ML20237A833

Safety Evaluation Accepting USI A-46 Program Implementation at Plant,Unit 1

ML20237A833
Person / Time
Site:	Three Mile Island
Issue date:	08/12/1998
From:	NRC (Affiliation Not Assigned)
To:
Shared Package
ML20237A827	List: ML20237A825 ML20237A833
References
REF-GTECI-A-46, REF-GTECI-SC, TASK-A-46, TASK-OR GL-87-02, GL-87-2, NUDOCS 9808170169
Download: ML20237A833 (28)

Text

,

g

>S urg\ UNITED STATES g j NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 30806-0001 o  !

{CT/ .. +

SAFETY EVALUATION OF USl A-46 PROGRAM IMPLEMENTATION AT THREE MILE ISLAND, UNIT 1 DOCKET NO. 50-289, OPERATING LICENSE NO. DPR-50

1.0 BACKGROUND

On February 19,1987, the NRC issued Generic Letter (GL) 87-02, " Verification of Seismic Adequacy of Mechanical and Electrical Equipment in Operating Reactors, Unresolved Safety issue (USI) A-46." GL 87-02 encouraged licensees to participate in a generic program to resolve the seismic verification issues associated with USl A-46. As a result, the Seismic Qualification Utility Group (SQUG) developed the " Generic implementation Procedure (GlP) for Seismic Verification of Nuclear Plant Equipment," Revision 2 (GlP-2, Reference 1).

On May 22,1992, the NRC issued Supplement 1 to GL 87-02 including the staffs Supplemental Safety Evaluation Report No. 2 (SSER-2, Reference 2), pursuant to the provisions of 10 CFR 50.54(f), which required that all addressees provide either (1) a commitment to use both the SQUG commitments and the implementation guidance described in GlP-2 as supplemented by tre staffs SSER-2, or (2) an attemative method for responding to GL 87-02. The supplement also required that those addressees committing to implement GIP-2 provide an implementation schedule as well as detailed information including the procedures and criteria used to generate l the in-structure response spectra (IRS) to be used for USI A-46.

By letter da'ed September 21,1992, (Reference 3) GPU Nuclear, Inc., (GPUN, the licensee) provided its response to Supplement 1 to GL 87-02 for Three Mile Island, Unit 1 (TMI-1). In that letter, GPUN committed to follow the SQUG commitments set forth in GlP-2, including the clarifications and exceptions identified in SSER 2. The staffs evaluation of the licensee's response was issued in a letter dated November 18,1992, (Reference 4).

i GPUN conducted the USl A-46 program and submitted a summary report on May 17,1995, (Reference 5). The staff reviewed the summary report and requested additionalinformation (RAI) on August 29,1995, (Reference 6) and February 24,1997, (Reference 7). GPUN responded to the staffa RAls on September 27,1995, (Reference 8) and May 28,1997, l

(Reference 9). After reviewing GPUN's responses, the staff noted that some areas in the RAls were not satisfactorily addressed. Therefore, the staff decided to conduct a site audit on July 21-l~

i 25,1997. During the audit, the staff obtained clarification to earlier GPUN responses and reviewed additional documentation including calculations and field walkdown sheets. The staff also performed a field inspection of a selected sample of equipment, cable raceway and conduit support systems and tanks.

l This report provides the staff evaluation of the licensee's USI A-46 implementation program based on the staffs review of the summary report, responses to RAls, and further clarification and documentation provided to the staff during and after the site audit. The site audit report is attached as Appendix A.

ENCLOSURE 9908170169 990812 PDR ADOCK 05000289 P PDR , ,

2 2.0 DISCUSSION AND EVALUATION The summary report (Reference 5) provides the licensee's implementation results of the USl A-46 program at TMI Unit 1. The report contains safe shutdown equipment identification, seismic screening verification and walkdown of mechanical and electrical equipment, the relay l

evaluation, seismic adequacy of tanks and heat exchangers, seismic adequacy of cable and conduit raceways, and outlier identification and resolutions. The report also provides the equipment modification schedule for resolving outliers.

2.1 Seismic Demand Determination (Ground Spectra and in structure Response Spectra)

GPUN utilized the licensing basis design response spectra (DRS), as defined in TMI-1 Final Safety Analysis Report (FSAR), as the seismic input at grade level to develop the in-structure response spectra (IRS) at floor elevations in the Reactor Building, intermediate Building (IB),

Auxiliary / Fuel / Control Building, and the Turbine Building. The Safe Shutdown Earthquake (SSE)

DRS horizontal components have a peak ground acceleration (PGA) of 0.12g. The corresponding vertical PGA is 0.08g. The method of developing IRS had been evaluated by the

staff and documented in Reference 10.

The staff found the licensee's approach for determining seismic demand to be consistent with the provisions described in Section 4.2.4 of GIP-2 for " conservative design spectra," and are therefore acceptable.

2.2 Seismic Evaluation Personnel The summary report states that the seismic walkdowns were managed by GPUN project management personnel and performed by two separate walkdown teams (Reference 5). Each l

team consisted of an EQE seismic capability engineer (SCE) and a GPUN SCE. The summary report further states that the SCEs met the qualification requirements of GIP 2. An organizational chart and several resum6s are also included in the summary report. During the site audit, the staff leamed that although the formal walkdown was performed by two teams, the

! individual members in a particular team did not necessarily remain with that team and sometimes l members switched from one team to another. At any particular time, any of the two teams consisted of at least two SCEs with at least one from EQE.

l The licensee reported during the audit that the seismic engineers worked with the systems and  !

operations engineers to obtain clarification on equipment functionality. Furthermore, after the systems engineers had identified the safe shutdown equipment items, the electncal and instrumentation engineers prepared a list of the associated relays. Through a review of the resum4s and interaction with GPU SCEs during the audit, the staff observed that the SCEs were .

l competent and their qualifications exceeded the minimum requirement of GIP-2. The staff also noted that the " Third Party" review was conducted by an independent expert highly recognized in l the field of seismic evaluation of structures, systems and components.

i

l I

)

3 2.3 Safe-Shutdown Path GL 87-02 specifies that the licensee should be able to bring the plant to, and maintain, a hot shutdown condition during the first 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> following a safe shutdown earthquake (SSE). To meet this provision, in its submittal of May 17,1995, (Reference 5) the licer,see addressed the following plant safety functions: reactivity control; pressure control; inventory control; and decay heat removal. A primary and attemale path were identified for each of these safety functions to ensure that the plant is capable of being brought to, and maintained in a hot shutdown condition for 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> following an SSE.

Short-term reactivity control is achieved through insertion of control rod assemblies while long-term reactivity control is achieved through injection of borated water from either the makeup tank or the borated water storage tank (BWST) to the reactor coolant system (RCS). Injection is achieved via the safety-related makeup pumps and associated flow paths. Reactor coolant inventory is maintained by injection of water from the makeup tank or BWST to the RCS via the safoty-related makeup pumps and associated flow paths. Reactor decay heat removalis accomplished by secondary heat removal. This operation consists of feeding water from the condensate storage tanks (CSTs) to the steam generators (SGs) via the safety-related einergency feedwater pumps and allowing steaming (boil off) of the SGs through the main steam atmospheric dump valves. Other sources of water for feeding the SGs include the condenser hot well, the demineralized water storage tank, and the reactor building river water.

The licensee has identified several diverse water sources to ensure that the plant safety functions are maintained. The seismically qualified sources used for direct injection into the RCS are the BWST and the makeup tank with capacities of 350,000 gallons and 2,200 gallons, respectively. The seismically qualified sources for makeup into the SGs are the two CSTs which have capacities of 150,000 gallons each and are projected to last for approximately 42 hours4.861111e-4 days <br />0.0117 hours <br />6.944444e-5 weeks <br />1.5981e-5 months <br />.

The other qualified water source is the reactor building river water which has an entire seismically-qualified flow path. Other non-seismically qualified sources which can be used for l makeup to the SGs are the condenser hot well and the demineralized water storage tank with l capacities of 764,000 gallons and 1,000,000 gallons, respectively. If available, these non-I seismically qualified sources would be used upon depletion of the CSTs and prior to injection of reactor building river water into the SGs.

i in the unlikely event that all of these cooling sources are unavailable, the operators would i

implement the primary feed and bleed mode of cooling which uses the BWST and the high head safety injections pumps. The licensee stated that components and equipment within this cooling path are seismically qualified in accordance with the original plant design and are determined to l be seismically adequate.

A composite list of the equipment necessary for the above-mentioned paths is included in Appendix A," Composita Safe Shutdown Equipment List (SSEL)," of the licensee's USl A-46 Seismic Evaluation Report (Reference 5). The components contained in this list were walked down by the licensee. In addition, the licensee performed reviews and simulator training I

l 1

9

i l

l 4-  !

exercises to verify the adequacy of the SSEL and plant procedures, and the compatibility of the SSEL with these procedures.

The numerous water sources, flow paths, and time available for achieving the operations alignments provide reasonable assurance that adequate decay heat removal capability is available. Symptoms-based abnormal and emergency operating procedures for lining up cooling sources are available-st the site.

In addition to the general criteria recommended in GlP-2 (Reference 1), certain plant-specific criteria and assumptions were used for selecting the safe-shutdown path and identifying equipment needed for safe shutdown. During the site audit, the licensee reported that the safe shutdown equipment list (SSEL) was prepared in several joint sessions and the formal "walkdown" was preceded by a preliminary " walk by." The " walk by" gave the engineers an opportunity to become familiar with the SSEL items, to obtain plant perspectives regarding interaction and structural concems, and to identify " branch lines" (i.e., auxiliary equipment items).

The licensee illustrated its equipment identification process by considering one safe shutdown path. The staff also loamed during the audit that a large number of equipment items that were in the safe shutdown paths were not included in the safe shutdown iist since they were judged to be passive or structurally " rugged." The licensee presented a list of the missing items (which were about as many as those in the SSEL). The staff reviewed the list and concurred with the licensee's judgment. In summary, the staff finds that the licensee has identified the safe shutdown path following the GIP-2 criteria and, in general, has considered the needed equipment items.

The staff concludes that the licensee's approach to achieve and maintain hot shutdown for 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> following a seismic event is acceptable.

2.4 Selsmic Screening Verification and Walkdown of Mechanical and Electrical Equipment The seismic screening and walkdown included verification of 658 equipment items in the 20 j classes of equipment included in GIP-2 (Reference 1) and the "other" equipment class not covered in GIP-2, 2.4.1 Equipment Seismic Capacity Compared to Seismic Demand The seismic capacity of equipment classes included in GlP-2 (Reference 1) was used by the licensee for determination of the seismic adequacy and compared with either the ground response spectra or the in-structure response spectra as recommended by GIP-2. The capacity )

spectra are provided in GIP-2 as a Bounding Spectrum, a Reference Spectrum (RS) (i.e.,

Bounding Spectrum multiplied by 1.5) or Generic Equipment Ruggedness Spectra (GERS).

Examples of how the capacity spectra were used to compare with the demand were reviewed at the site audit. In response to the staffs questions on the use of Method A from Table 4.1 of GIP-2 for comparing the seismic demand to seismic capacity of the equipment, the licensee stated

, (Reference 9, item 26) that in general, it has used Method A for equipment located within 40 feet above the effective grade level. However, recognizing the staffs concems regarding the use of l this method, where (1) the corresponding IRS exceeds the RS (1.5 times the GIP-2 bounding

)

s ,

5 spectrum), and (2) the building is embedded in shallow soil layers, the licensee has evaluated the relevant equipment using Method B from Table 4-1 of GIP-2. By letter dated April 1,1998, (Reference 12), the licensee stated that a total of 71 items out of all SSEL are located in areas of the plant where the IRS exceeds the RS, including four feedwater valves (FW-V-16 A&B) identified during the site audit. The licensee concluded that all outliers have been resolved without modifications, and that the screening Evaluation Work Sheets (SEWS) have been revised to document the revised evaluation using Method B of Table 4.1 of GlP-2. The staff considers this item resolved.

2.4.2 Assessment of Equipment" Caveats" in order to apply an experience-based approach and use the equipment seismic capacity defined in GIP-2, the plant-specific equipment must meet some restrictions or caveats described in GlP-

2. The licensee's SCEs verified that the inclusion and exclusion rules (called " caveats") listed in GIP-2 for each equipment class were met for TMI-1. The results were summarized in an appendix (Reference 5, Appendix H). In many cases, the caveats were not met, which resulted in the classification of these conditions as outliers. In a large number of cases, the licensee stated that the intent rather than wording of the caveats was met. GIP-2 allows engineers to verify whether the plant equipment conditions satisfy the caveats specified for a particular equipment class byjudging whether these conditions meet the " intent of the caveats" and not necessarily the words of the caveats. The staff requested additional information for a few related equipment items to understand how the intent of the caveats was met rather than the wording.

The staff found that additional analyses, data, or field modifications were required to demonstrate seismic adequacy of certain equipment, and, as such, determined that this equipment population should have been declared as outliers not meeting the caveats, but was resolved as a result of additional work, such as data collection and analyses, or field modifications.

This aspect of the licensee's implementation program was farther reviewed during the site audit.

For the sample cases selected for the audit, the staff fount that, for some cases, the plant equipment conditions can be considered adequate to satisfy the words of the caveats; justifications were needed for some other cases. The staff audit team noted that, in a number of cases, the licensee should have identified certain equipment as outliers instead of classifying them as " meeting the intent of caveats." The licensee agreed to reexamine all cases (not just the audit sample items) in the classification of ' meeting the intent of caveats" to determine if they should be reclassified as outliers. By letter dated April 1,1998, (Reference 12), the licensee stated that it has completed a review of all SEWS for thi. Items listed in Appendix J of Reference 5 to assure that the approach taken is consit tent with tiv approach described in GlP-2. As a result of the review, SEWS documentation has %n revised for 125 cases. Of these,30 cases were found to be outliers which have bet,. resolved without the need for any modifications. The staff considers this item resolved.

1 According to GlP-2 criteria, one general equipment caveat, "Any other concem," applies to all classes of equipment. During the site audit, the staffidentified a concem regarding the seismic adequacy of the grounding resistor, EG-Y-1B-GR, in the Diesel Generator Building. The concem

1 6

regards the brittle ceramic insulator for the grounding resistor. The licensee has performed a conventional structural analysis with questionable assumptions for determining the stresses ofthe ceramic insulator. It is noted that GIP-2 recommends the use of 20% of the ultimate strength as the tensile strength allowable for cast iron. The staff audit team questioned whether there is any test data to support the use of 40% of ultimate strength as the tensile strength allowable for ceramics, which have less energy absorption capacity than cast iron. By letter dated April 1, 1998, (Reference 12), the licenses stated that it has performed additional analysis of the forces on the diesel generator grounding resistors, taking into account the calculated frequency of the unit. The calculation resulted in a minimum factor of safety of 4.0 against the manufacturer's ultimate load for the ceramic insulators, and thus, assuring the seismic adequacy of the insulators. SEWS documentation has been revised to include the reference to the additional calculations. The staff considers this item resolved.

2.4.3 Equipment Anchorage The licensee verified equipment anchorage during the walkdown and documented the acceptability of equipment on the Screening Verification Data Sheets (SVDS, Appendix H, Reference 5). In certain cases, the anchorage was not visually inspected due to inaccessibility.

Torque checks were generally performed except in a few cases that involved plant operational concems. Anchorage modification was required for some equipment items. Subsequently, from the review of data packages and limited inspection performed during the site audit, the staff concluded that, in general, equipment was adequately anchored. The staff had two findings.

One finding is that the licensee used a type of anchor bolt called "Liebig Expansion Anchor," to l anchor some equipment. GIP-2 does not provide safe allowable loads for these bolts. The licensee provided a copy of a report which described the tests performed by the manufacturer at l the TMI site to arrive at the static load capacity of this anchor. The licensee had used a factor of

! safety of four for the allowable loads in shear and tension. The staff considers this safety factor reasonable.

The other finding is the use of plug welds that are relied upon as the sole means of connecting l equipment to the building structure. As the name signifies, the plug welds are used to plug holes in an overiying component (e.g., a plate) by depositing weld materials on an underlying structural component. This process usually ensures resistance against sliding of the overtying component over the undedying one, (i.e., shear resistance) by achieving fusion of the undertying metal.

However, the resistance against uplifts of the overtying component depends on the degree of g fusion achieved in its metal along the periphery of the hole. Unlike fillet welds, where fusion of

! both connecting components is a requirement, in plug welds, adequate fusion of metal of the l- component with the hole is not a requirement, although partial fusion at some areas is possible.

l Furthermore, there is a concem about buming of thin sheet metals. Therefore, the uplifting

resistance (i.e., tension capacity) of plug welds is uncertain. One example is the plug wolds used for the Foxboro power supply cabinet in the auxiliary building (PS-2), where the staff had difficulty in confirming the adequacy of the welds. Plug welds were also observed for engineered safeguards switchgear (IE-4160V-ES) located in the control building. The licensee agreed to confirm the adequacy of the plug-weld anchorage for these cabinets and for other equipment with plug-weld anchorages, and document the resolution. By letter dated April 1,1998,

7 (Reference 12), the licensee stated that it has completed a review of all SEWS documentatim to identify equipment which relies on plug welds for base anchorage. A total of 18 components with plug welds in the anchorage configuration were identified, which included six motor control centers, six 480 volt switchgear, two 4160 voit switchgear and four instnament and control cabinets. The licensee concluded that the anchorage for all 18 components is adequate without modifications. The staff considers this item resolved.

The staff audit team also noted, during its examination of the Engineering Safeguards Motor Control Center (IA-480-SHES) in the screen house, that there were four small breakers in the racked-out position. The licensee agreed to take appropriate action regarding this observation in accordance with NRC Information Notice 97-53, dated July 18,1997.

2.4.4 Seismic Spatial interaction Evaluation I

During the site inspection, the staff observed that the licensee had taken actions to preclude  !

spatial interactions by tioing vulnerable components. The acceptability of such interactions was judged and summarized with a "yes" or "no" answer in the SVDS (Appendix H, Reference 5).

There are examples where potential seismic interactions were identified by the licensee as concems. The technical data included in the licensee's summary report was not sufficient for an i independent evaluation. Selected cases were reviewed by the staff during the site audit. The i staff audit team identified four cases involving spatial interaction concems. One case involved

, the connection of a six-section motor control center (MCC) with adjacent switchgear in the

! screen house. No calculations were available for the design of this connection. The licensee agreed to verify the design to ensure that this connection is at least as strong as the connection between adjacent sections of the MCC. Another case involved the Reactor Protection System Cabinet, RPS-1B/2B located in the control room. The staff observed loose parts, such as large glass jars and cooking pots behind the cabinet in the kitchen area. Loose items, such as computers, paper trays, bottles, and trash cans were also observed in the console area of the 4 control room, where the operators sit. All these loose parts could fall Nwn, create obstructions to operation or could even injure the operators during a safe shutdoo earthquake (SSE). The third case involved the 480-V Engineering Safeguards Valve MCC, IC-480V-ESV which is very close (about 1/4 inch gap) to an adjacent conduit. The fourth case involved the large Boric Acid Water Storage Tank (BWST), due to its proximity to two other narrow vertical tanks (non l

seismic), that are separated from the BWST by space frames. The licensee was informed of the l

staff audit team's spatialinteraction concoms. By letter dated April 1,1908, (Reference it), the licensee stated that it had completed the review of these potential spatial interaction findings identified during the staff site audit. Most of the findings were resolved by performing new calculations or evaluations for compliance with the seismic housekeeping procedure (1401-18).

l The interaction potential between 1C 480V-ESV (MCC) and a nearby conduit has been

! eliminated by rerouting tbs conduit to provide a sufficient clearance with the MCC. The licensee I has also stated that it had looked for and found no similar spatial interaction concoms that have not been previously evaluated. The staff considers this issue resolved.

u__ ___ ._ _

8 2.5 Tanks and Heat Exchangers in Reference 5, the licensee identified 4g tanks and heat exchangers. In response to the staff questions on vertical tanks- (item 22, Reference 9), the licensee provided SEWS for 14 vertical tanks. As the evaluation of large flat-bottom vertical tanks is of special concem to the staff for the resolution of USI A-46, the staff reviewed the relevant SEWS in detail. A review of SEWS Indicated that the licensee's consultant EQE Intemational (EQE) had, in general, used the GlP-2 criteria for verifying the seismic adequacy of three large flat-bottomed vertica! tanks - two Condensate Storage Tanks (CST), and one BWST. The staff audit team also examined the Surge Tanks (NS T-1 and IC-T-1) and Air Starter Reservoirs 1 and 2 (EG-T001 A-1 and 2). The staff audit team's observations were as follows:

a. The analysis for the CSTs was performed by EQE, Inc. When outliers were identified, EQE used the source documents (the documents which formed the basis for the GlP-2 criteria) to calculate pertinent parameters. For example, the height to the radius ratio (H/R=0.816) of the CST does not meet the minimum ratio of 1.0 specified in GIP-2. As a first approximation, EQE used the GIP-2 steps using H/R=1.0, and performed sensitivity studies to assure that this is a conservative assumption. Using this approximation, the overtuming moment capacity calculated in Step 18 of Section 7.3 of GIP-2 was less than the demand moment calculated in Step 6. EQE refined the calculations using the correct H/R ratio, and proper consideration of horizontal impulsive modes and contributing water mass from each of the modes. The demand moment was verified to be less than the capacity. In the reFned analysis, EQE followed the GlP-2 criteria on safety factors and linear material behavior. The staff finds the analytical approaches used for verifying the tanks and heat exchangers acceptable.

During the site audit, the staff observed cracks in the foundations of the CST at some of the anchor-bolt locations. By letter dated April 1,1998 (Reference 12), the licensee agreed to evaluate the significance of these cracks as required by 10 CFR 50.65 (Maintenance Rule), using the evaluation criteria of ACI 349-3R-96, " Evaluation of '

Existing Nuclear Safety-Related Concrete CMures," and document the resolution. The l staff considers the licensee's commitment adequate to resolve this finding.

b. Boric Water Storage Tank l

This is a large tank located outdoors in the vicinity of the condensate storage tank. The licensee has verified the BWST tank seismic adequacy using criteria provided in Section 7 of GIP-2. However, two other long, narrow tanks separated by space frames l are located close to the BWST. As discussed in Subsection 2.4.4 of this SE, there is a i

potential scismic spatial interaction between the space frame and the BWST. By letter l dated April 1,1998, (Reference 12), the licensee stated that the Borsted Water Storage l Tank (DH-T-0001) was reviewed regarding the concem for potential interaction with an adjacent space frame supporting an access platform. A/E design calculations document the seismic design for the structure. The calculated maximum SSE deflection at the top of the platform structure is 0.18 inches. This deflection will not cause impact or damage I

L . -

g to the BWST. SEWS documentation has beer se ised to address the seismic deflection of the access platform. The staff considers this item resolved.

c. Surge Tanks, NS T-1 and !C-T-1 Both tanks are supported on legs and located in the fuel handling building at an elevation of 348 feet. The larger tank (MS-T-1) has cross bracings on the support legs while the legs of the other tank (IC T-1), which are longer, do not have cross bracings. This will .

potentially introduce bending in the angle iron legs oilC-T-1. The staff reviewed this as a l part of the site visit. The licensee's calculation indicated that the tank, its legs, and its anchorage are seismic. ally adequate. The sight gauge for NS-T-1 appeared to be made of glass. The licensee stated that even if the sight gauge breaks or the connecting pipe at the top of the tank is severed, resulting in substantial leakage, there is no signifcant safety concem in the event of a safe shutdown earthquake based on a plant-spacife calculation on loss of surge line inventory.

In a subsequent letter to the NRC from the licensee dated September 2,19g7, the licensee stated that GPUN calculation number C-1101-541536N2, 'NSCCW Loss of Surge Tank inventory," establishes that sufficient not positive suct on head (NPSH) is avail 6ble if tank inventory is lost. It is also stated that this calculation bounds the consequences of a break and loss of function of both the sight glass and the attached small-bore pipe, and that no further specife evaluation of the appurtenances is required.

This item is resolved.

d. Air Starter Reservoirs 1 and 2, EG-T-0001A-1 and 2 The air starter reservoirs are located in the diesel generator building at an elevation of 305 feet. These lightweight tanks are bolted to the pedestal through shims and are braced about halfway up the wall. The seismic adequacy of these tanks was verifed by the licensee and found to be acceptable by the staff.

2.6 Cable and Conduit Raceways A review of Reference 5 indicated that the iconsee had followed the guidelines and inclusion rules provided in Section 8 of GIP-2. Several outliers were identifed during the Seismic Review l Team (SRT) walkdowns of the raceway configurations in the Seismic Category I buildings. Many of these outliers were related to the seismic interaction potential in the Turbine Building.

Examples included (1) a traveling crane potentially impacting a conduit run was noted; (2) unanchored or inadequately anchored equipment (not on the Safe Shutdown Equipment List -

SSEL) might introduce an interaction hazard to adjacent raceways; and (3) seismic capacity of masor4ry block walls in Qe vicinity of cable trays and conduit runs was unknown. The affected Turbine Building raceways were evaluated in more detail and appropriate actions were taken to ensure that the safe shutdown capacity of the associated equipment would not be adversely affected. Detailed review of these issues was performed during the staffs site audit. The staff noted the following observations:

L

10

s. In the limited analytical review (l.AR), as a first scrooning, the licensee assumed all cable trays to be full. If the associated support did not pass the GIP-2 criterion for a vertical capacity check, the licensee used existing cable fill loads to demonstrate their seismic adequacy. As the selected supports represented the worst-case scenarios, the audit team asked the licensee if the partial-fill analyses were applicable to the non-analyzed supports, and if there are procedures that would ensure against adding more cables in the future. The licensee responded that the written design controls at the plant would not l allow adding more cables without reanalyzing the affected supports. The staff audit team l considered the response acceptable.

b. According to GIP-2 guidelines, the licensee selected 18 raceway supports for an L.AR. ,

These supports were judged to be representative bounding samples of the major types of J

the raceway support configurations in the plant. All supports passed the analytical review l using the acceptance criteria of Section 8 of GlP-2, except for five supports which failed the vertical capacity check. These five supports (four supports were characterized as )

' ductile,' and one support was characterized as 'non-ductile') were further analyzed for the i

lateral load check. These outliers were resolved on the basis that the lateral load check acceptance criteria were satisfied,

c. During the inspection, the staff audit team identified two support systems that might not have been enveloped by the samples selected for the LAR. The licensee was asked to analyze these two support configurations to dernonstrate the robustness of the selection process in the LAR. In its September 2,1997, letter (Reference 4), the licensee stated that it had evaluated the two support systems as supplemental cases, and determined that neither of the systems would pass the GIP-2 vertical capacity check criterion. The two support systems were subsequently identified as outliers. The outliers were resolved by evaluating them for the calculated lateral load check, and the supports were found to be acceptable.

~

Overall, the staff considers the licensee's evaluation of cable and conduit raceways to be acceptable.

2.7 Essential Relays The licensee stated that the essential relays were examined during a walkdown by a seismic l capability engineer and a relay evaluation engineer. The walkdown included spot checking of mounting, orientation, model number, load path, possible interaction, and cable slack in accordance with GlP-2 criteria and procedures. The licensee judged that the " low ruggedness" relays in radium and low voltage switchgear to be acceptable by taking advantage of operator l

action. The staff had a concem about the potential over-reliance on operator action. The licensee's proposed approach in response to the staffs audit concem is described in Section 3.4 of Appendix A, and is considered reasonable. The staffs overail review of operator action is

detailed in Section 2.8 of this SER.

l I

11 Outlier relays were identified in the summary report (Reference 5, Appendix K). The staff further evaluated selected relays during the site audit. Many relay models at TMI-1 were not covered by the original SQUG/EPRI database that was reviewed by the staff during the GIP-2 approval process. However, the licensee stated that these additional out-of-GIP-2 scope relays were tested and found to be acceptable (GPUN relay calculation C-1101-900-5326-025). The licensee's action was considered reasonable and acceptable.

2.8 Human Factors Aspect The licensee provided information which outlined the use of the " desk-top" evaluation method and simulator exercises by a licensed operating crew consisting of an actual control room shift and the typical compliment of auxiliary operators (e.g., a shift supervisor, two shift foremen, three control room operators, five raxiliary operators) to verify that existing normal, abnormal and emergency operating procedures were adequate to mitigate the postulated transient and that operators could place and maintain the plant in a safe shutdown condition. The staff verified that the licensee had considered its operator training programs and verified that its training was sufficient to ensure that those actions specified in the procedures could be accomplished by the operating crews. The licensee stated that both initial training and requalification training includes utilization of procedure EP-1202-30, " Earthquake," to respond to seismic events. The actions required to place the plant in a hot shutdown condition, using only the equipment available in the SSEL, following a seismic event are included in initial licensed operator training and requalification training.

In addition, the staff requested verification that the licensee nad adequately evaluated potential challenges to operators, such as lost or diminished lighting, harsh environmental conditions, potential for damaged equipment interfering with the operator's tasks, and the potential for placing an operator in unfamiliar or inhospitable surroundings. The licensee provided information regarding the simulator evaluations performed on February 21,1995, and the " desk-top" evaluation to substantiate that operator's actions could be accomplished in a timeframe required to mitigate the transient. Specifically, the licensee provided assurance that ample time existed for operators to take the required actions to safely shut down the plant and evaluated all equipment in terms of location, availability of operators, and complexity of the task. The licensee verified that existing procedures, availability of lighting equipment, and operator training were adequate to ensure the operators could perform the required actions credited in the submittal.

The licensee verified that all required actions were grouped in easily accessible and familiar areas of the plant. The licensee further stated that earthquake experience has shown that typical industrial grade equipment and structures are inherently rugged and not susceptible to damage at USl A-46 plant SSE levels. Therefore, the potential for physical barriers resulting from equipment or structural earthquake damage which could inhibit operator ability to access plant equipment is not considered to be a significant hazard. The licensee has provided the staff with sufficient information to demonstrate conformance with the NRC-approved review methodology outlined in GIP-2.

2.9 Outleer identification and Resolution l The licensee identified the equipment outliers resulting from the A-46 implementation effort in the summary report (Reference 5, Appendix K). In most cases, a resolution path was also described. The staff found that there were other equipment items which shouH have been on

}

4 ,

l 12 )

the outlier list, (e.g., equipment not covered by GIP-2 or not meeting the caveats). This concem was expressed in the RAI (Reference 7), and the response (Reference 9) provided by the l licensee did not include adequate data to address this issue. This was further evaluated during the site audit and the resolution is described in Section 2.4.2 of this report. In limited cases, the licensee resolved equipment outliers by using technical information that is neither contained in GlP-2 nor was reviewed by the staff during the GlP-2 review process. For example, the licensee used certain test data for an expansion anchor (Liebig expansion anchor) not included in GIP-2, which was further rev6ewed by the staff after the audit, and the evaluation is described in Section 2.4.3 of this report. (

in summary, the licensee stated that its seismic walkdowns identified 152 components as outliers .

(not meeting certain provisions in GIP-2, or requiring further information). Outliers for a total of 53 mechanical and electrical components were analyzed further and resolved as acceptable 4 without modifications. A total of 99 mechanical and electrical components required some form of modification to ensure their seismic adequacy. The licensee indicated that the modifications for all 99 outlier components identified in the Summary Report (Reference 5) were completed at the i conclusion of Refueling Outage UR in 1997. Note that, as a result of " caveat" reassessments described in Section 2.4.2, additional equipment items may be reclassified as outlier items. The licensee agreed to resolve all outlier items accordingly and document the resolutions. l 3.0

SUMMARY

OF STAFF FINDINGS 4

The staffs findings resulting from its review of the licensee's USl A-46 Implementation program are provided for each review area discussed above. The staff did not find any significant or programmatic deviation from GlP-2 regarding the walkdown and the seismic adequacy evaluation 3 at TMI-1. i l

4.0 CONCLUSION

~

The licensee's USl A-46 program at TMI-1 was established in response to Supplement 1 to GL 87-02, issued May 22,1992. This GL requested information regarding the seismic adequacy of mechanical and electrical equipment under the provisions of 10 CFR 50.54(f). The licensee conducted the USl A-46 program implementation in accordance with GIP-2. The licensee's submittal on the A-46 implementation indicated that the safe shutdown equipment list contained 679 components of which 658 were examined during walkdowns to verify their seismic l

adequacy. The licensee indicated that of the 658 components examined during the walkdowns,506 components were verified to be seismically adequate, and that the balance,152 components, were classified as outliers. According to the licensee,53 of the outliers were resolved by analysis and found to be acceptable, and 99 needed some form of modification. The licensee indicated that the modifications for all 99 outlier components identified in the Summary l Report (Reference 5) were completed at the conclusion of Refueling Outage 12R in 1997. In its i

April 1,1998, letter (Reference 12), the licensee provided an updated tally of outliers as a result l

of staffs concems descnbod in Sections 2.4.2 and 2.9. The total number of outliers, the number of components that required modifications, and the number of outliers resolved by analysis was slightly revised from the numbers cited above.

The licensee's implementation report stated thet the walkdown has identified five components f that were found to have design basis deficiencies. The licensee indicated that these deficiencies l

l

• A 13 were documented and repaired immediately with the issuance of Material Non-Conformance Reports (MNCR). The licensee did not identify any other instances where the operability of a particular system or component was questionable. After the staffs site audit, the licensee examined Appendix K of the Summary Report (Reference 5). The resulting updated tally of outliers is reported in the letter dated April 1,1998 (Reference 12).

The staff concludes that the licensee's A 46 implementation program has, in general, met the purpose and intent of the criteria in GIP-2 and the staffs SSER No. 2 on GIP-2 for the resolution of USl A-46. The staff has determined that the licensee's already completed actions and l commitments to resolve the staffs review and audit findings discussed in Section 3.0 of this l SER, will result in safety enhancements, in certain aspects, that are beyond the original licensing l basis. As a result, the licensee's actions provide sufficient basis to close the USl A 46 review at l .the facility. The staff also concludes that its findings regarding the licensee's implementation of USl A-46 do not warrant any further regulatory action under the provisions of 10 CFR 50.54(f).

Licensee activities related to the USI A-46 implementation may be subject to NRC inspection.

Regarding the future use of GlP-2 in licensed activities, the licensee may revise its licenshg l

basis in accordance with the guidance in Section 1.2.3 of the staffs SSER No. 2 on SQUG/GlP-2,

! (Reference 2) and the staffs letter to SQUG's Chairman, Mr. Nsil Smith on June 19,1996.

Where plants have specific commitments in the licensing basis with respect to seismic l qualification, these commitments should be carefully considered. The overall cumulative effect l of the incorporation of the GIP-2 methodology, considered as a whole, should be assessed in l making a determination under 10 CFR 50.59. An overall conclusion that no unroviewed safety l question (USQ) is involved is acceptable so long as any changes in specific commitments in the l licensing basis have been thoroughly evaluated in reaching the overall conclusion. If the overall i

cumulative assessment leads a licensee to conclude a USQ is involved, incorporation of the

! GlP-2 methodology into the licensing basis would require the licensee to seek an amendment under the provisions of 10 CFR 50.59.

Principal Contributors: P. Y. Chen, H. Ashar, K. Desai, G. Galletti Date: August 12, 1998 i

l l

a 4

1 l

1 14

5.0 REFERENCES

1. " Generic implementation Procedure (GIP) for Seismic Verification of Nuclear Power Plant Equipment," Revision 2, corrected February 14,1992, Seismic Qualification Utility Group.

2. NRC " Supplement No.1 to Generic Letter 87-02 Including Supplemental Safety Evaluation Report No. 2 on Seismic Qualification Utility Group's Generic implementation ,

Procedure, Revision 2, corrected February 14,1992," dated May 22,1992.

3. Letter, GPU Nuclear Corporation to NRC Document Control Desk, " Response to l Supplement 1 to Generic Letter 87-02, SQUG Resolution of USl A-46," dated '

September 21,1992.

4. Letter, NRC to GPU Nuclear Corporation, " Evaluation of the Three Mile Island Nuclear Station, Unit 1,120-Day Response to Supplement No.1 to Generic Letter 87-02," dated November 18,1992.

5. "Three Mile Island Unit 1 USl A 46 Seismic Evaluation Report," EQE Intemational Report No. 42105-R-002, dated May 11,1995, transmitted with letter from T.G. Broughton (GPUN) to NRC Document Control Desk, dated May 17,1995.

6. Letter, NRC to GPU Nuclear Corporation, " Request for Additional Information," dated August 29,1995.

7. Letter, B. C. Buckley, NRC, to J. Kniebel, GPUN,

• Request for Additional information on the Resolution of Unresolved Safety issue (USI) A-46 (Generic Letter 87-02) at Three Mile Island Nuclear Station, Unit 1," dated February 24,1997.

8. Letter, J. Kniebel, GPUN, to NRC Document Control Desk, " Response to NRC Request for Additional information," dated September 27,1995.

9. Letter from J. W. Langenbach, GPUN, to NRC Document Control Desk, " Response to the NRC's Request for Additional information on the Resolution of Unresolved Safety issue A-46 at Three Mile Island Nuclear Station, Unit 1," dated May 28,1997.

Letter, R. W. Heman, NRC, to T. G. Broughton, GPUN, " Evaluation of in-Structure 10.

Response Spectra for Resolution of Unresolved Safety issue A-46 (TAC No. M69486),"

dated September 26,1994.

11. Letter, Ted Noble, GPUN, to NRC (Attn.: P.Y. Chen), " Additional Material- July 1997 TMI USI A-46 Audit," dated September 2,1997.

12. Letter, J. W. Langenbach, GPUN, to NRC Document Control Desk, " Response to the

! NRC's Request for Additional Information Regarding USl A-46 (GL 87-02, Verification of l

Seismic Adequacy of Mechanical and Electrical Equipment in Operating Reactors)," dated l April 1,1998.

4 ,

l l APPENDIX A

( A SITE AUDIT REPORT E9B l USI A-46 PROGRAM IMPLEMENTATION AT THREE MILE ISLAND. UNIT 1 DOCKET NO 50-289. OPERATING LICENSE DPR-50

1.0 INTRODUCTION

Duiing the week of July 21-25,1997, a team of two NRR staff members, from the Mechanical l Engineering Branch and the Civil Engineering and Geosciences Branch, and one contractor from l Brookhaven National Laboratory (BNL), conducted an audit of USl A-46 seismic walkdown j results performed by the licensee of the Three Mile Island, Unit 1, plant (TMI-1). A Senior inspector from Region I, joined the audit team on the aftemoon of July 24,1997. The licensee (GPU Nuclear, Inc., or GPUN) has adopted the Generic implementation Procedure, Revision 2 (GlP-2), developed by the Seismic Qualification Utility Group (SQUG) and approved by the NRC in its supplemental Safety Evaluation Report No. 2 (SSER 2), dated May 22,1992, for the A-46 program implementation at TMI-1. The objective of the audit review was to verify that the l licensee's implementation of the USl A-46 program was completed and documented in accordance with the licensee's commitments and the general requirements of GIP 2 criteria and l procedures, supplemented by the staff's SSER-2.

l l An entrance meeting on July 21,1997, and an exit meeting on July 25,1997, were held at the

! plant site. The participants of these two meetin9s are listed in Attachment 1.

This audit report details the staff team's audit results, observations, summary and conclusions.

I The staff audit team examined the licensee's implementation of the USl A-46 program at TMI-1.

l The audit team found that, in general, the licensee conformed to GIP-2 in implementing the USl A-46 program at TMI-1. The audit team observed a few anomalies that required further actions; these were noted to the licenses and are described in this audit report.

GPU Nuclear, Inc., (GPUN) conducted a seismic walkdown for the USl A-46 program at Three Mile Island Nuclear Station, Unit 1 (TMI-1) and submitted a summary report on May 17,1995, (Reference 1). The staff reviewed the summary report and sent two requests for additional information (RAI) to GPU on August 29,1995, and February 24,1997. GPUN responded to the RAls in letters dated September 27,1995, (Reference 2) and May 28,1997, (Reference 3).

After reviewing References 2 and 3, the staff noted that some areas in the RAls were not satisfactorily addressed. Therefore, the staff decided to conduct a site audit of the licensee's seismic implementation program on July 21-25,1997. ~ his report describes the audit activ; ties, the staff audit team's observations and conclusions.

2.0 AUDIT INSPECTIONS 2.1 NRC Audit Team Members The staff audit team was led by PebYing Chen of NRR's Mechanical Enginsedng Branch (EMEB), and included Hansnqi Asher of NRR's Civil Engineering and Geosciences Branch (ECGB), and Kamal Bs,r.?icpst,ri of Brookhaven National Laboratory (BNL), a contractor.

i

I 4 ,

2 2.2 Audit Process The audit was conducted in accordance with a plan given to the licensee preceding the audit.

The audit process consisted of the following three major elements:

. review of the overall USl A-46 program impamentation process through ,

discussion with the licensee's engineers and examination of documents e discussion of the licensee's responses to the RAls, including review of additional data e inspection of selected mechanical and electncal equipment including relays, tanks, and cable and conduit supports Each of these elements is further described in the sections that follow.

2.2.1 Review of overall USl A-46 Program The staff audit team discussed the overallimplementation of the USl A-46 program at TMi-1 with i the licensee's engineers, and focused on the following topics: )

. selection process of a USl A-46 safe shutdown equipment list

. personnel qualification and team organization of review engineers

. participation of, and interaction with systems and plant operations engineers e relay review process and resolution of " low ruggedness" relays e selection of cable and conduit raceway supports for analytical evaluation

! . Judgment of " meeting the intent of caveats" e resolution of outilers j e completion schedule for field modifications and further calculations The staff audit team discussed various subjects with the appropriate systems, operations, mechanical, electrical, Instrumentation, and structural engineers who participated in the program.

The following is a list of key participating engineers:

Ted Noble GPUN Gordon Skillman GPUN Rick McGoey GPUN Jack Mancinelli GPUN William Heysok GPUN/TMl Yosh Nagai GPUN l Walt Duda GPUN l Jim Snell NSA I Steve Ku GPUN Jim White EQE Paul Baughman EQE Richard Starck MPR Associates G.J. Bixler GPUN/TMI f

I i

l

{

l

! 3 The licensee reported that the safe shutdown equipment list (SSEL) was prepared in several joint sessions and the formal walkdown was preceded by a preliminary " walk by." The " walk by" I gave the review engineers an opportunity to become familiar with the SSEL items, to obtain a preliminary assessment of the plant conditions, and to ider.tify auxiliary equipment items that t would help in preparing logistics for the subsequent formal walkdown. The licensee illustrated,

for the staff audit team, its equipment identification process by considering one safe-shutdown path. -

The formal walkdown was performed by two teams, although in a particular team the individual rasmbers were not necessarily the same and sometimes members switched from one team to

, another. Each team consisted of at least two seismic capability engineers (SCEs) with at least

! one from EQE Intemational. The following SCEs participated in the walkdown:

I R. D. Augustine EQE Intemational l l- P. D. Baughman EQE Intemational J. P. Conscente EQE Intemational R. W. Cushing EQE Intemational T. R. Kipp EQE Intemational J. L White EQE Intemational l R. A. Svotells Gilbert Commonwealth M. Gotthard GPUN l S. M. Lazorchak GPUN I T. H. Noble GPUN S. C. Ramdeen GPUN E.Y. Tang GPUN K. L Whitmore GPUN l Dr. J. Stevenson, of Stevenson Associates, performed the " Third Party" review.

After the identification of safe-shutdown equipment items by the systems engineers, the associated relays were listed by the electrical and instrumentation engineers.

The staff audit team reviewed a sample of the licensee's documentation including calculations for equipment anchorage, tanks, and cable tray and conduit supports.

l 2.2.2 Response to Requests for Additional Information The staff audit team discussed with the licensee the responses to the staffs RAls (References 2 and 3). During the audit, the licensee provided further clarification and additional data for appropriate disposition of the RAI items. The following were the major topics:

• Potential overreliance on operator action

• Equipment seismic capacity compared to salsmic demand (i.e., in structure response spectra exceeding 1.5 times the bounding spectrum)

• Judgment of " meeting the intent of caveats," vs. classification as " outliers"

4

. Additional data packages for selected equipment items The licensee's responses contained the resolution of RAI items regarding third party review comments.

2.2.3 inspection of Selected Components The following electrical and mechanical equipment items, relays, cable tray and conduit supports, and tanks were selected for a sample inspection:

• Reactor Protection System Cabinet, RPS 1B/2B

. Engineered Safeguards Bus 1E,4160-V ES

• Relay Panel, NRP1B e Regulated Transformer, 20-VA 480/120-V,1 B

. Foxboro Power Supply Cabinet, PS-2

. Isolation Valve, NR-V-0016A

.- Surge Tanks, NS T-1 and IC-T-1

. Condensate Storage Tank (CST)

. Bonc Water Storage Tank (BWST)

. Air Starter ReservoirTanks

. Feedwater Block Valves, FWV928 and FWV16B, e River Water Pump, NR-P-1B

. Screen House Engineered Safeguards Motor Control Center,1 AES-MCC

• Conduit Support in Screen House

. Grounding Resistor, EG-Y-1B-GR

• 1CES Valve and Heating MCC The plant was in operation during the audit inspection and most electrical equipment was energized. Therefore, the access was limited and the inspection was mostly visual and extemal.

Some of the equipment in the preceding list was selected for confirmation of certain aspects of the licensee's reviews, such as spatial interaction orjudgment of " meeting the intent of a caveat,"

and the sample inspection was limited to serve these objectives. The extent of the inspection and the observations are discussed on an item-by-item basis in Attachment 2.

3.0 OBSERVATIONS The major observations that resulted from the staff audit team's inspection are described in this section. Specific observations regarding the components inspected during the audit are described in Attachment 2.

1 3.1 Overall USl A-46 Program implementation in general, the licensee conformed to GIP-2 in implementing the USI A-46 program. The documentation provided before 6nd during the audit were sufficient ar,d adequate for the staff to evaluate this program. Although GIP-2 allows and even emphasizes exercise of judgment over standard engineering calculations, the licensee improved the quality of the program by L

5 performing analytical calculations in many situations where the licensee concluded that the calculation was necessary to support judgments. Structural modifications were made in many instances to demonstrate seismic adequacy. The staff audit team noted a few anomalies that required further action, and these were noted to the licensee and described in Section 4.0 of this l report.

3.2 Status of USl A-46 Program At the time of the alte audit, the licenses expected to complete the remaining modifications and necessary documentation, hv the 12R outage or before December 31,1997. In a subsequent telephone conversation wie, the licensee, the licensee indicated that the modifications for all 99 outlier components identified in the Summary Report (Reference 1) were completed at the conclusion of Refueling Outage 12R in 1997.

3.3 Personnel Qualification of Review Engineers The review engineers were found to be competent and their qualifications exceeded the minimum requirement of GlP-2. The licensee reported that during the review process, the seismic engineers worked with the systems and operating engineers to obtain clarifications on equipment functionality.

3.4 OperatorAction l A review of the summary report (Reference 1) indicated that a potential overreliance on operator action may be required for safe shutdown after an SSE (Reference 2). During the audit, the licensee explained that many of the operator action items are for equipment on the third train (i.e., optional according to GIP-2). Furthermore, recovery of all failed items may not be necessary for the safe-shutdown sequence. The licensee informed the staff audit team that at any given time, there will be at least three main operators in the control room and five auxiliary i

operators located in any of the six buildings in the plant. The licensee maintains that the SSE will not cause structural damage to the extent that the operators' accessibility and ability to function l will be restricted. Although the staff audit team noticed adequate anchoring or tioing of non i

seismic equipment, there were examples of components with potential damage vulnerability.

Specifically, the staff audit team noted the control room condition where the operators may be injured if non-structural items should fall. The licensee agreed to correct this potential spatial interaction issue as described in item 1 of Attachment 2.

In order to evaluate the concem of potential overreliance on operator actions in light of the overall plant procedure for safe shutdown, the licensee agreed to take the following actions:

a. The licensee will list all cases in which credit was taken for operator actions (or list equipment that relied on operator actions to perform its function in the event of an earthquake) and the estimated time required for GPUN operators to arrive at the equipment location to remedy the undesired conditens.

I i ,

O l

1 1

b. The licensee will identify and document equipment in the third train that was reported 1 in the summary report, and for which GPUN considers that operator actions may no I longer be required.

c. The licensee will document information about the number of operators on a shift and identify each individual's normal duties and additional responsibilities in covering the above-mentioned operators' actions. A marked-up set of General Arrangement Drawings will identify the area in the plant where the equipment listed in item (a) above are located.

The staff finds the licensee's proposed approach outlined above to be reasonable. The staffs

( overall review of the human factors aspect relating to the A-46 program implementation, is

! discussed in Section 2.8 of the SER.

3.5 Equipment Seismic Capacity Compared to Seismic Demand i

The seismic c pacity of equipment classes covered in GIP-2 (Reference 1) was used by the licensee for determination of the seismic adequacy and compared with either the ground response spectra or the irkstructure response spectra as recommended by GIP-2. The capacity l spectra are provided in GIP-2 as a Bounding Spectrum, a Reference Spectrum (i.e., Bounding l Spectrum multiplied by 1.5) or Generic Equipment Ruggedness Spectra (GERS). Examples of how the capacity spectra were used to compare wkh the demand were reviewed at the site audit.

In response to the staffs questions on the use of Method A from Table 4.1 of GIP-2 for

comparing the seismic demand to the seismic capacity of the equipment, which can be generically represented by the GIP-2 bounding spectrum, the licensee dated (item 26, Reference 9 of the SER) that, in general, it has used Method A for equipment located within 40 feet above the effective grade of the building. However, recognizing the staffs concems regarding the use of this method (1) where the corresponding IRS exceeds 1.5 times the GlP-2 bounding spectrum, and (2) where the building is embedded in shallow soil layers, the licensee has evaluated the relevant equipment using Method B from Table 4-1 of GlP-2. During the NRC site audit, the licensee informed the staff audit team that it has identified two valves (feedwater valves FWV92B and FWV168) in the Intermediate Building that required further evaluation using Method B. The licensee agreed to document the resolution.

3.6 Meeting " Intent of Caveats" GIP-2 allows engineers to verify whether the plant equipment conditions satisfy the inclusion and exclusion rules (called

• caveats") specified for a particular equipment class byjudging whether l

these conditions meet the " intent of the caveats

• and not necessarily the written words. The summary report (Reference 2) identifies a large number of cases in which the judgment of

" meeting the intent of caveats" was used. For the sample cases selected for the audit, the staff found that, in some cases, the plant equipment conditions can be considered to satisfy the words l of the caveats; justifications were needed for some other cases. The staff audit team noted that in a number of cases, the licensee should have identified certain equipment as outisers instead of providing justifications that they satisfy the intent of caveats. The licensee agreed to reexandne 1

l l

l

7 all cases (and not just the sample audit items) in the classification of " meeting the intent of caveats" to determine if they should be reclassified as outliers.

3.7 Anchorage From the review of data packages and limited inspection during the staff's site audit, the staff audit team concluded that in general, equipment was adequately anchored. The staff audit team had two findings. One finding is that the licensee had used a type of anchor bolt called "Liebig Expansion Anchor," to anchor some equipment items. At the staff audit team's request, the licensee provided a test report that has been reviewed by the staff and is discussed in Section 2.4.3 of the SER. The other finding is the use of plug welds that are relied upon as the sole means of connecting equipment to the building structure. One example is the plug wolds used for the Foxboro power supply cabinets in the auxiliary building (PS-2), where the staff had difficulty confirming the adequacy of the welds. The licensee agreed to confirm the adequacy of the plug-weld anchorage for these cabinets and for other equipment with plug-weld anchorage and document the resolutions.

The staff audit team also noted, during its examination of the Engineering Safeguards Motor Control Center (IA 480-SHES) in the screen house, that there were four small breakers in the racked-ou' positions. The licensee agreed to take the appropriate actions to address this observat.*.a in accordance with NRC Information Notice 97-53, dated July 18,1997, 3.8 SpatialInteraction During the site inspection, the staff audit team observed that spatial interaction was, in general, precluded by tioing vulnerable components. However, the staff questioned the adequacy of the connection of a six-section motor control center (MCC) with adjacent switchgear in the screen house. No calculations were available for the design of this connection. The licensee agreed to ]

j verify the design and to ensure that this connection is at least as strong as the connection .

l between adjacent sections of the MCC or switchgears. Additionalitems relating to spatial interaction issues are described in Attachment 2 to this audit report. j 3.9 Data Outside GIP in limited cases, the licensee used technical information that is neither contained in GIP-2 nor was reviewed by the staff during the GlP-2 review process. For example, the licensee used certain test data for an expansion anchor (Liebig anchor bolt) not included in GIP-2. The licensee proMed information relating to this anchor in support ofits use to establish the seismic adequacy of the affected equipment. Another example related to the uncertainty involving the seismic adequacy of ceramic insulators in a grounding resistor. The staff audit team questioned the adequacy of the ceramic insulators when subjected to dynamic loading involving potential impact. It is noted that GlP-2 recommends the use of 20% of the ultimate strength as the tensile strength allowable for cast iron. The staff audit team questioned the licensee's use of 40% of the ultimate strength as the tensile strength allowable for the bnttle ceramics that have less energy 4

l J

8 j 1

(

absorption capacity than cast iron. The licensee agreed to reassess the seismic adequacy of the i ceramic insulators. {

3.10 Relay Evaluation Many relay models at TMI-1 were not covered by the original SQUG/EPRI database that was q

reviewed by the staff during GIP-2 approval process. However, these additional out of-GIP scope relays were tested and found acceptable (GPUN relay calculation C 110190Ch532Ch025).

There are several " low ruggedness" relays at TMI-1. The licensee justified their use either by showing that they perform only an annunciation function or by taking credit for operator action.

Regarding the latter, the staN questioned an apparent over-reliance on operator schon. This issue is discussed in Section 3.4 of this audit report.

3.11 Cable Tray and Conduit Raceway Supports The staff observed that the licensee, in general, followed the GIP-2 approach in evaluating the cable tray and conduit raceway supports. Regarding the sample supports selected for analysis, the staff observed the following:

a. As a first screening, the licensee assumed all cable trays to be full. Some of the supports did not pass GlP-2 criteria with this conservative assumption.

Subsequently, for those cases, the licensee used existing cable tray fill loads to show their seismic adequacy. As the selected supports represented conditions of maximum design capacity, the staff audit team asked the licensee if the partiabfill analyses were applicable to the non-analyzed supports and if there are procedures in '

place that would ensure against adding more cables in the future. The licensee responded that the design control at the plant would not allow adding more cables without reanalyzing the affected supports. The staff audit team considered the response acceptable, ,

b. The licensee assumed that some of the cable tray support systems are " ductile," as the term was used in GIP-2. Because of this assumption, the licensee did not have to {

perform a lateralload check according to GIP-2. However, the staff audit team i noticed one particular support system configuration (Selection No. 2 of the Cable Tray Limited Analytical Review, one of the 18 raceway supports selected for a limited analytical review mentioned in Section 2.6 of this SER) that may not fit the definition of the " ductile" support implied in GIP-2. The licensee agreed to reassess the resolution for this support configuration.

a During the examination of cable tray supports, the staff audit team identified two ,

support systems that may not be enveloped by the samples selected for the analyses.

The licentee was asked to compare these two supports with the samples selected for l analys!s and confirm that its selection process was robust. In a letter dated 1 i

September 2,1997, the licensee stated that it had evaluated the two support systems as supplemental cases, and determined that neither of the systems would pass GIP-2

]i l

I

9 l verticalload check. They were identifed as outliers. Subsequently, the outliers were

! resolved by evaluating them for the calculated lateral load check, and the supports i were found to be acceptable.

i 3.12 Tanks l

l The staff audit team observed that the licensee used GlP-2 criteria wherever applicable to l evaluate the TMI-1 tanks. When the TMI-1 tank parameters were not enveloped by GIP-2, reasonable methods were used for evaluating those tanks. Specific evaluations are described in items 7,8,9, and 10 of Attachment 2 to this audit report.

4.0

SUMMARY

AND CONCLUSIONS

( The licensee's USI A-46 program was, in general, implemented in accordance with GIP-2. High

! reliance has been placed on operator actions and the licensee has provided additional information to justify the apparent overreliance on operator action to verify seismic adequacy of

equipment. The licensee provided additional information and data that were outside GIP-2 and l were used for the TMI-1 seismic implementation program. On the basis of the discussions described above and in Attachment 2, the staff audit team has identified severalitems that the licensee has acknowledged as requiring further action. These items are detailed in Section 3.0 cf the SER.

5.0 REFERENCES

l 1. Letter, T. G. Broughton, GPUN, to NRC, "USl A-46 Seismic Evaluation Report,"

May 17,1995.

l 2. Letter, J. Kniebel, GPUN, to NRC Document Control Desk, " Response to NRC Request for Additional information," dated September 27,1995.

3. Letter, J. W. Langenbach, GPUN, to NRC," Response to the NRC's Request for Additional Information on the Resolution of Unresolved Safety issue A-46 at Three Mile Island Nuclear Station, Unit No.1," dated May 28,1997.

I i

1 l

1 i

-__ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . ]

w- e List of Meetina Attendees SamDie Audit of USl A-46 Seismic Walkdown Results at TMI-1 Entrance Meetina on July 21.1997 HRmt Company Hansraj Ashar NRC, ECGB/DE Pei-Yino Chen NRC, EMEB/DE Sam Hansell NRC Resident inspector Tom Basso GPUN l S. C. Ramdeen GPUN Ted Noble GPUN Gordon R. Skillman GPUN Rick McGoey GPUN Jack Mancinelli GPUN William Heysek GPUN/TMl Yosh Nagai GPUN Walt Duda GPUN Jim Snell NSA Steve Ku GPUN ,

Jim White EQE Paul Baughman EQE ,

Kamal Sandyopadhyay BNL )

Richard Starck MPR Associates G. J. Bixler GPUN/TMI-1 Exit Meetina on July 25.1997 Namt Company l

S. C. Ramdeen GPUN Pei-Ying Chen NRC/NRR Kama! Bandyopadhyay BNL l Jay Moore, Jr. GPUN/NSCE .

Yosh Nagai GPUN l William Heysek GPUN/TMI-1 ,

Ken Whitmore GPUN Paul Baugham EQE '

Richard Starck MPR Associates Tom Basso GPUN/TMI-1 i

, Jim Whde EQE l Ted Noble GPUN/TMI-1 Jim Snell GPUN Walter Duda GPUN i

ATTACHMENT 1 i

l E__

l w * , )

l ATTACHMENT 2 TO THE TMl SITE AUDIT REPORT EQUIPMENT-SPECIFIC AUDIT INFORMATION:

The NRC staff audit team examined the following components during the site audit.

1. Reactor Protection System Cabinet, RPS-1B/2B This is a two-bay cabinet located at an elevation of 355 feet in the control building, control room.

The cabinet is welded to an embedded steel plate, which is anchored to an elevated concrete pedestal. Five outlier items for this cabinet were resolved. However, loose parts, such as large glass jars and cooking pots, were observed behind the RPS cabinet that may create a spatial interaction. Loose items, such as computers, paper trays, bottles, and trash cans, were also observed in the console area of the control room, where the operators sit. Loose objects may fall down, create obstructions to operation or could even injure the operators during the safe-shutdown earthquake (SSE). The licensee agreed to correct this potential spatial interaction issue.

2. Engineered Safeguards Bus,1E-4160V-ES This is a 14-section switchgear assembly le::sted at an elevation of 338 feet 6 inches in the control b#1ing, switchgear room. The switchgear frame is plug welded to an embedded steel plate. Thw switchgear assembly item was initially identified as an outlier because it contains 21 essential relays that were not evaluated. These relays were later evaluated and resolved in r

accordance with GIP 2 criteria and procedures.

3. Relay Panel, NRP1B The relay panel is located in the intake pump house (IPH) building at the base elevation of 312 feet. The panel enclosure is bolted to the wall. The enclosure doors were observed to remain loose even after locking. These doors will likely rattle and may cause relay chatter during an earthquake; a tighter locking system would mitigate such an undesirable effect. The licensee agreed to correct the condition.

4. 120-V Regulated Transformer,1B l The 20K-VA 480/120 V transformer is located in the control building switchgear area at an elevation of 322 feet. The transformer cabinet is bolted to the floor through clip angles. A photo shown during the audit indicated that the transformer coil was anchored to a steel frame. This

! installation was considered acceptable.

5. Foxboro Power Supply Cabinet, PS-2 The power supply cabinet is located in the auxiliary building at an elevation of 305 feet. The i

cabinet is anchored to an embedded steel plate with plug welds, some of which were observed l to be partially filled. The licensee agreed to confirm the adequacy of the anchorage of this cabinet and document the resolution.

ATTACHMENT 2

t

• t l 2

6. Isolation Valve, NR-V-0016A This isolation valve is located in the heat exchanger vault of the auxiliary building at a floor elevation of 271 feet. The Limitorque Operator (model SMB-0002) was mounted upside down.

The licensee agreed to verify the seismic adequacy of this unusual mounting with the vendor and doc'Jmont the resolution.

7. Surge Tanks, NS-T-1 and IC-T-1 l Both tanks are supported on legs and located in the fuel handling building at an elevation of 348 feet. The larger tank (NS-T-1) has cross bracings on the support legs; the legs of the other tank l

(IC T-1), which are longer, do not have cross bracings. The licensee presented the results of its calculation which indicated that the tank and its anchorage are seismically adequate. The sight gauge for NS-T-1 could be damaged during the SSE. The licensee stated that even if the sight l gauge breaks or the connecting pipe at the top of the tank is severed, resulting in substantial leakage, there is no significant safety concem during a safe shutdown earthquake based on a plant-specific calculation on loss of surge tank inventory.

l In a subsequent letter to the NRC from the licensee dated September 2,1997, the licensee stated that GPUN calculation number C-1101541-5360-006,"NSCCW Loss of Surge Tank inventory," establishes that sufficient not positive suction head (NPSH) is available if tank inventory is lost. It also stated that this calculation bounds the consequences of a break and loss i of function of both the sight glass and the attached small-bore pipe, and that no further specific evaluation of the appurtenances is required. This item is resolved.

8. Condensate Storage Tank l

This large tank with ring stiffeners is located outdoors. The concrete foundation was observed to l have vertical cracks at several of the anchor bolt locations. As discussed in the teleconference

! of October 2,1997, the licensee will evaluate the effects of concrete cracks in accordance with the evaluation criteria of ACI 349-3R-96 and take the necessary' corrective action. The licensee agreed to document the resolution of this finding.

! 9. Bonc Water Storage Tank (BWST)

This is a large tank located outdoors in the vicinity of the condensate storage tank. Two other long, narrow tanks separated by space frames are located close to the BWST. As discussed in the teleconference of October 2,1997, the licensee will evaluate the potential interaction between the space frame and the BWST and take corrective action, if necessary. The licensee ,

agreed to document the resolut;on.

ATTACHMENT 2

a t 3

10. Air Stari Reservoirs 1 and 2, EG T-0001 A-1 and 2 The air start reservoirs are located in the diesel generator building at an elevation of 305 feet.

These lightweight tanks are bolted to the pedestal through shims and are braced about halfway up the wall. The staff's audit team did not identify any deviations from GlP-2 and, therefore, the support anchorage of these reservoirs was found to be acceptable.

11. FeedwaterValves, FWV928 and FWV168 These two valves were determined to be seismically adequate based on Method A of GIP-2 for comparison of seismic capacity versus demand. Because of the issue raised by the staff concoming Method A, the licensee stated that a seismic analysis is being performed for the piping system, including these valves, to determine the seismic adequacy of these two valves using the in-structure response spectra approved by the staff for the TMi-1 USI A-46 program.

The licensee agreed to confirm the seismic adequacy of these two valves using Method B from Table 4-1 of GlP-2, and document the resolution.

12. River Water Pump, NR-P-1B i

l This vertical pump in located is the scan house at an elevation of 308 feet. The casing and shaft length cantilevered more than 20 feet. This item should have been classified as an

" outlier." The extended part above the base plate is braced to the nearest wall. The base plate is anchored to the pedestal and is also braced to the same wall. Some of the bolts connecting the braces to the wall were observed to be missing. The casing of the pump shaft below the floor was corroded. The staff audit team expressed concem about the corroded shaft casing and bolt connections. The licensee responded that the corrosion condition is being monitored.

Maintenance work has been initiated for some of the shaft casings in the vicinity. The cantilever shaft was analyzed and the deflection and stresses were found acceptable. This item is resolved. ,

13. Screen House Engineered Safeguards Motor Control Center, IA-480-SHES l

This 480-V MCC is located on a platform in the screen house at a floor elevation of 308 feet.

The MCC is welded to the platform steel below and is connected to an adjacent switchgear assembly with about 1/4-inch-diameter screws / bolts. The four 1/4-inch screws / bolts appeared to be hadequate and no design calculations were available on record The licensee agreed to verify the adequacy of the side attachment to ensure that it is at least as strong as that between adjacent MCC sections. The staff audit team also noted at the time of the inspection that there were four small breakers in the racked-out positions. The licensee agreed to take appropriate action regarding this observation in accordance with NRC Information Notice 97-53, dated July 18,1997.

• • r 4

14. Base Plates Located in the intake Screen and Pump House (ISPH)

The staff audit team noted corrosion in floor-mounted carbon steel base plates located in the ISPH at an elevation of 305 feet. The base plates are part of the conduit supports.

The condition was identified by the licensee's Seismic Review Team and was classified as an '

ouuser.

The corroded conditions cannot be corrected under the Plant Preservation Program, which is limited to application of coatings and minor corrective maintenance. Therefore, the licensee has i initiated Administrative Work Request #789103 to provide the repair specifications. The licensee I egreed to document the corrective action regarding this outlier.

15. Grounding Resistor, EG Y-1B-GR The grounding resistor for the emergency diesel generator is located in the diesel generator ,

building at an elevation of 305 feet. The main concem is the brittle ceramic insulator. TL. l licensee had performed a conventional structural analysis with questionable assumptions for determining the maximum stresses of the ceramic insulator due to an SSE loading. It is noted .

that GIP recommends the use of 20% of the ultimate strength as the tensile strength allowable I for cast iren. The staff audit team questioned whether there is any test data to support the use of 40% of ultimate strength as the tensile strength allowable for ceramics, which have less energy absorption capacity than cast iron. The licensee agreed to reassess the seismic adequacy of this equipment and document the resolution.

16. 480-V Enginee:ing Safeguards Valves MCC,1C-480V.ESV This equipment is located in the fuel handling building at an elevation of 281 feet. The MCC was inspected only for the spatial interaction concem relative to an adjacent conduit, which is very close (about 1/4-inch gap). In its letter to the NRC dated September 2,1997, GPUN confirmed

< the potential for interaction and has initiated Administrative Work Request #788927 to reroute the conduit in order to resolve this concem. The licensee agreed to document the resolution.

l l

l