unit. The IDI team determined that the minimum

. voltage required to operate these-relays was 90 volts and independently i calculated a voltage drop in the circuit of greater than 60 volts. This would require that 150 volts 4xist at the de bus. FSAR Section 8.3.2 indicates that the maximum voltage available would be less than 140 volts with the battery on 3 equalize charge and 105 volts with the battery at minimum design discharge. Therefore, successful transfer of control from the control room to the auxiliary control panel may not be possible.

RESPONSE

,                                          Upon initiation of a transfer from the main control room to the auxiliary j                                           control panel, a master relay initiates operation of multiple slave relays.

While operating, these slave relays draw a momentary current of 6.6 amps for approximately 35 milliseconds. The total momentary current drawn by these i relays approaches 125 amps. We concur that if all the relays operated during this 35ms period the. voltage drop in the #10AWG cable could be such that the ' voltage across the relay might be less than the minimum relay operating voltage. While we believe this to be improbable the specific cables associated with this circuit have been revised from a #10AWG to a #2AWG. The above condition is considered unique since no other panel or cabinet in the plant is designed with this master-slave relay arrangement where the number of slave relays (approximately 28) causes a momentary inrush current of j this magnitude. l l

  . - . . - . _ _ . . . . . . . . . - , . . , _ , , , . _ . _ . _ _ , _ . . _ _ . - _ ,           .._---,__.,,__.__,.m,,           .-,_..,___m-,,y,      .,____.,m._-,,.. m,-.

I 1 I D5.8-3 (DEFICIENCY) REACTOR COOLANT PUMP POWER CABLE VOLTAGE DROP CALCULATIONS ' t DESCRIPTION Ebasco Engineering Procedure E-46 specifies requirements for conducting an auxiliary electrical system load study. Guidance for conducting this study is ! provided by Ebasco Engineering Guide S104-5. This analysis is presently

conducted using Ebasco Computer Program V2AUXSYS2027. This program calculates

the voltage conditions for various configurations of operating equipment, and is updated on an as needed basis. Input data for the computer program are 4

provided from vendor-supplied data via the motor and electrical load list.

The analysis which was conducted for.the reactor coolant pumps includes two

. . operating conditions: motor starting (7000 hp) and normal full-load running (6302 hp). However, the information contained on the motor and electrical

load list gives the value for normal full-load as 7121 hp, provides no data for motor starting load, and lists a third value for pump running with a cold I plant as 9290 hp. The IDI team was informed that Westinghouse provided two 1

different sets of pump data, one via a letter of transmittal and the other in the technical manuals submitted as required by contract. The existing

                   -calculation did not use the most limiting pump data available. There was no

assurance that either set of data was correct. The accuracy of the motor and

electrical load list was in question, and the discrepancy concerning the i

available pump data had not been resolved. Additionally, the data contained

on the motor and electrical load list have not been properly controlled

between and within design organizations.

l RESPONSE As stated in the " Post Cutoff" section of the IDI report, Westinghouse has . submitted a letter (CQL-8543 dated 1/11/85) providing verified brake l horsepower values for the Reactor Coolant Pump Motors for both a start-up and running condition. An electrical auxiliary system study was perfonned and the results verified the acceptability of the electrical system design. The motor and electrical load list is not a controlled document and as such cannot be used as a design input to the electrical auxiliary system program . V2AUXSYS2027. Input to this document is supplied from vendor data. The motor , ! and electrical load list is a supplemental technical document used in the l preliminary stages of a project to establish bus voltage levels and short

circuit ratings.

i , Because the IDI team reviewed other motor loads and no other discrepancies

were identified the above finding is not considered systematic.

f ( t

t 4-D5.9-1 (DEFICIENCY) REACTOR VES ZL LEVEL INSTRUMENTATION SYSTEM RCP INPUTS I DESCRIPTION-The Reactor Vessel Level Instrumentation System (RVLIS) is a redundant system which determines the level of coolant in the reactor vessel. This level detemination is dependent upon reactor coolant pump operation. Therefore,

              -the system requires independent inputs to each redundant system from the
              -reactor coolant pump breakers.

RESPONSE

i

The RVLIS system is used to satisfy the FSAR commitments for the Inadequate i

Core Cooling (ICC) Monitoring for Reg Guide 1.97 compliance. The non-safety 4 inputs are the Reactor Coolant Pump (RCP) breaker status (run vs. stop) which are used via non-safety to safety electrical isolation. Since the ICC is classified as a Category I variable, two redundant safety trains are provided. Therefore, inputs from each of the three RCP breakers must be used

in each RVLIS train. When the RVLIS sytem was initially implemented, a sufficient number of isolation (non-safety to safety) relays were not available. In light af this, RCP breaker status was input to an available

, A-train non-safety to safety isolation relay. From this point, the input was l- ' split; one signal going directly to the A-train RVLIS, a redundant signal

              . going via an isolation relay (safety train to safety train) to the B-train RVLIS. The use of a safety related isolation relay powered from a safety battery isolation relay for input multiplication does not degrade system reliability inasmuch as the RCP breaker is a non-safety device powered from a non-class 1E source.

Since it is not possible to provide non-safety inputs which satisfy single failure criteria and it is not a design requirement, this does not represent a deficiency in design. h Subsequent to the IDI additional quantities of isolation releys became available due to an' unrelated design change. This enabled enhancement of the , i circuit design to provide the IDI recommended independence. Control Wiring j Diagram CAR-2166-B-401, Sheet 140, will be revised to provide separate RCP ~ status input to each RVLIS train. Design Change Notice 251-527 has been ! issued to implement this change. i l i

I D5.10-1 (DEFICIENCY) SITE ENGINEERING DESIGN CHANGE CONTROL Two 125 volt de circuits for auxiliary transfer panels used two conductor #10 AWG cables connected to 100 ampere circuit breakers (Cables 10821 A). The cables should have been limited to 24 amperes. Auxiliary panel design was done by Ebasco via DCN-251-200, but breaker sizing and cable selection and routing was done by the Harris Plant Engineering Section (HPES) Electrical Unit without ampacity calculations, analysis to determine circuit current requirements, nor evidence of the basis for breaker selection.

RESPONSE

Complete load information was unknown at the time the cable routing effort was initiated. An arbitrary cable size was selected to assure an adequate route for the cables in question; however, the cable card was inadvertently released to the field prior to final verification. A review of other cables associated with this work (DCN-251-200) will be performed. We have issued a field change request to correct the cable size of the identified cables above and changeout , has been completed. No change in breaker size is required. We are preparing a design guideline to place adequate control and verification on cables and breakers sized by HPES. Ampacity calculations, circuit analysis, and breaker selection criteria will be included. N ! I I i

l l D6.1 -1 l ! (DEFICIENCY) INSTRUMENT LIST DATA BASE l DESCRIPTION i l The instrument list and the setpoint document are controlled, common data base

drawings which provide a comprehensive listing of instrument tag numbers and related radiation class, seismic class, setpoint and range, vendor print references and other information. '

Errors and inconsistencies in the instrument list have not been corrected through the design process. The instrument list is a commonly used reference; i for example, computer sorts have been used for design reviews which require - the data base to be accurate. The following errors and inconsistencies were noted by the IDI team in the data base. i 1. Reactor Coolant Pump A bearing water temperature element 1TE-131 was j licted incorrectly as 1TE-132. , 2. Reactor Coolant Pump A seal water flow transmitter 1FT0130 was listed ! with a range of 0-100 inches of water column instead of the correct 0-178 j inches of water column. i

3 Letdown valve hand control 1HC-0142.1 was input into the list as part of

the residual heat removal instead of the chemical and volume control j system.

4. Reactor Coolant Pump B seal differential pressure indicator PI-155A1 and i anion bed domineralizer differential pressure gauge did not contain the j "W" that EBASCO uses to designate Westinghouse supplied instrumentation.

EBASCO personnel stated that the "W" designation may be used to provide a j computer sort for reviews or studies.

5. Rector Coolant system wide range pressure transmitter 1PT-403SB was shown l as 0-300 psig instead of 0-3000 psig.

! 6 EBASCO has not consistently input classifications (pipe, seismic, i radiation, safety) into the instrument list. The following are typical ! examples of inconsistencies in documentation of instrumentation , classification (ISC codes) noted by the team. Auxiliary Feedwater Pump ! interlock relay 1AF-1158-CR4 in the auxiliary relay panel has the IAC l code but the Steam Generator 1A blowdown time delay relay 1BD-119402-1 which is also in the auxiliary relay panel does not have the IAC codes documented. Numerous other examples were noted by the IDI team.

D6.1 -1 (DEFICIENCY) INSTRUMENT LIST DATA BASE (Cont'd)

7. The data base entries for reactor coolant pump seal bypass and leakoff flow elements 1FE156,1FE155,1FE154,1FE1548,1FE1558, and 1FE156B were not updated to show the current Westinghouse supplied design.

8. Component cooling heat exchanger 1B-SB outlet flow transmitter 1FT-0653 was shown as a differential pressure of 0-3250 inches of water column instead of 0-400 inches of water column.

RESPONSE

The Instrument List contains over 26,000 instruments and control devices, each having more than twenty pieces of information. The Instrument List provides a comprehensive compilation that aids a user in quickly locating pertinent sources of information such as applicable specification, vendor drawings, I control wiring diagrams and installation details. The IDI team identified I approximately 25 cases of missing or incorrect inputs. Although valid, these i discrepancies were minor in nature and did not affect design or installation. , This information was inadvertently omitted or incorrectly transposed from  ! input data sheets. The following is a detailed evaluation of the aforementioned errors or i inconsistencies.

1. Although temperature element ITE-131 was incorrectly identified under

       " vendor tag" column as 1TE-132, the Ebasco tag number designation was correctly noted at TEIRC131. This is a random error.

2. The measurement range 0-178 in. W.C. for IFT0130 was incorrectly transposed from the Westinghouse data sheets. This is a random error.

3. The system designation (RH) applied to letdown valve hand control was inconsistent with the piping system for CVCS (CS). This is a random error.

4. The use of a "W"'in the Ebasco tag number for Westinghouse supplied field equipment is intended only to highlight that fact. If a review or study of the Westinghouse field equipment is ever deemed necessary, a sort of the instrument list would only be an aid since the same source information is available from the Westinghouse data sheets. The correct Westinghouse tag number IPI-155A1 was indicated. This is a random error.

D6.1-1 (DEFICIENCY) INSTRUMENT LIST DATA BASE (Cont'd) -

5. Although the range (0-300 psig) was incorrectly listed for 1PT-403SB, the redundant transmitter,1PT-402SA, was correctly listed (0-3000 psig).

This was a random input error. ,- _6. Ebssco is responsible only for inputting the I&C Classification Code for instrumentation or control devices that were processed under specific Ebasco generated specifications. The instrument list contains tag numbers for items supplied by Ebasco and "Others". As such, the Classification Code does not apply to items by Others. For the two cases cited, both relays, Ebasco had procured these items in bulk quantities that were not originally in the instrument list. They were added for inventory control and setpoint documentation purposes. Ebasco has been and still is in the process of completing all necessary information for

                   ,, relays in general.                                      '

7. The information on the data base for the R.C. pump seal bypass and

          ,           leakoff elements had not been updated to Westinghouse supplied design because Westinghouse had not submitted revised instrument data sheets.

Ebasco had previously verbally requested this information (prior to the inspection cutoff) but Westinghouse had not responded until 2/85 - CQL 8603 and 8605. Since the instrument list was consistent with the data

              ,.      sheets current at the time of the inspection, this is not an error.
             '8.      The range for 1FT-0653 was shown correctly in revision 36 of the instrument list which was reviewed. Additionally, the redundant safety train transmitter 1FT-0652 was correctly shown (0-400 In. W.C. ).

Therefore, this is not an error. Since the instrument list is a controlled drawing, any inconsistencies or errors will be noted and changed by a controlled process (Field Change Request or Design Change Notices). Design Change Notice DCN-251-536 has been issued to correct the noted discrepancies. Additionally, as part of the start up procedure preparation and pre-operational check-out, the Shearon llarris Nuclear Project is further screening the instrument list and issuing FCRs as required.

   'i l

d

D6.1-2 (DEFICIENCY) FSAR/ INSTRUMENT INDEX CONSISTENCY DESCRIPTION A comparison was made among the Instrument Index Purchase Order Specifications, and the FSAR for a sampling of instrumentation ranges and setpoints. The following discrepancies and inconsistencies were noted by the IDI team:

1. The Instrument Index entry and purchase order specification for containment temperature indicator TI741SB shows a range of 0-400 degrees F. FSAR Table 7.5.1-1 shows a range of 0-250 degrees F. FSAR Table 7.5.1-1 shows a range of 0-250 degrees F.

2. FSAR Table 7.5.1-2 shows a range of 0-226 kpph (1000 lb per hour) for auxiliary feedwater flow indicators FI-2050 A1, B1, C1. FSAR Table 7.4.1-1 lists the same indicators as 0-550 gpm (0-266 kpph). The Instrument Index shows a range of 0-266 kpph.

3. FSAR Table 7.5.1-5 identifies the control room filter temperature indicators as TI-7824 A1/Bl. The Instrument Index identifies them as TI-7824 A/B.

4. FSAR Table 7.5.1-5 identifies that control room emergency filter flow indicators FI-7817 A/B have 0-1200 cfm range. The Instrument Index shows 0-500 cfm.

5. FSAR Table 7.5.1-11 lists turbine first stage pressure indicators PI-446 and PI-447 with a 0-700 psig range. The Instrument Index shows 0-785 psig range.

6. FSAR Table 7.5.1-13 lists two component cooling water heat exchanger temperature indicators TI-674, TI-674.2, TI-675-1, and TI-675.2 providing the same identification.

7. FSAR Table 7.4.1-1 indicated that residual heat exchanger outlet flow indicators FI-0688 B, A2, Al have a 1500-7000 gpm range. The instrument index shows a 0-7000 gpm range.

8. FSAR Table 7.4.1-1 has reactor make-up flow totalizer FIS-0114 with a 0-160 gallon range. The totalizer has a range of 0-999, 999 gallons as shown on the Instrument Index. The inputs to the totalizer are from 0-160 gpm source.

9. FSAR Table 7.1.0-1 does not show instrumentation commitment to Regulatory Gudies 1.21,1.38,1.60, or 1.97; however, these are included in many instrumentation specifications. FSAR Section 7.1.2.15 provides information for Regulatory Guide 1.100 but it is not included in Table 7.1.0-1 either. This indicates to the IDI team that the FSAR is not used as the design input source of regulatory commitments to purchase specifications.

                                                             ?
                 ,.~                                                      'i
               ,                                        yf                        e D6.1-2 (DEFICIENCY) FSAR/ INSTRUMENT INDEX CONSISTENCY (Cont'd)                      3               ,

The FSAR does not. reflect the actual design for instrumentation ranges. These' cases appear to indicate that the FSAR is used to record information after ~ design is complete rather than as a design input document. These items appear to be systematic due to'the quantity of inconsistencies identified. ,; , RESPONSE-t,

• The Instrument List is an integral part of the design process, as such it sometimes leads the FSAR with respect to detailed system design. System /

                                                                                                  ^

changes occuq as part of the normal process of finalizing design and may ' require modification of the FSAR. When necessary, the areas of the FSAR that need to be revised are noted and scheduled for revision. It should be noted that although Ebasco prchedures permit the FSAR to be used as a design input, calculations or other design documents are available to detai1 ~, supplemerit or ccatirm FSAR information. Shearon Harris Nuclear Project procedures do not permit the FSAR to be used as,a design inpet except for commitments only. We have reviewed the identified discrepancies and inconsistencies with the following response. f' ' ' g

                                                                 .'     a

1. Inconsistency does exist, FSAR Table 7.5.1 will be updated to show correct range of 0-400 degrees F for TI-75418SB. <

2. Inconsistencies b'etween Tables 7.4.1 and 7.5.1 were corrected in FSAR '

   >>     Amendment 11. The instrument list showing only 0-266 kpph is valid since it is equivalent to 0-550 gpm for auxiliary feedwater flow, i

3. Inconsistency does exist, FSAR Table 7.5.1 will be updated to show j correct instrument tag numbers TI-7824 A/8.

s y ! 4. -Inconsistency does exist and the FSAR tehlebill be updated. However,

                                     ~

this task cannot be done until finalization of the scheduled HVAC static-pressure calculations which could affett the flow range shown both in , the FSAR and instrument list.

5. Inconsistency does exist, FSAR Table 7.5.1-will be updated t'o show the correct range of 0-785 psig for turbine first stage pressure.

i e

6. Inconsistency does exist because the component heat exchange temperature indicators are located on both the main control and auxilipry control boards. Although the instrument loop number is consisteqt(the use of the suffixes .1 & .2 are used to denote MCB/ACP locations when

        -indications are duplicated. The FSAR will be updated to reflect this.

7.- 'Although an inconsistency does exist presently, Ebasco has issued a design change notice, DCN-251-451 that. listed FSAR-Table 7.4.1 as impacted by the change of range of FI-0688 B, A2, Al to 0-7000 gpm.

        . Therefore, the FSAR change had been identified and had been scheduled.

Jj 'l , j f, .. / s 1

D6.1-2 (DEFICIENCY) FSAR/ INSTRUMENT INDEX CONSISTENCY (Cont'd) RESPONSE (Cont'd)

8. We do not consider this item to be a discrepancy or inconsistency due to the fact the FIS is a batch totalizer that translates total gallons measured (0-999, 999) based on a flow rate (0-160 gpm) input. Since the associated multi-pen recorder is calibrated for a range of 0-160 gpm, the FSAR Table 7.4.1 will be clarified to reflect total gallons measured.

9. We take exception to this item based on the purpose of Table 7.1.0. FSAR Table 7.1.0 lists any regulatory guides that were specifically mentioned in FSAR Chapter 7. In FSAR section 1.8 the commitments to the regulatory guides applicable to the Shearon Harris Nuclear Plant are delineated. This minimizes the need to repeat commitments that are generic to the plant (Reg. Guides 1.21, 1.38, 1.60). Reg. Guide 1.97, Post Accident Monitoring, has a specific position as detailed in the TMI Appendix to the FSAR, thus negating the need in Chapter 7.

Ebasco procedure (E-65SH) provides for a detailed listing of applicable standards and codes which are derived from the PSAR and FSAR for the Shearon Harris Nuclear Project. The purchase specifications as a minimum meet the FSAR requirements. Detailed design must occur prior to listing specific values in the FSAR. Therefore, the FSAR necessarily lags design and should not be utilized for detailed design purposes. In accordance with Ebasco procedures, all FSAR amendments are reviewed for consistency and accuracy. Additionally, Project Procedure E-69SH requires identification of FSAR impacts for design changes and the noted inconsistencies identified and scheduled for change. Therefore, this deficiency is not considered to be systematic. i

                                                   - - , _ .     -_.       v -

06.1-3 (DEFICIENCY) DIESEL GENERATOR STARTING AIR COMPRESSOR PRESSURE ALARM DESCRIPTION A low pressure alarm is provided for each diesel generator starting air compressor. The alarm setpoint was established by the diesel generator supplier, and was listed in a setpoint document. A low pressure alarm value of 200 psig was established for the 1A, 2A,18 and 2B starting air compressors associated with the two diesel generators. The setpoint document listed this value for three compressors, however, a value of 225 psig was shown for compressor 2A.

RESPONSE

The cause of the deficiency was an error in transferring of the correct value (200 psig) for the alarm setting for Diesel Generator Starting Air Compressor 2A from the Setpoint Worksheets to the computerized listing, 2166-8-508, via a data input sheet. This was an inadvertent oversight in checking the computer output against the input sheet and worksheets. This condition occurred on one (1) of four (4) setpoints for the DG Air Starting System (Trains A&B) that are to be identical. Therefore, it is considered to be random deficiency. The value had been corrected during the inspection immediately after its detection (12/21/84). A Design Change Notice, DCN-251-510, has been issued to formally correct the value shown in the Setpoint Document. This DCN has also been incorporated into Revision 22 of the Setpoint Document.

06.1-4 (DEFICIENCY) COMPONENT IDENTIFICATION ON CONTROL WIRING AND PROCESS CONTROL BLOCKS DIAGRAMS DESCRIPTION A control wiring diagram does not attempt to detail all of the components provided in a Westinghouse supplied instrument loop; rather, the control wiring diagram describes input and output termination connections for process instrument cabinets. Specific details of an instrument loop within the cabinet are described on other Westinghouse drawings, such as a process control block diagram. Minor differences in instrument tag numbering between nuclear steam supply and balance of plant drawings were agreed upon in 1978. The IDI team examined drawings for the boric acid tank level instrument loop for consistency in this Westinghouse to Ebasco interface. The Instrument List described the combination of LY/106A and LY/106B as

LS-1CS-0106A-SA-W. This latter component numbering choice was inconsistent l with both the control wiring diagram and the Westinghouse process control l block diagram. These inconsistencies in component numbering should either j have been identified and corrected or design guideline modifications should have been made. The IDI team noted that the boric acid tank level

 . instrumentation control wiring diagram was revised in both January and June 1984, and contained these inconsistencies at the time of its turnover to

Carolina Power & Light.

1 l Component numbering inconsistencies for instrument loop outputs to the annunciator have not been resolved through the design process to conform with agreed upon designations. Such inconsistencies introduce unnecessary confusion for document users since they must exercise judgement with correlating a Westinghouse drawing with either an Ebasco drawing or instrument Ifst.

RESPONSE

To identify instrumentation and control equipment for NSSS loops, the Shearon Harris Nuclear Project utilizes Westinghouse instrument tag designations. For , Balance of Plant (BOP) or Ebasco supplied equipment, functional tags are l applied rather than discrete output tags. This convention has been followed consistently in the drawing interface between CWD's and Westinghouse's PIC drawings. However, because the process loop for the Boric Acid Tank level was mistakenly attributed to Ebasco scope of supply, this convention was misapplied. Although the tank and temperature loops were supplied by Ebasco, the level indicator was supplied by ! Westinghouse. Since the actual cable and connector / terminations are correct, L this inconsistency does not affect physical design. Additionally, this tag l identification deviation would have been routinely identified during CP&L's review of the cable termination information as part of the cable termination card preparation. This is considered as an isolated error, with no effect on system design. Nonetheless, a design change notice, DCN-251-539 has been issued to correct the noted inconsistency between the CWD and Instrument List. l l \

06.1-5 (0BSERVATION) VOLUME CONTROL TANK ISOLATION VALVE INTERLOCK DESCRIPTION Two Volume Control Tank level instruments are listed as the same instrument on Westinghouse interlock CVC-8 (type).

RESPONSE

Level instruments LT-112 and LT-115 attached to the Volume Control Tank automatically realign the suction of the centrifugal charging pumps from the Volume Control Tank to the Refueling Water Storage Tank when low-low water level in the Volume Control Tank is sensed by each of the level transmitters. Westinghouse interlock diagram CVC-8 depicts the logic for closing the VCT outlet valves Lev-1150/E. Level instrument L-ll2 on interlock CVC-8 is incorrectly labeled as L-115 so that L-115 appears twice instead of just once. The mis-labeling is a typographical error as noted by handmarked correction on the master copy of the interlock sheets. The process control block diagram 1080803, Sheet 23, Sub 8, depicting the instrument loops for LT-112/115 are correct as are the elementary wiring diagrams for the actual valve control circuits (271C518, Sheet 174). Interlock sheet CVC-8 was corrected and re-issued to the appropriate Westinghouse functional groups. 1 i i I

06.1-ti (OBSERVATION) CYCS DESIGN BASIS DESCRIPTION The IDI team reviewed the " Design Basis Document," (DBD) prepared for the CVCS system and noted a discrepancy between what was stated in the DBD and the actual design classification of a level transmitter (LT-380) for the chiller surge tank. The DBD stated that "all CVCS instrument lines ... are designed to maintain boundary integrity following a seismic event;" the FSAR Table 3.2.1-1 identifies- the chiller surge tank as non-nuclear safety equipment but the same table identifies the instrumentation for that system as IE; the I&C Instrument Index (2166-8432, Rev. 33) for level transmitter LT-380 indicates that this instrument is classified as N3 (non-seismic).

RESPONSE

The CVCS is divided into three systems. One is the charging and letdown portion of the system described in DBD No.107. The Boron Thermal l' Regenerative Portion of the CVCS is described in DBD No.108. The portion of the CVCS which processes and recycles borated radioactive effluent from the reactor coolant system is described in DBD No.109. Section 1.0 of DBD NO. 107 clearly describes this system breakup. We disagree with the conclusions reached by the IDI team. Our review of the applicable documents, DBD, FSAR and Instrument Index indicates that the documents are consistent and do reflect the correct classifications of components. In particular the chiller surge tank and level transmitter (LT-380) are part of the Baron Thermal Regeneration System (DBD #108). The FSAR Table 3.2.1-1 correctly lists the chiller surge tank as NNS and the ISC Instrument Index (2166-B432, Rev 33) correctly lists the level transmitter (LT-380) as N3 (non-seismic). 3

l l D6.1-7 (DEFICIENCY) PROCESS INSTRUMENTATION CABINET INTERCONNECTIONS DESCRIPTION l Electric cables connecting balance of plant instruments to process l instrumentation cabinets are depicted on Ebasco control wiring diagrams and i are also described in the Ebasco instrument list data base. When process instrumentation cabinet C19 was added for a fire protection design change, control wiring diagram connections to this process instrumentation cabinet (Ref 1) were shown for three devices:

1) Charging header flow transmitter FT-1CS-0122;

2) Charging header auxiliary control board flow indicator FI-1CS-0122A.2, I and

3) Charging header flow control valve current to pneumatic converter I/P-1CS-0122. However, the instrument list continued to show cable connections for these devices to process instrumentation cabinet C6 rather than C19.

The Instrument List was inconsistent with the control wiring diagram for three cable connections.

RESPONSE

Although this control loop is classified as non-safety, it has been identified as required to attain a cold shutdown condition under an Appendix R scenario for evacuation of the Main Control Room. This resulted in a modification of this control loop, which was originally designed by Westinghouse. The modification was designed to minimize the impact to installed equipment (i.e., PIC cabinet 6 and the Main Control Board). The Instrument List at the time of the inspection reflected the original design since the overall Appendix "R" . modification relative to the control loops was still in progress and not finalized. Under this modification, portions of this loop were designed into a new process control cabinet, (i.e., PIC-C19). Since the changes were being implemented by DCN and the final design not complete, the update of the Instrument List was to be done after the Appendix "R" changes were ccmpleted. The three cable connections referenced on CWD sheet 310 are correct, although preliminary. This allowed field routing and termination in an expeditious manner. Since it is in the work plan to update the Instrument List for the Appendix "R" modification, no further action is required other than the originally identified and monitored under the existing change control procedure followed by pro, ject personnel.

06.1-8 (OBSERVATION) FLOW INDICATING SWITCH DESIGN PRESSURE AND TEMPERATURE DESCRIPTION In reviewing the. file containing the master pages for specification IN-11 the , IDI team noted inconsistencies in both design pressure and operating

  -temperature values for condensate pump seal water flow indicating switches FIS-1CE1900 A and B between the Addenda I and II sections of this specification. Design pressure values of 300 psig and 150 psig and maximum operating temperture values of 150 and 160 degrees F were stated for these instruments. During discussions with Ebasco personnel, it was determined that these instruments had been superseded by FS-lCE2340A and B which had been procured to values of 300 psig and 160 degrees. In this instance, the master file for the IN-ll specification contained obsolete technical data pages that were no longer applicable to the Shearon Harris Nuclear Project but had not been marked " void" to preclude inadvertent use.

RESPONSE

All specifications issued on the Shearon Harris Nuclear Project are transmitted to their respective vendors via purchase order contract supplement. This procedure provides a controlled method of distributing current information externally and internally to the Shearon Harris Nuclear Project. During the time of the inspection, the purchase order files containing the supplements were being prepared for transfer to the job site purchasing department. In order to facilitate the IDI the original of the specification was provided. As a matter of keeping a historical record of past information changes, the original of the previous addendum was maintained in the same " workday. file" which is not the formal project file. The IDI team noted that the operating pressure and temperature for the l Condensate Pump seal water line listed in spe:ification CAR-SH-IN-11 was not c consistent with the Line List, 2165-B-070. .Upon further investigation it was l determined that, inadvertently, the IDI team had reviewed a superseded i specification addendum that had subsequently been revised to reflect actual operating pressure and temperature requirements. This specification revision had b,een completed prior to start of the IDI. Since the addendun is generated from the same data base as the instrument list data base, consistent design information is maintained. Therefore, obsolete information cannot be used because of the controlled purchase order contract

distribution and instrument list consistency, 1

l l

D6.1-9 (DEFICIENCY) EBASCO PROCUREMENT SPECIFICATION DESCRIPTION Ebasco Procedure E-65-SH, defines the Standards and Regulatory Guides to be referenced in procurement documents. The specifications cited did not contain those standards and regulatory guides actually referenced . The IDI team concludes that "No technical impact on purchased instrumentation is anticipated" and further that "No hardware or analysis impact is anticipated since the procurement specifications provided specific data for environmental and seismic qualification where required".

RESPONSE

The nature of this deficiency appears to be that E-65-SH, in the opinion of the IDI team, does not reference a sufficient number of standards and regulatory guides. Actually, some of the Ebasco procurement documents reference too many. The tabulations provided by the IDI team as examples of the inconsistencies point out the IDI team's interpretation of the guides and standards and Ebasco's conservative over-referencing. Although procedure E-65-SH does not list the dates of issue for the various I&C Guides and Standards, those dates are defined for the project in the FSAR, or guidance is given to use the issue in effect at the time the purchase order is placed. IEEE Standards 279, 308, 379, 384 and Regulatory Guidos 1.29, 1.75 define and provide design guidance for those systems and components which should be seismically designed and qualified, and those systems and components that should be designed to withstand single failures. These design responsibilities lie with Ebasco and cannot be imposed upon an instrument supplier, because an instrument supplier has no way of assuring overall system design. Therefore, referencing these standards and guides in a purhcase docunent is academic. IEEE Standards 383, 332, 344 and Regulatory Guides 1.53,1.89 and 1.100 provide guidance and requirements for seismic and environmental testing. Scismic and environmental qualification requirements are specifically stated , in addenda to purchase documents where required. These addenda impose Shearon Harris Nuclear Project specific requirements which satisfy the FSAR commitments. The tabulaton provided as examples of inconsistency or omission merely highlight that some Ebasco purchase documents reference more than they should. However, the IDI team suggests that " systems level" references should be invoked for devices (instrument stands, level switches, thermocouples, pressure transmitters) where the vendor cannot have system responsibility. 1 l

D6.1-9 (DEFICIENCY)'EBASCO PROCUREMENT SPECIFICATION (Cont'd) RESPONSE (Cont'd) The input to E-65-SH has been revised to reflect the codes , standards and reg. guides listed as requirements in the specifications generated by the I&C _ group. Since the reviewer has stated the " procurement specifications provided

   . specific data for environmental and seismic qualification where required" no deficiency exists.

D6.1-10 (DEFICIENCY) INCOMPLETE AND UNISSUED DRAFTING MANUAL DESCRIPTION Site design work has been ongoing in the Electrical and Instrumentation and Control disciplines for approximately six months without having a drafting manual issued.

RESPONSE

The drafting manual being used at the time of the audit was an uncontrolled copy of Ebasco's drafting manual. We have written a drafting guide to provide guidance for the instrumentation and control discipline. It was reviewed by the principal engineers in each of the electrical and I&C disciplines and has been formally issued. CP&L's Quality Assurance group conducted an audit in March 1985 of the drawings issued without the use of the formal drafting guide. All of the drawing discrepancies noted were minor and no design changes were deemed necessa ry. Consequently, there is no hardware nor analysis impact. All drafting errors indicated by the QA audit have been identified for incorporation into later drawing revisions. Also as a further check, the ongoing FSAR consistency review (of change documents and drawing revisions versus the FSAR) will likely detect any discrepancies up to the present time and provide further assurance that drawing revisions have been properly accomplished, j v 4

 -y-                  -       ,.r     y-   e,--- - -    -v y--,, ,,q e- - - - - -   -
                                                                                      -w- .- 4 w.--. e ----w y   e-r-re.+4g

r I 06.3-1 (OBSERVATION) " TRAIN" TERMIN0 LOGY l DESCRIPTION The use of the phrase " safety train A or B" is used in the Shearon Harris Nuclear Project design to denote a redundant safety system. The IDI team discussion with Ebasco personnel indicated that the terminology " train A" is commonly meant as a redundant safety system. Inconsistent use of safety equipment technology might result in confusion in design information between Ebasco and Carolina Power & Light. The IDI team identified no evidence of any failure to use the correct power supply or safety related wiring as a result of the confusing terminology. Therefore, the IDI team recommends a consistent approach to designating safety / non-safety items.

RESPONSE

The design criteria stated in drawing CAR-2166-B-060 relative to safety and non-safety train designations was viewed by the IDI team to be potentially misleading. Since the design criteria in B-060 is specific with regard to the installation of safety and non-safety equipment and no violations of this design criteria were identified, the Shearon Harris Nuclear Project judges the existing notes to be satisfactory. Additionally, the terminology "SA" or "SB" is synonomous with safety train A or safety train B. Therefore, there is no inconsistency between AEP-1, specification IN-39, and associated sketch SK-350. The use of the term " train A" does not necessarily mean safety train related. It is not used to mean anything specific other than there is another redundant (safety or non-safety) train, process or control. l I L l l

k D6.3-2 (DEFICIENCY) CONDUIT SEPARATION DESCRIPTION Physical independence of redundant safety systems in instrumentation and control panels is maintained by barriers such as flexible conduit and/or air space between'the wiring. Separation guidelines are provided by NRC Regulatory Guide 1.75 and IEEE standards 384 and 420.

.The Shearon Harris Nuclear Project maintains separation in auxiliary control panels by means of flexible metallic conduit Lin addition to other methods.

! The flexible conduits of redundant safety channels are allowed to touch and in many cases are tie-wrapped together. The Shearon Harris Nuclear Project committed to follow Regulatory Guide 1.75, Revision 1, IEEE 384-1974, and IEEE 420-1978 in FSAR Section 8.3.1.2. 1.- Analysis of specific installation, such as the auxiliary control panels, is allowed by IEEE-Standard 384 (Section 5.6.2) and Regulatory Guide 1.75 to permit relaxation of separation criteria. Analyses of these types were not perfonned for Shearon Harris instrumentation and control panel internal wiring. The Shearon Harris Nuclear Project interpretation of separation is based 2. on IEEE Standard 420, Section 4.2.1 (1), which states that metallic conduit is an acceptable barrier and IEEE Standard 384, Section 5.6, which states that separation may be maintained by barriers. The more 4 conservative IDI team interpretation is based on IEEE Standard 384 sections 5.6.6 and 5.1.3, which state that redundant cable inside control ,

boards may use conduits for separation if a one-inch air gap is '

i. maintained.

4

3. The Ebasco electrical notes and details do not specify balance of plant internal panel separation methods.  ;

RESPONSE

The Shearon Harris Nuclear Project's position is con:.istent with commitments >

stated in the project's FSAR for Pegulatory Guide 1.75, Rev 1, which endorses , i IEEE 279-1971 and IEEE 384-1974. Paragraph 5.6.2 of IEEE 384 specifically states that for wiring inside of switchboards (i.e., control panels) separation must be maintained by: (1) use of flame retardant cables with analysis or (2) use of a 6" minimum spacial separation or (3) barriers between - redundant Class lE wiring.. The Shearon Harris Nuclear Project's commitments for Regulatory Guide 1.7.5 (FSAR Section 8.3.1.2.14), IEEE 279 (FSAR Section i

8. 3.1. 2. 22 ) , ' and IEEE 384 (FSAR Section 8.3.1.2.300) are consistent with the above criteria. Drawing 2166-B-060, sheet 7CA, Note 15, requires equipment manufacturers to provide separation by means of barriers, wireways and/or

conduit. This is in accordance with option 3 defined above.

The Auxiliary Control Panel (ACP) and Auxiliary Equipment-Panel-1 (AEP-1) { . designs were patterned after the basic design of the Main Control Board (MCB) [ supplied by Westinghouse, the NSSS supplier, which. utilizes flexible conduit to maintain separation. A numeric analysis is not deemed necessary because the following design considerations eliminate the possibility of thermal damage to wires in adjacent conduit:

e D6.3-2 (DEFICIENCY) CONDUIT SEPARATION (Cont'd) RESPONSE (Cont'd)

1. The wire size was determined for the particular load in accordance with the National Electric Code. This limits thermal build-up in normally operating equipment.

2. All control circuits are fused, effectively limiting the amount of overcurrent that the circuit would be forced to carry, thereby preventing thermal damage during fault conditions.

3. Separation by voltage level limits possible fault voltage to 140VDC or 118VAC, this eliminates the possibility of damage to adjacent conduit due to arcing.

4. The control wiring utilizes a rugged thermoplastic (Tefzel) insulation fabricated under military specification MIL-W-168780 code, rated 600V and 150C.

The combination of the wire sizing, as determined by approved codes, fuse protection and insulation type precludes any possibility of a fire being generated or propagated between trains or between train and non-train conduits (i.e., safety and non-safety). Based on the above, it is our position that the ACP and AEP-1 meet the intent of Regulatory Guide 1.75 and IEEE 279. Nevertheless, an engineering analysis has been performed to provide additional assurance that flexible conduit is adequate to maintain separation.

U6.3-3 (UNRESOLVED ITEM) INSTRUMENT IMPULSE LINE SEPARATION DISTANCE i DESCRIPTION

   ~

A design change (FCR-1029 Rev.1) which reduced the minimum separation , distance:between safety and non-safety tubing was designated " Minor Change" and did not undergo the design' verification process of ANSI N45.2.11. Therefore, the required assessment of the potential for adverse safety impacts was not performed. Additionally, detailed guidance on what constitutes a . " Minor Change" .is not provided in CP&L procedures and " Major Change" is not defined at all.

RESPONSE

The Shearon Harris Nuclear Project considers instrument tubing as defined by Specification M-71 (Rev.1) to be subject to the basic design requirements of

,     piping. As such, several provisions of FSAR Section 3.6. titled " Protection Against Dynamic Effects Associated with the Postulated Rupture.of Piping" are
'~

useful in addressing instrument tube rupture. In particular, Paragraph 3.6.1.2.4b states "where it is not feasible or practical to isolate the

     . Seismic Category I piping, the adjacent non-seismic Category I piping was 4      seismically designed in accordance with C.2 and C.4 of Regulatory Guide i      1.29". - This prevents non-seismic piping / tubing from causing the failure of seismic piping / tubing.

With regard to tubing / piping ruptures, FSAR Safety Evaluation Section 3.6.1.3 specifically defines how this type of failure is addressed. Namely: l 3.6.1.3 e) Unrestrained whipping pipes are considered capable of:

1. Rupturing impacted pipes of smaller nominal pipe sizes and

2, developing through wall leakage cracks in equal or larger nominal pipe sizes with thinner wall thickness.

3. 6.1. 3 f) Jet Impingement forces from a given pipe of specified nominal pipe i size and wall thickness are considered cabable of:

! 1. rupturing targeted pipes of smaller nominal pipe size, and

2. developing through wall leakage cracks in pipe of. larger j nominal pipe size and thinner wall thickness.

In general, instrument tubing in any particular area is of the same outside dimension and wall thickness. Therefore, consideration of failures of safety. l instrument lines due to the rupture of equal or smaller size non-safety lines i is not required. Further, the nomally postulated types of breaks or leakage cracks in fluid systems are not considered for piping less than one inch nominal pipe size. This exception is specifically designed in FSAR Paragraph 3.6.2.1.5. General NRC acceptance of this FSAR Pipe Rupture criteria and compliance with GDC-4, BTP ASB 3-1 and BTP MEB 3-1 are documented in the SHNPP SER (NUREG-1038) section 3.6. L

U6.3-3 (UNRESOLVED ITEM) INSTRUMENT IMPULSE LINE SEPARATION DISTANCE (Cont'd) RESPONSE - (Cont'd) Presently, the only instrumentation tubing greater than one inch nominal pipe size. is utilized in very limited, low energy, Seismic Category 1 applications. As such, instrument tubing does not need to be reviewed for the above

   -postulated failures. This is further addressed in the Ebasco Instrumentation and Design Criteria, Revision 2, dated November 8,1984, where it states ia Paragraph 2.20A, that where a. single credible failure could impair or destroy                                                  i both redundant safety class instrument lines, one safety channel relocation or a suitable barrier would be necessary. Paragraph 2.20A also states that instrument lines are constructed from the same high quality ASME material and that the potential for failure of one line caused by an adjacent line, regardless of channel designation, is not considered.

Based upon the above design criteria, .it is clear that the concern of the design is a single, credible, external common mode failure which could impact both redundant safety trains. Tube to tube failure is not considered a credible event.

   . It should also be noted that this design criteria and philosophy has been under discussion between CP&L and Ebasco for several months. This has led to a methodical, carefully considered evaluation of criteria which bridges design and construction activities. During initial Ebasco design activities, a five foot separation requirement was imposed on the designers. . This resulted in a general Ebasco routed separation of safety channels from each other as well as from non-safety channels. This routing was shown diagramatically in plan view on Ebasco produced Instrument Location and Arrangement drawings. End points, key penetrations and general area routing are shown. Precise routing dimensions are not shown. The five foot separation criteria also initially appeared in the Ebasco drawing CAR-2166-B-431, Instrument Installation Details.                                                  :

This criteria was removed in 1983 because it was design criteria not construction criteria. Ebasco had informed CP&L that if the specified diagrammatic routing was followed no further separation criteria need be considered because it had already been designed in. The reference to five foot separation was removed from the installation detail drawing. When tolerance questions. arose from the craft and inspection personnel, a memorandum was sent from site engineering to site Quality Assurance, (HNPD-840384, June 22,1984). This memorandum reiterated that the design separation requirement had already been taken into account by Ebasco design and that the diagrammatic rcuting was all that need be considered. However, a measurable tolerance was requested so FCR-I-1029 Rev. O was approved in August 1984 based on a telephone resolution with Ebasco. This FCR allowed a + 18 inch tolerance from the scaled tube routing as shown on the location / -- arrangement drawing. Note 1.18 was then added by Ebasco to the installation notes (drawing B-431). The 18 inch tolerance is based on the initial Ebasco design requirement of 5 feet minus 2X18 inches, which equals a 2 foot minimum separation distance and which envelopes the 18 inch Westinghouse requirement for redundant impulse lines. The note, however, still contained reference to separation between safety and non-safety tubes. l

 ;  , . - . -            w,.... .- e--_,m,..-,, .,n,.,.,n,,,,_,- - _ , , ,,,--,-_n._,,,._,-.-_,..-,-,.-.-,,-,.,-.,-,.,__-,n- - - , ,

1 U6.3-3 (UNRESOLVED ITEM)-INSTRUMENT IMPULSE LINE SEPARATION DISTANCE (Cont'd RESPONSE - (Cont'd) When site inspection personnel inquired about the non-safety inspection requirements, FCR-I-1029 Rev. I was issued in November 1984 to eliminate the non-safety separation requirement. There was considerable discussion since it was initially a criteria consideration. Final resolution involved both CP&L and Ebasco supervising I&C Engineers. Telephone conversation memorandum of November 5,1984 confirmed, once again that the Ebasco design criteria concern was a single, external, common-mode failure which could damage both safety trains. . FCR-I-1029 Rev. I clearly concerned only one safety channel and its proximity to a non-safety channel. Note 1.18 of installation notes drawing B-431 has subsequently been revised by Ebasco on February 7,1985 to remove the safety to non-safety tube separation requirement. It should be clearly recognized that FCR-I-1029 Rev.1 addressed only tube to tube interactions resulting from a tube failure. Interdiscipline clearance criteria (based on thermal growth, accessibility, vibration, etc) are addressed in B-431 Note 1.01. Regulatory Guide 1.29 concerns are addressed in B-431 Note 2.09b and state that non-seismic installations shall not be allowed to impact sa fe ty-rel a ted/se i smic . i ns tal l a ti on s. This requirement is satisfied by the Ebasco general area routing, by field identification of potential problem areas, and by a sitewide Regulatory Guide 1.29 walkdown program which was already in progress. This program specifically evaluates systems interaction concerns. Neither of the above notes were affected by the sub;ect FCR. With . respect to the " Minor Change" designation, safety related #CRs are initially evaluated as to effect on the design. They may be resolved by Ebasco or Harris Plant Engineering Section (HPES) as appropriata. If the resolution is by Ebasco, HPES does not dictate whether design verification is -required. .If the resolution is by HPES, internal departmental procedures and section instructions apply. Both documents contain design verification instructions and the conditions under which they apply. The procedures do not require design verification for a " Minor Change". Minor changes are those items which do not involve changing the safety review, design calculations, or design basis. A major change is not defined since design verification is required for everything that is not a minor change. In summary, we do not agree that the subject change requires verification. The design basis requirement of separation between redundant safety-related instrument tubing was not changed. This fact was discussed at length on numerous occasions between CP&L and Ebasco. Conversations ind correspondence are documented, though not included as part of the original justification sheet. Since the subject FCR is in accordance with regulatory criteria and SHNPP licensing ccmmitments, and in no way reduces the safety of design, the designation " minor" does apply and further action is not required.

U6.3-3 (UNRESOLVED ITEM) INSTRUMENT IMPULSE LINE SEPARATION DISTANCE (Cont'd) RESPONSE - (Cont'd) In order to provide a concise historical perspective, a copy of this response and the above referenced telecon and site engineering memo has been filed with

   . the existing FCR justification sheet. Also, since the instrument location arrangement drawings have been developed depicting routing using the original separation criteria, this condition would only apply to FCR's involving tubing reroutes. Numerous FCR's were reviewed by Ebasco for any unacceptable consequences of the reroutes from the view-point of jet impingement, single failure criteria or separation and general interference checks. None have been found.

Finf ally, other " Minor Change" FCRs which the IDI team may have considered as being inappropriately designated are discussed in Chapter 7.4 of this response.

D6.4-1 (0BSERVATION) ITT BARTON DIFFERENTIAL PRESSURE SWITCHES DESCRIPTION The Ebasco equipment qualification group prepares packages which should include all of the environmental qualification testing and reviews. They are required to interface with the other disciplines to ensure that the package is complete and accurate. Ebasco's preparation of the package No. 36.5, Differential Pressure Switches,

 - failed to reveal the change in equipment. There also appears to be a lack of communication between the instrumentation and control group and the equipment qualification group. The change in equipment had been identified and accepted by the instrumentation and control group and no impact on safety has resulted.

RESPONSE

The Equipment Qualification Package for differential pressure switches (No. 36.5) was not consistent with the applicable specification CAR-SH-IN-46. The accuracy of the pressure switches was originally specified as + 0.5%. The qualification test report verified the accuracy to be + 2% and + 5% for the specific models supplied. Revision 0 of the specification was iised for bidding purposes. At that time the vendor had taken several technical exceptions, one of which was the required accuracy. Long delivery schedules necessitated placement of the contract with the understanding that the specification would be revised to reflect actual accuracy values. This was done af ter the qualified accuracy values were determined to be acceptable and compatible with the system application. This process was initiated prior to the start of the IDI. Since the vendor had required acceptance of all qualification prior to the contract release, Ebasco perfern'ed a review and determined that the reports were fundamentally acceptable. Once this had been done, associated EQ , packages were prepared by I&C engineers assigned to the EQ group. Since these engine.frs were in verbal contact with the engineers responsible for the contract, the accuracy deviation was known. Although the specification i revision was "in-progress" the E0 package had to accurately reflect the qualification reports with any deviations so noted, i The specification CAR-SH-IN-46, Revision 1, has been approved and transmitted to the vendor via a purchase contract supplement. Since there are no outstanding items the EQ package will not have to be nodified. 4

D6.4-2 (DEFICIENCY) VENDOR CONFORMANCE TO SPECIFICATION l DESCRIPTION  ; l The containment hydrogen analyzer system that was supplied for use at the Shearon Harris Nuclear Plant does not comply with the design requirements of Specification CAR-SH-N-45 because it was not designed as Class 1, Division 2, Group 8. Additionally, the vendor did not take exception to the specification, nor did it provide analysis to support the deviation from the specified design. Ebasco did not comment on the deviation as part of its review and acceptance of vendor-supplied infomation as required by Company Procedure E-6

RESPONSE

Paragraphs 6.15 and 7.01.d.7 of specification CAR-SH-N45 classify electrical components which can come in contact with a flammable concentration of hydrogen as Class I, Division 2, Group B. This requirement applies to the Hydrogen Analyzer Cabinet (Par 6.15) and the Remote Sample Dilution Panel (RSDP) (Par 7.01.d.7). The equipment vendor, did not provide an explosion proof design but has instead provided analyses to demonstrate that flammable concentrations of hydrogen cannot occur. The RSDP is equipped with an exhaust blower designed to maintain a minimum of one (1) inch water negative pressure within the panel at a minimum of 75 CFM airflow. This design provides a forced air circulation which prevents the buildup of a flammable concentration inside the panel during a worst case gas leakage condition. The Hydrogen Analyzer is designed for a maximum process flow rate of 1 1/2 ft 3/ minute (which is less than one liter per second). Vendor calculations demonstrate that for a leakage of up to one liter per second, the hydrogen concentration build-up is less than 4% which is below the flammable potential. The vendor calculation is based on a conservatively calculated natural ventilation rate. Although an explosion proof design has not been provided, the intent of the specification has been satisfied by ensuring that flammable hydrogen concentrations cannot occur. The applicable sections of the specification will be revised to reflect this. 1 l

U6.5-1 (UNRESOLVED ITEM) DESIGN BASIS FOR SAFETY RELATED INSTRUMENT SETPOINTS DESCRIPTION To assess the technical adequacy of individual setpoint values, the IDI team selected a sample of 17 balance of plant safety related instruments in five systems, and examined the design basis documentation for their safety related setpoints. The Shearon Harris Nuclear Project's position is that only those setpoints identified in the Harris Technical Specifications were subject to this design basis documentation requirement. FSAR Section 1.8, which commits to Regulatory Guide 1.105, only specifically refers to NSSS and balance of plant setpoints which are to be contained in the plant Technical Specifications. The instrument sample selected by the IDI team provided automatic protective or control actions within safety related systems or provided Class 1E alarm indications to support subsequent operator actions. The IDI team identified:

1. one setpoint that did not appear to be technically adequate,

2. one setpoint that did not use the current FSAR design input,

3. a number of inconsistencies among EBASCO design documents,

4. extensive use of undocumented engineering judgement by EBASCO to establish setpoint values and tolerances, and

5. setpoint documentation deficiencies in four of the five systems inspected.

A number of safety related instruments in four of five balance of plant systems reviewed by the team did not have a documented design basis for setpoint and setpoint tolerance values as required by IEEE Std. 279-1971, commitments to NRC Regulatory Guide 1.105 and Ebasco procedure E-77. Other omissions and inconsistencies among the design documents regarding setpoints and tolerances violate Ebasco precedure E-77. Inadequate design basis documentation appears to be systematic in HVAC, emergency service water, and the reactor auxiliary butiding sump level safety related setpoints based on the sample examined by the team. Use of an obsolete FSAR design input for the spent fuel temperature setpoint appears to be a random error. l l 1

U6.5-1 (UNRESOLVED ITEM) DESIGN BASIS FOR SAFETY RELATED INSTRUMENT SETPOINTS (Cont'd)

RESPONSE

The scope of the Setpoint Document is consistent with Regulatory Guide 1.105, Instrument Setpoints. The Regulatory Guide defines specific guidance or explicit methodology concerning parameter (e.g., instrument errors) to be addressed in formulating instrument setpoints used for automatic initiation of protective actions and alarms.

 -The Shearon Harris Nuclear Project utilizes this explicit setpoint methodology only for Plant Protection System input setpoints (ESFAS and RPS). The Westinghouse Precautions, Limitation, and Setpoints (PLS) Document provides the values for the great majority of these items. Westinghouse has also provided in the setpoint methodology document all the calculated values that were used for determining the ESFAS/RPS setpoints. Ebasco's responsibility is limited to the explicit setpoints for Containment Radiation and Chlorine Level for the Containment Ventilation and Control Room Ventilation Isolation signals respectively. The Shearon Harris Nuclear Project considers this to be consistent with position Cl of Regulatory Guide 1.105.

All other setpoints (non-ESFAS/RPS) are considered implicit and as such do not require the same degree of documentation of calculation basis. Therefore, the values listed in the Setpoint Documents which reflect fluid processes have sufficient documentation. It is the Shearon Harris Nuclear Project's position that these values will be maintained during plant operations or modified to account for the physical and/or operating environment. Therefore, the Setpoint Document has been prepared in accordance with applicable criteria and its accuracy is not in question. Each of the areas noted by the IDI team will be reviewed and, where necessary, additional documentation to support the issued setpoints 9b-508) will be provided. In the event there are changes to existing setpoints, a design change notice will be initiated. The following addresses the areas noted by the IDI team:

1. The setpoint of the containmert dome temperature elements was initially selected using engineering judgement pending confimation of the setpoint by calculation. The current setpoint (1250F) of the containment dome temperature elements has been verified by calculation.

l

2. The HVAC ESF local coolers have been sized and specified in conformance with calculations based on design space temperatures and corresponding cooling loads. There are two (2) types of coolers, by function. Some

of the coolers operate during normal operation and following an accident. The remaining coolers operate only following an accident.

The interlock and annunciation temperature setpoints for the local coolers should conform to the space design temperature under both scenarios. It was determined that in some cases, the temperature setpoints of the local coolers did not conform to the requirement. I

U6.5-1 (UNRESOLVED ITEM) DESIGN BASIS FOR SAFETY RELATED INSTRUMENT SETPOINTS (Cont'd) RESPONSE - (Cont'd The HVAC area interlock and annunciation temperature setpoints of the ESF local coolers have been. reviewed. Revised input will be provided to the I&C Department, where required, based on cooler function during normal and accident conditions and the corresponding space design

<                                        temperature..

All other area temperature setpoints will be reviewed'as part of a l _ program to evaluate and document HVAC system setpoints.

3. - An assumption made in establishing the Emergency Service Water header setpoint pressure was incorrect. Ebasco has revised the associated

calculation EQS-19, and will incorporate the revised values into the -

l setpoint document.

4. Setpoint data for the RAB sump pumps was determined from Sump Pump Specification CAR-SH-M-18, which is a non-nuclear specification. .I AC i selected setpoints for safety-related level switch based upon data in the specification.

5. An FSAR Amendment provided fuel pool equilibrium temperatures based on

, an analysis which had not been verified. " Calculation of Equilibrium i Temperatures" and " Calculation of Rates of Temperature Increase for North End and South End New and Spent Fuel Pools" has been completed, reviewed, approved and verified. The results indicated that the non-verified analysis referred to was somewhat conservative. An FSAR change has been . initiated to-reflect the results of the analysis.-- No i hardware changes are required. i HPES section Instruction 3.8 " Review and Approval of FSAR and ER Changes" will be revised to include a requirement that FSAR changes be supported by a design document, as appropriate. i i i-l I i i i }

D6.7-1 (DEFICIENCY) WESTINGHOUSE REACTOR COOLANT PUMP INSTRUMENTATION DESCRIPTION An inconsistency exists in Westinghouse instrument list specification sheets with regard to range of bypass flow elements. Also, a Westinghouse change to the specification was not incorporated into the Ebasco Instrument List.

RESPONSE

Westinghouse changed the bypass flow element's range from 0-728.6 inches DP to 0-60 inches DP, the range for the flow indicating switches however, was not revised. Westinghouse has now revised the specification sheet for the indicating switches so that they correspond to the flow elements specification. In addition, the Instrument List did not reflect the changes that Westinghouse had made. Ebasco was notified of these changes in letter CQL-8603. The Instrument Index will be updated so that it correlates with Westinghouse.

06.8-1 (OBSERVATION) DESIGN CHANGE NOTIFICATION SUPPORTING DATA DESCRIPTION The IDI team reviewed thirteen instrumentation and control design change notifications issued by.Ebasco during 1983 and 1984 for both technical content and conformance with the applicable procedure. In the IDI team's judgement, adequate detail regarding the approved technical change was provided in each of the reviewed design change notifications. The Ebasco procedure instructed the preparer to complete Form 612, ' Design Change Notification' (DCN), in detail, attach pertinent support data and required documentation to define the change, record of telephone calls, memoranda, notes, field change request, etc. The IDI team notes that this wording is subject to interpretation for both purpose and form. For example, it could be interpreted that the requirement is stated to direct the preparer to gather background information to ease the review process or that merely referencing background material (record . telephone calls, etc) was sufficient. The statement could also be interpreted that all pertinent background material should be retained with the official record copies of design change notices. In the instances noted by the IDI team neither the attachments required by the Ebasco procedure nor a direct reference to the pertinent background material were provided with the completed and approved design change notification.

RESPONSE

The Shearon Harris Project meets the stated procedural requirement by means of information either attached to or referenced on the DCN cover sheet and impact summary sheets. This information provides sufficient traceability to identify -the cause of the design change. Typically, items referenced which support this are the HPES authorization request, work task number and applicable telecons (under the comments section). Additionally, the project Drawing Closeout Log (DCL)-lists outstanding design changes to be incorporated on the drawing. As indicated by the IDI team, no action is required on the DCN's previously issued but all future DCN's will contain additional definition as to the cause of modifications. Although the design issued by these DCN's is adequate and the intent of the procedure is satisfied, the procedure E-69-SH and the associated DCN memo sheet (612 SH) have been revised to clarify the information required to be noted for a design change.

F 06.8-2 (OBSERVATION) SUMP PUMP CONTROL DESCRIPTION The IDI team reviewed the design criteria used by Ebasco to add the safety sump pump overflow instrumentation. During the review the team observed a non-safety design error associated with the sump pump start-stop level control system. The team noted that the pump vendor indicated on the sump pump drawings a maximum float travel of 21 inches for the level switch. The sump pump specification specified float settings in a range of float travel from 24 inches to 42 inches. This discrepancy was not discovered during the Ebasco vendor drawing review phase (circa 1979). In 1984 during the equipment setpoint effort this discrepancy was brought to the attention of the Ebasco Mechanical Engineering Group. Subsequently, Ebasco advised CP&L that all the sump pumps would be modified except for 5 items that either were.not required because of Unit 2 deletion or "they do not need to be revised because the 21 inch travel is not exceeded". The team discovered however that at least two of the five sump pumps (M-18 Items #10 & #15) which were exempt from the retrofit had a specified travel greater than the existing 21 inch vendor design. In addition the team noted that the Hi-Hf alarm (which is mounted independently) will not occur at the same time as the Hi-Hi pump control function (start second pump) if the equipment was mounted as -specified. The IDI team also expressed some concern regarding the transfer of information between intraorganization groups.

RESPONSE

Ebasco has reviewed the IDI team's findings in this area and acknowledges that the discrepancy between the manufacturers equipment data noted on the drawings and equipment specification should have been identified and resolved during the vendor drawing review phase. The discrepancy was discovered during the equipment setpoints effort and brought to the attention of Mechanical Engineering. Ebasco initiated the design modification with the manufacturer to correct this design error. While establishing the necessary information for the vendor to implement the design change, the travel range (21 inches) was incorrectly compared to the Hi level setting for two of the pumps in lieu of the Hi-Hf setting. In order to correct this error the Hi-Hi setting has been adjusted so that the level switch will operate within th( present range of the switch (21 inches) without impairing the system function. In addition the Hi-Hi level alarm (mounted independently) has been adjusted to correspond to the Hi-Hf pump control setting. This will alert the operator that the Hi-Hi water level has been reached. Regarding the concern expressed by the IDI team about the transfer of information between groups Ebasco believes that the existing design control process is adequate. The specific level switches identified by the team with this discrepancy are non-nuclear safety and their failure would not have jeopardized safe operatio1 of the Shearon Harris Nuclear Plant.

      ,  06.8-3 (OBSERVATION) LICENSING GROUP INTERFACE DESCRIPTION The Ebasco I&C design criteria state that I&C drawings related to nuclear safety shall carry the notation " NUCLEAR SAFETY RELATED" and be sent to the Ebasco licensing group for their safety review. In addition, the Ebasco interdisciplinary drawing review list states that licensing group will review all control wiring diagram changes. Ebasco stated that it is allowable to perform the required licensing review either at the time of Design Change Notice (DCN) issue or when the base drawing is issued incorporating the DCN.

As a result of the transfer of most of the drawings from Ebasco to Carolina Power & Light, the IDI team believes that a potential exists for safety DCN's written by Ebasco but incorporated by Carolina Power & Light to bypass licensing review. As a consequence, required FSAR changes may be missed.

RESPONSE

One element of Ebasco's interdisciplinary design review includes the review of safety-related design drawings by th? Licensing Department to ensure the incorporation of pertinent regulatory requirements / commitments and the accuracy of the FSAR versus project design documents. (See Engineering Procedures E-7, E-21, E-69 and Licensing Procedure L-7). The engineering procedures allow the lead discipline engineer (LDE) to determine if Licensing review is required for Design Change Notices (DCN's) based upon LDE assessment of the significance of the revisions. The transfer of drawings from Ebasco to CP&L has necessitated completion of the review cycle used by Ebasco Licensing to control the FSAR prior to transmittal to the Shear n Harris Site. Ebasco has revised Engineering Procedure E-69SH " Procedure for Design Change Notification / Field Change Request" to, require Licensing review of Design Change Notices with potential FSAR impact prior to final signout. This provides additional assurance that the FSAR will be maintained current with design. In order to assure further FSAR accuracy and completeness, the Shearon Harris Nuclear Project has initiated an FSAR Consistency Review. Under this program, a Licensing review of design change documents will be performed to ensure that these design changes are consistent with pertinent regulatory requirements, commitments and stated project positions as described in the SHNPP FSAR. The program also provides for FSAR updates should additional clarification be warranted. c -

06.8-4 (OBSERVATION) VOIDING FIELD CHANGE REQUESTS DESCRIPTION The .IDI team noted that a Field Change Request can be voided by its originator without obtaining the approval of the responsible supervisor.

RESPONSE

We have establis'hed the same level of approval for voiding FCR's as for an original issue. Procedure AP-lX-05 has been revised to require a discipline principal engineer's concurrence on all voided FCR's. The discipline principal engineers have been placed on standard distribution, by letter MS-854044, for all FCR/PW's to make sure they concur with all the voided FCR/PW's. No further action is required to correct the problem since no errors were identified before the. procedure for voiding was changed. I i i a nn-y---,,- ,,v., .. , w ~., ,,. ,,,..-nm.m-_--,

3 06.8-5 (DEFICIENCY) BATTERY ROOM SEl'VICE SINK The battery rooms include personnel safety and maintenance plumbing fixtures. The shower, eyewash and service sin't were inspected by the team, and it was noted that the support frame for the service sink is constructed of wood in lieu of metal as required by specification Be-10.

RESPONSE

The wooden supports have 'been replaced with metal. Material substitutions (both Safety and NNS) are-required by procedure to be documented by an approved change document. Our review of the FCR's written against Specification BE-10 indicates that this item represents an isolated case of material substitution without engineering review and approval. e l l .  ;> r y l l

                                   ?

4 l , 1 l 7 l \ h A

r-

                                                                                                                  -7 D6.8-6 (DEFICIENCY) CALCULATION BASIS FOR LICENSING AMENDMENT DESCRIPTION An FSAR amendment provided fuel pool equilibrium temperatures based on an analysis which had not been verified.

RESPONSE

Calculation of equilibrium temperatures and calculation of rates of temperature increase for north end and south end new and spent fuel pools has been completed, reviewed, approved and verified. The results indicated that the non-verified analysis referred to by the IDI team was somewhat conservative. An FSAR change has been initiated to reflect the rc;dits of the analysis. No hardware changes are required. HPES Section Instruction 3.8, " Review and Approval of FSAR and ER Changes", has been revised to include a requirement that FSAR changes be supported by a design document, as appropriate.

                         ,        . - , , - . , - - , - - - . . - n..- -. ~ ,- , - ---- ..c.n --,-., , - ,.-. ,e}}