0CAN059906, Forwards Response to NRC 990402 RAI Re GL 96-05, Periodic Verification of Design Basis Capability of Safety-Related Movs

Forwards Response to NRC 990402 RAI Re GL 96-05, Periodic Verification of Design Basis Capability of Safety-Related Movs

ML20195C021
Person / Time
Site:	Arkansas Nuclear
Issue date:	05/28/1999
From:	Vandergrift J ENTERGY OPERATIONS, INC.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
0CAN059906, CAN59906, GL-96-05, GL-96-5, TAC-M97013, TAC-M97014, NUDOCS 9906030340
Download: ML20195C021 (13)

Text

.

W 7 Entergy operations,inc.

@ 1448 S R. 333 RtzsdtA AR 72801 Td 501858-f000 i

May 28,1999 OCAN059906 U. S. Nuclear Regulatory Commission Document Control Desk Mail Station OPI-17 Washington, DC 20555

Subject:

Arkansas Nuclear One- Units 1 and 2 Docket Nos. 50-313 and 50-368 License Nos. DPR-51 and NPF-6 Response to NRC Request for Additional Information on GL %-05; Periodic Verification of MOVs (NRC TAC Nos. M97013 and M97014)

Gentlemen-On September 18,1996, the Nuclear Regulatory Commission (NRC) issued Generic Letter  !

(GL) 96-05, Periodic Venfication of Design Basis Capability of Safety-Related Motor Operated Values. This GL requested that nuclear power plant licensees establish a program, or ensure the effectiveness of their current program, to verify, on a periodic basis, that safety-related motor-operated valves (MOVs) continue to be capable of performing their safety functions within the current licensing basis of the facility.

On March 17,1997, Entergy Operations, Inc. on behalf of Arkansas Nuclear One, Units 1 and 2 (ANO-1/2), and other Entergy nuclear sites submitted a response to GL %-05 indicating our intent to implement the provisions of a joint owners group (JOG) program on MOV periodic verification. The NRC Staff encouraged licensees to participate in the industry-wide {f JOG program to provide a benefit in reactor safety by sharing expertise and information on MOV performance and to increase the efficiency of GL %-05 activities at nuclear plants. On January 11,1999, Entergy Operations Inc. submitted an updated response to GL %-05 that indicated we would continue to participate in the JOG MOV Periodic Verification Program and will implement the program elements described in Topical Report NEDC-32719, Revision 2. hl As a result of ongoing NRC review oflicensee's MOV periodic verification programs, the B NRC Staff has issued requests for additional information (RAls) for closing GL %-05 programs. In a letter dated April 2,1999, the NRC Staff requested that ANO provide additional information regarding GL 96-05 within 60 days of the date of the letter. In accordance with the RAI, ANO is hereby providing the attached response for your review.

Please contact me if you have any additional questions.

9906030340 990528 4 PDR ADOCK 05000313:.

P PDR }

1 U. S. NRC May 28,1999 OCAN059906 Page 2 i

Very tru,1y yours,

~ // ,

D andergri Director, Nuclear Safety JDV/sab Attachment cc: Mr. Ellis W. Merschoff Regional Administrator U. S. Nuclear Regulatory Commission Region IV 611 Ryan Plaza Drive, Suite 400 Arlington, TX 76011-8064 NRC Senior Resident Inspector Arkansas Nuclear One P.O. Box 310 London, AR 72847 Mr. Nick Hilton NRR Project Manager Region IV/ANO-1 U. S. Nuclear Regulatory Commission NRR Mail Stop 13-D-18 One White Flint North 11555 Rockville Pike Rockville, MD 20852 Mr. Chris Nolan NRR Project Manager Region IV/ANO-2 U. S. Nuclear Regulatory Commission NRR Mail Stop 13-D-18 One White Flint North 11555 Rockville Pike Rockville, MD 20852

l Attachment to OCAN059906 Page 1 of12

. ARKANSAS NUCLEAR ONE, UNITS 1 AND 2 RESPONSE TO NRC REQUEST FOR ADDITIONAL INFORMATION ON GENERIC LE' ITER 96-05 NRC RAI No.1:

l In NRC [ Nuclear Regulatory Commission] Inspection Report No. 50-313 and 368/% .23, the NRC Staff closed its review of the motor-operated valve (MOV) program implemented at Arkansas Nuclear One, Units 1 and 2 (ANO) in response to Generic Letter (GL) 89-10, " Safety-Related Motor-Operated Valve Testing and Surve'd lance." In the inspection report, the NRC Staff discussed certain aspects of the licensee's MOV 4 program to be addressed over the long term. For example, the inspectors noted that (1) l the licensee was still evaluating dynamic unseating loads; (2) the licensee intended to l perform modifications to increase the actuator capabilities for several valves that had less than 10 percent margin; and (3) the licensee acknowledged the need to improve its trending procedures. Describe the actions taken to address the specific long-term aspects j of the MOV program at ANO noted in the NRC inspection report.

ANO Response to RAI No.1: i The following sections provide a synopsis of how each of the three main issues in ANO Inspection Report No. 50-313 and 368/96-23 was addressed Dynamic Umscating Loads: The concern raised by NRC during the closcout inspection regarding unseating loads was that they might increase with increased dynamic pressure across the valve. The basic premise was that the unseating loads and the differential pressure loads were additive loads. While ANO does not have any multiple pressure point testing, we have tested approximately 150 MOVs of various valve types and sizes and under varying pressure conditions. Dynamic unseating load data was compared to corresponding static unseating data. Globe valves and hard-seated butterfly valves exhibited little or no unseating load under either static or dynamic conditions. A population of 70 solid and flex-wedge gate valves did provide enough information to indicate that the mechanical unwedging load is a separate and non-additive load to the differential pressure load seen under dynamic conditions. There were indications of differences between individual static and dynamic unseating loads for particular valves, but many ANO valves exhibit similar variances between two successive periodic static tests.

MOV Modifications: ANO has modified many of the safety-related MOVs having reduced or non-conservative margins since Generic Letter 89-10 closure. New issues, such as Limitorque Technical Update 98-01, have resulted in additional MOVs with less than 10% margin. MOVs with less than the required margin (0% based on design criteria inputs) are addressed by condition reports having operability evaluations and corrective actions accordingly issued. MOVs with less than 10% margin are tested on a more

Attachment to OCAN059906 Page 2 of12 frequent basis as required by Entergy Operations, Inc. (EOI) Nuclear Management Manual DC-Il0, Periodic Venfcation, and Rev. 2 to the joint owners group (JOG) report on periodic verification. Actions to improve design margins for these MOVs are being considered by either modification or design basis reviews The amount of available margin and the cost ofincreased testing primarily determine the need to modify the valve.

MOV Treading: ANO has made improvements to the MOV trending program based on berd.w.eAleg efforts at other sites. Currently all testing data is input into an ANO MOV program database where the data can be more easily extracted for review and trended, as required. Standard trending packages have been developed which include all relevant trend data. With the use of the database all data points can be trended on a case-by-case basis should the standard trend package indicate a negative trend. Records of testing anomalies have also been added to the trending database so that they can be tracked.

MOV test trending is an evolving area within the industry and changes to the program may be required as more information becomes available. When changes are required, they will be handled as a function of overall MOV program activities.

In addition to the above, another area for improvement included enhancements to maintenance procedures for skill of the craft activities. ANO Procedure 1025.011, Motor Operated Value Maintemmce Program, was enhanced to more positively control internal l valve characteristics. l l

1 l

i

At*-hawat to OCAN059906 Page 3 cf12 NRC RAI No. 2:

In a letter dated March 17, 1997, the licensee stated that it is participating in the joint owners group (JOG) program on MOV periodic veri 6 cation in response to GL %-05. It is recognized that the owners group for ANO Unit 1 (a Babcock & Wilcox designed reactor) did not pa:ticipate in the JOG program. On August 6,1997, the Combustion Engineering (CE) Owners Group submitted Revision 2 of Topical Report MPR-1807 on the JOG program on MOV periodic verification. On October 30,1997, the NRC Staff completed a safety evaluation concluding that the JOG program is an acceptable industry-wide response to GL 96-05, with certain conditions and limitations.

The JOG program specifies that the methodology and discrimination criteria for ranking MOVs according to their safety significance are the responsibility of each participating licensee In its letter dated March 17, 1997, the licensee stated that static diagnastic testing would be based, in paat, on an expert review and ANO's Probabilistic Safety Assessment (PSA). Describe the methodologies used for risk ranking MOVs at ANO in detail, iachWag the specific high-risk MOVs at each ANO unit. In responding to this request, the licensee might apply insights from the guidance provided in the Westinghouse Owners Group (WOG) Engineering Report V-EC-1658-A (Revision 2, dated August 13, 1998), " Risk Ranking Approach for Motor-Operated Valves in Response to Generic Letter 96-05," and the NRC safety evaluation dated April 14, 1998, on the WOG ,

methodology for risk ranking MOVs at West inghan- designed reactors. The licensee i could also cbtain insights from an MOV risk-ranking methodology developed by the Boiling Water Reactor Owners Group.

.tNO Response to RAI No. 2:

A detailed discussion of the ANO MOV risk ranking methodology and its application is provided below. The overall approach is depicted in the flowchart shown in Figure 1.

The five primary steps employed in this methodology are summarized below:

Sten 1: Irlanti&mtina of GL 89-10 MOVs and Func+ional Faihire Modes In this first step, the ANO Program MOVs for each unit and their corresponding functional failure modes (FFM) were identified. The term " functional failure mode" refers to the combination of safety function (e.g., isolate containment, allow ECCS flow to the RCS, etc.) and component failure mode (e.g., failure to close, failure to open, etc.).

Step 2: PSA-Based MOV Safety Importance Analysis The safety significance ranking of the GL 89-10 MOV failure modes was primarily based on the use of existing plant-specific PSA event tree / fault tree model and cutset results.

The risk achievement worth (RAW) of all failure modes of each MOV was used as the

At*=ch=amit to OCAN059906 Page 4 cf12 primary PSA safety importance measure. The major steps in the PSA-based safety )

importance analysis process are presented below.

Substep 2a: Identifv PSA Model and Results The safety importance of GL 89-10 MOV failures in the PSA model was determined using the PSA Rev. O at-power core damage frequency (CDF) cutset results. ,

Although the general methodology allows for explicit accounting of other risk '

measures, such as large early release frequency (LERF), external events, and shutdown risk, in the actual analysis engineering judgment was used to identify the risk importance of all MOVs in Step 2e (Risk insights) and Step 4 (Expert Panel).

Substen 2b: Identify MOV FFMs in PSA Model A list of the MOV FFMs in the PSA model was developed by reviewing the basic events in the PSA model. The MOV failure modes represented in the model *mclude failure to open, failure to close, failure to operate, transfers open, and transfers closed.

Consistent with GL 89-10, only active MOV FFMs (i.e., failure to open and failure to close) were tabulated.

I MOV failure events in the PSA model were grouped according to function. The grouping of similar MOVs helps to eliminate the effect of simplifying PSA model i asymmetries and assumptions. As a conservative measure, all of the MOVs in a group are placed in the category associated with the highest importance MOV.

Substen 2c: Enrich the PSA Model Results with MOV Failures The number of cutsets containing MOV active failures was increased by artificially increasing the active MOV failure rates in the PSA model to a bounding high value (e.g.,0.1/ demand) and then re-quantifying the PSA model. Cutsets which contained active MOV failures that were not represented in the Rev. O cutset fde were added to the cutset file. The MOV failure rates in the resulting cutset (hereafter called the

" reference" cutset) were restored to their nominal value and operator recoveries were applied to these cutsets. This process " enriched" the number of MOV failure events, which occur in the model cutset results. This method helped to assure that the Level-1 cutsets contained more relatively low safety-importance active MOV failures.

1 Substen 2d: Perform the PSA Safety Importance Screeninn Analysis 1

The PSA safety importance evaluation was performed on those MOV FFMs that occurred in the reference cutset results. Several risk importance measures were considered for ranking the safety significance of MOVs. The RAW importance measure was selected as the most appropriate measure for ranking MOV FFM safety importance This choice was based on the fact that a decision to defer testing on an MOV may result in not detecting its potential failure and may result in an increase in

Attachment ts OCAN059906 Page 5 of12 risk. Since RAW is a measure of the increase in risk associated with failing a component, it was chosen as the most appropriate safety importance measure.

The RAW of a basic event (BEi) from the PSA is defined u follows:

RiskAchievenent Wwth (RA W,) = '" where, CDF(REF)

CDMBEl = 1) = Perturbed core damage frequency, (i.e., CDF given the probability of basic event BElis set to 1.0 (or TRUE)) and CDMREF) = Reference core damage frequency.

For the Level-1 MOV safety importance analysis, the RAW associated with each MOV was determined via cutset manipulations on the reference cutset using a cutset editor. The RAW value was based on the perturbed CDF in which all of the MOVs FFMs are concurrently set to TRUE.

The flow chart shown in Figure 2 provides an overview of the PSA-based MOV safety importance ranking analysis process. The process involves two MOV " screenings."

These screenings rank MOVs included in the PSA model into one of three categories.

These categories are as follows:

(1) MOVs whose individual RAW values are greater than or equal to 2.0 when assigned a failure probability of unity are categorized as "High" PSA safety importance MOVs; (2) MOV groups which are not "High" whose cumulative RAW is less than or equal to 2.0 when assigned a failure probability of unity are categorized as " Low" PSA safety importance MOVs; (3) MOVs that are neither "High" nor " Low" are classified as " Medium" PSA safety importance MOVs.

The first screening, " Screen 1" in the figure, identifies the "High" PSA safety importance MOVs: each MOV in this category has a RAW of 2.0 or greater. The first step in this process was to group the MOV failures according to function (e.g., HPSI injection MOVs). The Level-1 ranking analysis consisted of setting the active failure mode events (all FFMs concurrently) for each MOV (one MOV at a time) to TRUE in the combined reference cutset anc! subsuming the cutset to eliminate non-minimal cutsets. If the reference CDF increased by a factor of two or more (i.e., RAW 2 2.0),

then the MOV was considered a "High" safety importance MOV. Note that the probability of the common cause failure event associated with each MOV was set to its beta factor during this process. To eliminate asymmetric modeling assumptions, all of the MOVs in a group were ranked according to the most important MOV.

E Attachment to  !

OCAN059906 i Page 6 of12 The next step in the screening process, teferred to as " Screen 2" in Figure 2, identified the " Low" PSA safety importance MOVs. In this screening process, the MOV functional groups were treated as single entities in order to eliminate asymmetry issues l associated with the PSA model In this process, all the valves in the least safety imponant MOV functional group were set to TRUE (i.e., all valves in the group were I assumed to simultaneously fail). If the resulting CDF was less than twice the reference  !

CDF, the process was repeated with both the lowest and second-lowest importance l MOV functional groups set to TRUE (i.e., both groups of valves simultaneously l failed). This process continued until a maximum number of MOVs which when set to l TRUE (i.e., all simultaneously failed) result in a penurbed CDF which was just less

]

than twice the reference CDF. The final selection of MOV groups has an overall  !

RAW of just under 2. This set of MOVs was classified as " Low" PSA safety importance MOVs.

1 The remaining MOVs not directly classified as either "High" or " Low" are classified as

" Medium" PSA safety important MOVs.

Substen 2e: Perform Focused PSAs and Develop PSA-Related Insinhts Results of the Level-1 at-power PSA model provided only a limited assessment of an MOV's risk importance Therefore, an additional review was performed to qualitatively incorporate the Level-1 risk issues associated with the ISLOCA, ATWS, j and Internal Flooding analyses submitted as part of the ANO-1 and ANO-2 IPEs and l to qualitatively incorporate the Level-2 results of the IPEs. These risk importance insights were used to increase the PSA-based MOV risk rankings wheie judged appropriate. These insights were presented to the expert panel.

Step 3: Expert PanelImoortance Assessment An expert panel, composed of a broad spectrum of ANO expertise (i.e., Operations, Maintenance, Systems Engineering, MOV Testing, MOV Design Engineering, and Nuclear Fagia-ing Design) was established to assure that the final MOV safety importance rankings were reasonable. The charter of this expert panel was to establish overall safety importance rankings for the GL 89-10 MOVs. Using the PSA-based MOV risk ranking results as att input, the members of this expert panel combined their operating experience and engineering judgment to establish an overall safety importance for each MOV. This panel assured that the MOV risk ranking accounted for the functional failure modes defined for the GL 89-10 MOVs and assured that significant safety and operational concerns in all plant operating conditions were addressed. The expen panel ranking represented the " final" and overall MOV imponance These results were used to establish the periodic testing schedules for the ANO-1 and ANO-2 MOVs. The high-risk importance MOVs for ANO-1 and ANO-2 are provided in Tables I and 2, respectively.

The MOV safety importance rankings provide information as to the consequence of an MOV failure.

Attachment to I OCAN059906 '

Page 7 of12 l l

Step 4: MOV Setun Marain Analyaig ANO utilizes setup margin as a measure of MOV failure probability. MOVs are periodically tested to ensure continued safe operation. As-leR static test results are i combined with the minimum required operating thrust or torque to determine the actual setup margin for the MOV. Margins are calculated in both the open and closed directions for both thrust and torque (as applicable) and the margin associated with the safety function of the valve becomes the actual setup margin.

1 Step 5: MOV Test Schedule Development The MOV test schedules for both units are based on the overall risk associated with failure i to operate during an accident condition. " Risk" is a combination of" probability" and

" consequence." The ANO MOV test schedule was developed by combining the MOV's I setup margin (a measure of" failure probability") and its safety importance (a measure of

" consequence").  !

The process used at ANO combines the three safety importance rankings (High, Medium, l and Low) and four setup margin categories (High-High, High, Medium, and Low) to place each of the GL 89-10 MOVs into one of four MOV testing frequency categories.

This process assures that the most safety important MOVs having the lowest setup margin will be tested most frequently. Likewise, MOVs with low safety importance and high setup margin will be tested least frequently.

Attehment to OCAN059906 Page 8 of12 Figure 1 Overall Process to Establish the GL 89-10 MOV Testing Schedule STEP 1

• I Ideanly OL 8910 OL 8910 MOVs MOVs and FTMs and Funcenal Failure Modes 1F For Each FFM:

Y v

Tressed in PSA Ya PSA Based STEP 2 i PSA. Based Safetyimportance

^"*'Y"' w v

Yes PSA. Based Focused PSA7  : --*

Impanance Assessment

--____ ____ ____ ___ _m______ ____ ___ ________ _ _ _ ..___

STEP 3 Expert Panel Expert PanelImportance ,

Impanano. A- Priority A-1 i

TFEP 4

MOV Margin Assessment  :

MOV haarsin Analysis

____________________3___________

MOV Testing hia Assessment STEP 5 MOV Test Sabedule

(__ _.

MOV Teenns Sched ie i

l

Attachment to i

. OCAN059906 Page 9 of12 l

Figure 2 PSA Based Safety Importance Ranking Process OL 39-10 MOV FFM or l Assive MOV FFM in PSA Model {

1 v

MOVs w/ j Cumulative i Yes RAW <2 FFM

PSA Screen 1 PSA Screm 2 Tressed in PSA Model?

No MOVs w/ Other y

RAM 2 MOVs u.n-Focuesd PSA No categorized or PSA-Related PSA Safety-lesights? Importance MOV Yes o

MOV Casecones bened on Medium Focuesd PSA and/or  :

PSA-Related High tow 0

i u r PSA Safety- PSA Safety- PSA Safety-importance importance importance MOV MoV -

MoV

Attachment ts

. OCAN059906 Page 10 of12 Table 1 ANO-1 High Risk Importance MOVs VALVE' SYSTEM FUNCTION CV1276 MU LPI DISCH TO HPI CV1277 MU LPI DISCH TO HPI CV1405 DH RB SUMP ISOL CV1406 DH RB SUMP ISOL.

CV1407 DH BWST OUTLET CV1408 DH BWST OLTILET CV1428 DH DHR COOLER OUTLET CV1429 DH DHR COOLER OUTLET CV2215 ICW ICW RTN ISOL.

CV2400 BS RB SPRAY ISOL. l CV2401 BS RB SPRAYISOL CV2625 FW MFWISOL CV2675 FW MFWISOL CV3643 SW ACW ISOLATION CV3806 SW EDG HX SW ISOL CV3807 SW EDG HX SWISOL CV3811 SW ICW LOOP 11 SW ISOL l CV3820 SW ICW/ LOOP I SW ISOL.

Table 2 ANO-2 High Risk Importance MOVs VALVE SYSTEM FUNCTION 2CV02052 MS EFW PUMP STEAM BYP.

, 2CV03402 MS EFW MS ISOL 2CV10251 EFW EFW TO SGA 2CV10751 EFW EFW'ID SGB 2CV14531 SW SDC SWISOL 2CV14562 SW SDC SW ISOL 2CV15031 SW EDG COOLERISOL 2CV15042 SW EDG COOLERISOL 2CV46981 RCS ECCS VENT 2CV47402 RCS ECCS/LTOP VENT 2CV50151 HPSI HPSIINECTION 2CV50162 HPSI HPSIINJECTION 2CV50351 HPSI HPSIINECTION 2CV50362 HPSI HPSIINECTION 2CV50551 HPSI HPSIINECTION  ;

2CV50562 HPSI HPSIINJECTION l 2CV50751 HPSI HPSIINECTION 2CV50762 HPSI HPSIINECTION l 2CV56121 BS RB SPRAY HDR ISOL. j 2CV56132 BS RB SPRAY HDR ISOL.

2CV56491 BS RB SUMP SPRAY SUCT 2CV56502 BS RB SUMP SPRAY SUCT i

'. Attachment to

.- OCAN059906 Page 11 of12 NRC RAI No. 3:

The JOG program focuses on the potential age-related increase in the thrust or torque required to operate valves under their design-basis conditions. In the NRC safety evaluation dated October 30,1997, on the JOG program, the NRC Staff specified that licensees are responsible for addressing the thrust or torque delivered by the MOV motor actuator and its potential degradation. Describe the plan at ANO for ensuring adequate ac and de MOV motor actuator output capability, including consideration of recent guidance in Limitorque Technical Update 98-01 and its Supplement 1.

ANO Response to RAI No.3:

On July 22,1998, ANO issued a condition report in response to Limitorque Technical Update (LTU) 98-01. ANO Program MOVs with limitorque actuators, AC motors, and rising stems were reviewed in accordance with new limitorque motor pullout torque (MPT) equation guidelines. Of the 183 MOVs reviewed,47 were found to have as-left settings outside of the newly defined maximum MPT. Operability for the 47 MOVs was established based on safety function of the valve, actual available voltage, and in some cases actuator capability to seal the valve for a one-time stroke.

Sixteen action items were issued as part of the condition report. Nine involved revising design basis documents to reflect a more accurate depiction of anticipated field voltages and pressures during design basis accident events. Two of the actions call for torque switch resets for a group of MOVs such that the new setting will be within the allowable torque window. Four of the actions call for the modification of specific MOVs. One of the corrective actions involved reviewing the quarter-turn valve population for impacts related to LTU 98-01.

Although LTU 98-01 explicitly states that it is only applicable to rising stem valves, ANO took the position that the issue was generic and applicable to quarter-turn MOVs, especially in cases where the MOV is torque switch controlled. The as-left settings of the 41 quarter-turn program MOVs were reviewed. Nine were found to have torque switches set higher than the allowable MPT torque established by LTU 98-01. ANO typically utilized 80% voltage as input into the setpoint calculations. Utilization of full under-voltage values in the MPT equations resulted in torque capabilities that exceeded the as-left settings for six of the nine MOVs. Operability for the remaining three MOVs was established and a corrective action issued to reset these MOVs to a lower torque switch setting.

In addition, ANO is cognizant of the issue regarding DC motor capability and has initiated a review of the electrical system and battery sizing, as well as the identification of potential modifications to the DC motor operated actuators. In general, ANO DC motor operated actuators have greater than 20% torque capability in excess of current as-left torque I i settings. Only one DC MOV has a torque margin ofless than 20% but it also has a thrust

'. Attehawnt to OCAN059906 Page 12 of12 setup margin of greater than 100% (thrust setting is more than twice the minimum I required thrust). Therefore, resetting the torque switch to a lower value is a viable option Tor this MOV. All potentially impacted ANO MOVs are operable based on current understanding of the issue.

i

)

1 1

1 1

1

!