ML18058A131

Ti 2515-195 Response to Flowserve Updated Part 21 Notification Issued on July 11, 2017

ML18058A131
Person / Time
Issue date:	07/11/2017
From:	Michael Farnan NRC/NRR/DE/EMIB
To:
Farnan, M
Shared Package
ML18058A128	List: ML18058A131
References
DC 18-005
Download: ML18058A131 (23)

Text

NRC INSPECTION MANUAL EMIB TEMPORARY INSTRUCTION 2515/195 RESPONSE TO FLOWSERVE UPDATED PART 21 NOTIFICATION ISSUED ON JULY 11, 2017 CORNERSTONE: MITIGATING SYSTEMS APPLICABILITY: This TI applies to all operating nuclear power reactor licensees that have Anchor/Darling Double Disc Gate Valves (A/D DDGVs) with threaded stem to upper wedge and a locking wedge pin design installed in safety-related systems.

2515/195-01 OBJECTIVES The objective of this TI is to support the U.S. Nuclear Regulatory Commission (NRC) review of licensees activities in response to Flowserves updated Part 21 notification, Stem-Wedge Separation of an Anchor/Darling Double Disc Gate Valve at Exelon, LaSalle County Station, Unit 2, July 2017 (ADAMS Accession No. ML17199F890). Inspection activities implemented under this TI will allow the staff to independently verify that safety-related A/D DDGVs are operationally ready to perform their safety function and that licensees are implementing adequate corrective actions.

2515/195-02 BACKGROUND On January 4, 2013, Tennessee Valley Authority (TVA) issued a 10 CFR Part 21 notification letter to the NRC subject titled Anti-Rotation Pin Failure in (Flowserve) Anchor/Darling Double Disc Gate Valve (ADAMS Accession No. ML13008A186). The notification referenced a then recent failure of a Browns Ferry Nuclear Plant (BFN) Unit 1 High Pressure Coolant Injection (HPCI) Inboard Steam Isolation Valve, which significantly exceeded its leakage rate administrative limit. Disassembly and inspection revealed that the valve stem to wedge anti-rotation wedge pin had broken and the disc retainer had fallen from the wedge assembly causing an obstruction to the valve discs. BFN had previous wedge pin failures on identical A/D DDGVs dating back to 2001 when BFN began replacing their HPCI system steam isolation valves to improve local leak rate performance.

On February 25, 2013, Flowserve issued a 10 CFR Part 21 notification, Wedge Pin Failure of an Anchor/Darling Double Disc Gate Valve at Browns Ferry Nuclear Plant Unit 1 (ADAMS Accession No. ML13064A012). The notification concluded that the root cause of the wedge pin failure at BFN Unit 1 was excessive load on the wedge pin. The stem operating torque exceeded the torque used to tighten the stem into the upper wedge before installation of the wedge pin, ultimately causing the wedge pin to shear. Flowserve stated that this situation could potentially occur in any A/D DDGV with a threaded stem to upper wedge connection (typically size 2.5 inch and larger) operated by an actuator that applies torque on the stem to produce the required valve operating force. This would typically include most A/D DDGVs in motor operated valve (MOV) applications that utilize Limitorque or Rotork actuators. Flowserve recommended that all critical A/D DDGVs with threaded stem to upper wedge connections and actuators that Issue Date: DRAFT 1 2515/195

produce torque on the stem be evaluated for potential wedge pin failure. Valves with electric motor actuators which produce high output torques would be the most susceptible to failure.

Valves that were assembled with stem torques that exceed the operating torque are not candidates for failure. Flowserve recommended that valves with operating torques greater than the stem installation torque be corrected by increasing the stem installation torque to at least the maximum expected operating stem torque. Flowserve further stated that there was no non-intrusive test or inspection method to determine whether the stems were adequately torqued into the upper wedge prior to pin installation. However, abnormal rotation of the stem immediately after valve seating or unseating, short valve stroke, or poor, unusual operation would be symptoms of a degraded or failed stem-disc connection and should be investigated.

On February 11, 2017, during a refueling outage at LaSalle County Station Unit 2, the licensee was attempting to fill and vent the High Pressure Core Spray (HPCS) system when the HPCS pump discharge valve (12 A/D DDGV) experienced a stem-disc separation. Disassembly and inspection revealed the anti-rotation pin had sheared and there was severe degradation on the valve stem and upper wedge threads. The licensee had been using industry guidance to perform visual examinations and diagnostic testing on the valve in response to operating experience and the 10 CFR Part 21 notifications issued on January 4, 2013, and February 25, 2013.

On May 31, 2017, in response to the 10 CFR Part 21 notifications and the recent LaSalle A/D DDGV failure, Columbia scheduled an as-found diagnostic test and disassembly/rework task for the HPCS injection valve. This valve is identical to the LaSalle model which failed on February 11, 2017. The as-found diagnostic test produced abnormal results. Disassembly and inspection revealed the upper wedge anti-rotation pin had sheared allowing the stem to rotate freely in the upper wedge. However, the stem and wedge were still connected with minor wear noted on the threads.

On June 15, 2017, the NRC staff issued IN 2017-03, ANCHOR/DARLING DOUBLE DISC GATE VALVE WEDGE PIN AND STEM-DISC SEPARATION FAILURES, (ADAMS Accession No. ML17153A053) to inform stakeholders of recent failures involving A/D DDGVs.

On July 11, 2017, Flowserve issued a 10 CFR Part 21 notification, Stem-Wedge Separation of an Anchor/Darling Double Disc Gate Valve at Exelon, LaSalle County Station, Unit 2, February 2017 (ADAMS Accession No. ML17194A825). This notification was an update to the Part 21 notice issued on February 25, 2013. The initial Part 21 notice concluded a wedge pin shear could lead to stem-wedge joint degradation and eventual stem-wedge failure. The updated Part 21 notice concluded that repeated valve cycles at high actuator loads can shear the wedge pin and eventually wear the wedge threads to the point of failure during a valve closing cycle resulting in a subsequent separation of the stem from the wedge. In addition, the updated 10 CFR Part 21 noted that the pressed fit collar design can be pushed up and out of position, thus reducing or eliminating any existing preload on the stem to wedge joint. Once the preload is lost, the wedge pin would be susceptible to shear from the actuator. The updated 10 CFR Part 21 notes that the capability to maintain the stem to wedge preload is much less for stems with pressed fit collars than for stems with integral collars, but that the scope remains the same as the previous notification. The Flowserve Part 21 update includes the following additional recommendations:

1. Torque the stem into the wedge to the maximum joint capacity

2. Replacement stems should have integral collars in lieu of pressed fit collars

3. Replacement wedge pins should be manufactured from high strength material Issue Date: DRAFT 2 2515/195

4. Verify the actuator stem thrust is less than the maximum allowed to maintain the stem preload (i.e., applied thrust less than the collar to stem connection thrust capability)

The following is a typical illustration of the A/D DDGV design:

Valve Description The primary valve components are the body, seats, discs, upper wedge, lower wedge, stem, and in some cases a stem disk retainer pin and retainer clips. The upper wedge holds the valve discs and fits inside the lower wedge. The stem is screwed into the upper wedge and secured with the anti-rotation pin, thus forming the disc/wedge assembly. In some valve designs, disc retainer clips are used to hold the discs together. The valve body accepts the stem wedge assembly which lowers into the body. The valve seals against liquid flow through contact between the discs and the seats.

The following diagrams show a simplified close and open operation:

Issue Date: DRAFT 3 2515/195

Issue Date: DRAFT 4 2515/195 The stem is attached to the disk/wedge assembly by one of three methods:

• Threaded two-piece stem with a pressed (interference) fit collar. The stem is screwed into the upper wedge to a recommended pre-torque against the interference fit collar. It is intended this torque be greater than the maximum operating torque of the actuator. This collar is subject to displacement if maximum thrust exceeds the collar capacity. This results in a loss of pre-torque.

• Threaded one-piece stem with an integral collar. The stem is screwed into the upper wedge to a recommended pre-torque against an integral collar. It is intended this torque be greater than the maximum operating torque of the actuator. This collar is not subject to displacement.

• T-Head configuration. This design is not subject to the failure modes discussed above and falls outside the scope of the Part 21 notification.

The following is a simple diagram of T-Head design:

The Flowserve Part 21 notification is applicable only to A/D DDGVs that have a threaded and pinned stem-to-stem wedge connection. The pin was designed to act as a lock washer to maintain the proper torque of the stem-to-stem wedge connection. The pin was not intended to be exposed to the full output of the valve actuator, so the pin was not initially evaluated for its shear strength capability Issue Date: DRAFT 5 2515/195

While the Flowserve Part 21 recommended repair and replacement, industry representatives (specifically the Boiling Water Reactor Owners Group (BWROG)) developed a guidance document that provides a graded approach to addressing the A/D DDGV issue. The guidance provides recommended prioritization and screening criteria, evaluation methods, inspection, use of diagnostics, repair methods, and recommended repair schedule.

The NRC staff was provided a copy of Revision 4 of the guidance and noted three provisions that could lead to non-conservative results, depending on their application. Specifically, they are (1) the use of stem-to-wedge thread friction, (2) the calculation of maximum actuator torque, and (3) the (potential) overreliance on testing and diagnostics. The NRC Special Inspection Team (SIT) team that reviewed the LaSalle Unit 2 MOV failure concluded that the use of diagnostics and stem rotation checks were not effective in determining active stem-to-wedge connection degradation. However, diagnostics and stem rotation checks are effective in determining gross failure of the stem-to-wedge connection. The BWROG program uses the stem rotation tests and diagnostics to establish near-term operability, sufficient to support one cycle of continued operation, in a structured repair program. This is intended to allow licensees to prioritize and focus efforts on the valves with the highest safety significance. The stem rotation checks and diagnostics should not be used to conclude that the stem-to-wedge joint is undamaged (i.e., concluding that there have been no historic over torque events that challenged the threaded joint). These three provisions are discussed in relevant portions of this TI.

Finally, as part of a voluntary initiative, all licensees submitted information on their safety-related A/D DDGVs. The licensees submittals followed a template developed by the Nuclear Energy Institute and were used, in part, to identify the overall inspection population and to determine whether further NRC staff actions are warranted. NRC staff will use the results of inspections conducted under this TI to determine whether further regulatory action is necessary.

2515/195-03 INSPECTION REQUIREMENTS This will be a one-time inspection with the goal to verify that licensees have: (1) identified the valve population at their sites affected by Flowserve Part 21 notification on A/D DDGVs and reflected those valves in their submittals to the NRC; (2) properly evaluated whether valves are susceptible to possible failure; and (3) instituted an acceptable corrective action plan. The inspection will be a three step approach:

1. Identify - The Flowserve Part 21 notice states that a valve failure can potentially occur on any A/D DDGV with a threaded stem to upper wedge connection (typically size 2 inch and larger) operated by an actuator that applies torque on the stem to produce the required valve operating force. This first step is to identify valves that meet the criteria.

2. Evaluate - Evaluate the valve population identified in step 1 for potential susceptibility to wedge pin and stem-disc connection failure to determine whether the licensee is appropriately determining whether valves require further corrective actions. This evaluation can be performed on a sampling basis.

3. Review - Review the licensees corrective action plan.

Issue Date: DRAFT 6 2515/195

2515/195-04 INSPECTION GUIDANCE 04.01 Identify Review the licensees evaluation of the population of safety-related A/D DDGVs at their facilities. For each identified valve (or a sample), verify that the licensee has properly identified the following:

1) Valve size

2) Motor and Actuator type and size

3) Associated plant system

4) Valve function

5) Normal valve position

6) Valve operational requirements (e.g. open, close, or both, cycles)

7) Stem/Disc connection type (e.g. threaded or T-head)

8) Stem collar type (e.g., pressed fit or integral)

9) Safety risk category (high, medium, or low)*

10) Previous test and/or repair results

• - All plants with the exception of Callaway have committed to the Joint Owners Group (JOG) final program plan when they addressed generic letter (GL) 96-05. The JOG final program plan incorporates diagnostic testing of MOVs at an interval based on risk and margin.

Typical documentation to assist in the verification process include:

1) Generic Letter (GL) 96-05 program listing of applicable MOVs

2) Plant Inservice Testing (IST) Program document

3) Documentation from Flowserve and/or A/D

4) Valve drawings

5) Plant equipment master listing

6) Maintenance history

7) Testing history

8) Valve setup documentation (thrust, torque, limit switches, torque switch, etc.)

9) Valve Weak Link Analyses

10) System diagram Compare the above information with the licensees submittal to the NRC to verify accuracy of the submittal. Note that there was some ambiguity as to whether licensees should include valves with a T-head joint configuration. T-head joints are not the focus of the Part 21 corrective actions and licensees did not need to include them in their submittals.

04.02 Evaluate Note - See Attachment 1, Assessment of Anchor/Darling Double Disc Gate Valves flow chart.

As a minimum, the inspector shall select a sample of the A/D DDGV population for evaluation at the applicable nuclear power plant that includes both (1) valves determined by the licensee to be susceptible to stem-disc connection degradation, and (2) valves determined by the licensee to be non-susceptible to stem-disc connection degradation. The sample size may increase dependent on the findings. The purpose of sampling non-susceptible valves is to verify that the valves were correctly identified as being non-susceptible. Suggested sample size minimum for Issue Date: DRAFT 7 2515/195

susceptible and non-susceptible valves is the greater of three valves or 30 percent of each population.

The Flowserve Part 21 notification is applicable to only those A/D DDGVs that have a threaded and pinned stem-to-stem wedge connection. The pin was designed to act as a lock washer to maintain the proper torque of the stem-to-stem wedge connection. The pin was not intended to be exposed to the full output of the associated actuator which resulted in the pin never being evaluated for its shear strength capability.

Following Attachment 1 flow chart, determine if the valve stem has been adequately torqued into the upper wedge assembly. There is no test or inspection method to determine if the stem has been completely torqued into the upper wedge prior to pin installation. Verification would have to be validated by Flowserve documentation and/or valve maintenance history. If documentation exists, verify that stem installation torque is greater than the maximum expected operating stem torque. The maximum expected operating torque is the maximum load torque that has ever been or could be applied to the valve stem. The absolute maximum load torque applied would result from a motor stall event. To calculate the actuator capability in a motor stall event use the following equations:

AC Motor Tqstall = Tqest Tq X OAR X Mstall eff X (Vact / Vrate)2 DC Motor Tqstall = Tqest Tq X OAR X Mstall eff X (Vact / Vrate)

Tqstall = Total estimated actuator stall torque Tqest Tq = Estimated motor stall torque. If using a motor curve, the motor stall torque is the motor torque at the 0 RPM point. If the curve has a knee with a torque above that at 0 RPM, the higher knee value should be used in the calculation. The motor curve value is generic. The motor curves were developed based on a small sample size.

Due to manufacturing variances, each motor will have slightly different speed versus torque curve. If no motor curve data is available, the estimated motor stall torque is 110% of the rated start torque.

OAR = Overall actuator ratio. This is the gear ratio between the electric motor and the output of the actuator.

Mstall eff = Motor stall efficiency. For Limitorque actuators a stall efficiency is provided in SEL-7 and SEL-8 (attachments 2 and 3). The stall efficiencies are not true efficiencies.

They have been increased above the stated efficiency to envelope inertial effects. If the motor stall efficiency value is not available, use the running efficiency of the actuator.

Vact = Actual Voltage. The actual voltage is the maximum voltage that might be applied to the motor during plant operations or testing during outages.

Vrate = Rated Voltage Issue Date: DRAFT 8 2515/195

Considerations when evaluating the maximum load torque applied to the valves include:

The BWROG guidance document is not clear on what constitutes actuator maximum load torque capability when assessing the wedge pin shear capability. During public meetings, licensees and Flowserve indicated that many end users are evaluating the maximum actuator load torque applied based on historical values obtained from diagnostic testing. By using test data and not the maximum motor stall torque, a valve may be susceptible to failure based on historical over thrust/torque events. Historical over-thrust/torque events can include:

1) An over-thrust/torque event could have occurred during testing before issues were identified with A/D DDGVs or the Part 21s were issued.

2) Valves could have had excessive force and torque applied during GL 89-10 testing in the late 80s and early 90s. The bulk of the testing was completed using equipment that was not as accurate as currently available test equipment. The inaccuracy of the equipment was noted in information notices IN 1991-61, IN 1992-23, and GL 89-10 Supplement 5.

3) Some valves might have been subjected to a pressure locking or thermal binding event.

Licensees were requested to review these phenomena in GL 95-07. These types of events can cause an actuator to exert significant thrust and torque on a valve stem without the knowledge of the plant operators by the motor applying its full capability to open the valve against an unexpected pressure locking or thermal binding condition.

4) A motor stall event can be caused by a sticky contactor such as in the motor control center (MCC). When the actuator torque switch circuit opens during valve travel, the open circuit drops out the contactor relay located at the MCC. If the contactor sticks in the energized position, the motor will continue to run eventually causing the motor to stall. This type of event can go unnoticed due to the erratic behavior of a sticky mechanical device.

Some licensees are determining Applied Wedge Pin Torque based on the limiting value of the spring pack capability or the valve/actuator weak link analysis of maximum acceptable torque.

However, in a motor stall event, the actuator output torque (and the corresponding valve stem force) is not limited by the spring pack limiter sleeve. The primary purpose of the limiter sleeve is to prevent over travel of the torque switch assembly.

If valve history and maintenance records validate that the valve has never experienced a motor stall event, then the historical maximum thrust and torque applied would be actual test data.

The test data should account for all possible uncertainties such as force sensor accuracy, data acquisition accuracy, changes in stem factor, changes in actual voltage at the valve, and possible pressure locking / thermal binding events. Note that changes in voltage at the valve can occur during refueling outages due to the MCC being unloaded and/or battery bank DC chargers being at elevated values. As noted in the actuator capability torque calculation, elevated voltages will raise motor torque capability.

If the licensee is applying the maximum torque based on test data and is not using the maximum actuator capability in a motor stall event to determine valve susceptibility to potential stem-disc degradation, evaluate how the licensee plans to manage future possible motor stall events. Considerations include:

1. Wedge pin shear capability has been included in the valve weak link analysis

2. Procedures have been updated as necessary

3. Valve has been modified to lower actuator capability and/or valve components replaced with stronger materials to withstand a motor stall event Issue Date: DRAFT 9 2515/195

4. Valve drawings have been updated with Part 21 information regarding pressed fit collars on the stem and/or updated to reflect the pressed fit collar stem being replaced with an integral collar stem.

5. Preventive Maintenance (PM) activity has been updated to monitor and assess MCC contactor mechanical performance and/or have a periodic contactor replacement strategy.

If the maximum torque applied is less than the validated threaded stem-to-stem wedge connection torque, the pin will not be subjected to the actuator torque and, thus, will not fail.

This assumes that the applied stem-to-stem wedge connection torque does not exceed the stem preload torque specified in the Flowserve Part 21 notice dated February 25, 2013, such that the integrity of the stem and upper wedge threads are not adversely affected. Valves that satisfy these criteria can be considered non-susceptible to degradation of the stem-disc connection.

After the maximum applied torque has been determined, evaluate the stem and upper wedge thread shear strength to withstand applied thrust. If thrust and torque test data are not available, thrust is calculated by:

Thrust = Tq SF Tq = Torque = Calculated motor stall torque and/or the as tested maximum torque value.

SF = Stem Factor To calculate Stem Factor for Standard ACME Screw Thread Type with a half thread angle of 14.5°:

SF = d (0.96815 tan + µ)

24 (0.96815 - µ tan )

tan = Stem lead

• d d = Ds - Pitch 2

Pitch = 1 TPI TPI = Stem threads per inch Stem Lead = Pitch

• Thread type (single, double, triple)

µ = Stem to stem nut friction coefficient (if unknown use .08)

DS = Stem diameter For stems with a pressed fit collar, if the maximum thrust applied is less than the pressed fit collar capability (after the stem with the pressed fit collar has been validated to be adequately torqued into the upper wedge assembly) then the applied actuator torque/thrust will not cause a Issue Date: DRAFT 10 2515/195

loss of preload due to slip of the collar. However, the thrust capability of a pressed fit collar is difficult to verify based on the potential degradation of the collar preload as a result of shipping, installation, and overthrust conditions.

After the maximum applied thrust value has been determined, calculate the stem and upper wedge thread shear capability. The following formulas for thread shear are equations from ASME B1.1 - 2003 Unified Inch Screw Threads (UN and UNR thread form).

First, calculate the thread geometric shear areas:

Where ASn = minimum thread shear area for internal threads ASs = minimum thread shear area for external threads 1/P = number of threads per inch LE = length of engagement d min. = minimum major diameter of external thread d2 min. = minimum pitch diameter of external thread D1 max. = maximum minor diameter of internal thread D2 max. = maximum pitch diameter of internal thread Issue Date: DRAFT 11 2515/195

Issue Date: DRAFT 12 2515/195 Next calculate the shear strength of both the stem threads (ASs) and the upper wedge threads (ASn):

Shear Strength of Threads = 0.5St (ASn or ASs)

St = ultimate tensile strength of material, psi.

If the maximum applied thrust is greater than the calculated shear strength of the stem or upper wedge threads, the capability of the stem-disc connection is challenged (given the assumptions in the analysis) and further justification is required.

Issue Date: DRAFT 13 2515/195

If the maximum applied thrust is less than the calculated shear strength of the stem and upper wedge threads, evaluate the capability of the wedge pin to withstand the applied torque. The basic equation for calculating the applied torque to the wedge pin as indicated in the BWROG Revision 4 guidance document is as follows:

Where ODPSTEM = Stem pitch diameter @ upper wedge ODPIN = Wedge pin diameter Suggested allowable wedge pin shear stress values in the BWROG Revision 4 guidance document are: (this may vary based on plant design and/or weak link analysis)

= 0.577*SY (pure shear limit per distortion energy theory)

= 0.50*SY (pure shear limit per maximum shear stress theory)

Recommend use of = 0.50*SY unless justified in the plant specific analysis and/or applicable code.

If the stem and upper wedge threads are capable of withstanding the maximum applied thrust and the wedge pin shear strength is capable of withstanding the maximum applied torque, the valve is not susceptible to degradation of its stem-disc connection. If the stem and upper wedge threads are capable of withstanding the maximum applied thrust but the wedge pin shear strength is less than the maximum applied torque, then the wedge pin is subject to failure and the valve is susceptible to degradation of its stem-disc connection.

The BWROG Revision 4 guidance document allows credit to be taken for thread resistance of the stem to upper wedge connection to reduce the torque induced shear load on the wedge pin provided that the valve being analyzed shows no indication of thread damage (e.g., no anomalous behavior from diagnostic trending that could potentially be attributed to thread /

upper wedge threaded joint damage) and assumed thread friction is conservative.

However, the NRC staff notes that there is no accepted guidance available in the industry regarding how to credit thread friction to resist applied torque. Power screw formulas would apply but uncertainty is very high due to unknown friction factors, actual condition of the threads, thread clearances, and the dynamic nature of the applied thrust and torque. The BWROG Revision 4 guidance document allows reasonable engineering judgement when crediting thread resistance. The NRC staff considers that thread resistance should only be used to determine valve operability for the short term and should not be relied upon for a long term fix. The staff considers a reasonable value to use when crediting thread friction for the short term is 0.10. Any higher value used to credit thread friction warrants additional attention (e.g., stem rotation check during each quarterly valve exercise). Currently the BWROG Issue Date: DRAFT 14 2515/195

guidance document requires a diagnostic/stem rotation check frequency every 2 years for susceptible valves with no symptoms of stem/wedge degradation. The NRC staff considers this to be reasonable until the valves can be reworked. Valves that credit thread friction should be considered susceptible and fall within the corrective action program. NRR/DE/EMIB can be contacted for assistance with evaluating thread friction.

04.03 Review Revision 4 of the BWROG guidance document provides that repair/replacement of valve internals will be prioritized based on risk and valve function as follows:

Category A - all applicable (e.g., susceptible) GL 96-05 valves that are high or medium risk and traverse multiple times to perform their safety function. Repair next outage/within 2 years; Category B - all remaining applicable GL 96-05 valves that are high or medium risk (e.g., only traverse once, open or closed, to perform their safety function). Repair at next outage or pass diagnostic test during the next outage and repair within 2 outages/4 years; and Category C - all remaining applicable GL 96-05 valves that are low risk. Repair at next outage or pass diagnostic test during each of the next 2 outage and repair within 3 outages/6 years.

Review the following:

1. Review the methodology the licensee used to apply risk insights in categorizing the valves. Licensees must use a risk ranking method that has been accepted by the NRC staff on a plant-specific or industry-wide basis with the conditions in the applicable safety evaluations. The inspector should contact NRR/DE/EMIB if the licensee is found to be applying a risk ranking method that has not been previously accepted. Reference documents include:

• ASME Code Case OMN-3 Requirements for Safety Significance Categorization of Components Using Risk Insights for Inservice Testing of LWR Power Plants with the conditions discussed in Regulatory Guide 1.192

• BWROG Topical Report on MOV risk categorization which was accepted with conditions by NRC in a safety evaluation issued 2/27/1996

• Westinghouse Owners Group (WOG) Topical Report on risk categorization which was accepted with conditions by NRC in a safety evaluation issued 4/14/1998

2. Review the actuator capability calculations for the thrust and torque being applied to the A/D DDGVs at the applicable nuclear power plant. The calculations must use conservative factors of applying maximum voltage to the motor and assume a conservative coefficient of friction of the stem to stem-nut interface. For this evaluation, a conservative coefficient of friction of the stem to stem-nut interface would be a low value, which would yield higher applied thrust to the valve stem. Licensees may be able to use actual measured friction factors that have been increased for expected variation and measurement uncertainties. If actual measured friction factors are not available a conservative coefficient of friction value of 0.1 may be used. Reference documents include:

Issue Date: DRAFT 15 2515/195

• NUREG/CR-6750 Performance of MOV Stem Lubricants at Elevated Temperatures

• Electric Power Research Institute (EPRI) TR-103237-R1 EPRI MOV Performance Prediction Program Revision 1

• EPRI TR-106563-V1 Application Guide for Motor-Operated Valves in Nuclear Power Plants Revision 1: Gate and Globe Valves

• Limitorque Selection Procedure Guides SEL 1 through 16 (available on the EPNB Component Performance & Testing share point web site http://fusion.nrc.gov/nrr/team/de/cpnb/cptb/default.aspx)

3. Review the structural capability and weak link calculations for the A/D DDGVs at the applicable nuclear power plant. Calculations may include wedge pin analysis, pressed on collar capability analysis, and stem to wedge thread capability analysis. Reference documents include:

• ASME B1.1-2003, Unified Inch Screw Threads (UN and UNR Thread Form),

American Society of Mechanical Engineers, 2003

4. Based on the review of the licensee calculations, identify the susceptible and non-susceptible valves from the A/D DDGV population. For example, non-susceptible valves as identified by the calculations above include (1) valves where the maximum actuator capability is less than the shear capability of the wedge pin, stem threads, and other weak link components; and (2) valves where the integral collar stem has been torqued into the wedge assembly at a higher value than the maximum actuator capability within the Flowserve Part 21 specified stem preload torque, and the wedge pin is properly installed. Other valves, such as those with a pressed fit collar, will need to be specifically evaluated for their short term or long term resolution as described in this TI. It is important to note that design basis documents (e.g., weak link analyses) of valves determined to be non-susceptible may need to be updated to include the structural capability of the wedge pin and the pressed fit stem collar.

5. If the maximum expected operating torque (used in determining that a valve is non-susceptible) is less than the motor stall torque capacity, review the licensees plans to manage an over-torque event as detailed above. Also verify that the licensees records acceptably confirm that there have been no historic over torque events.

6. Review the adequacy of the corrective action plan and the timeline to complete.

Justification is required for the schedule that exceeds the specifications of the stated valve categories in Revision 4 to the BWROG guidance document.

7. For those valves that have already been repaired, select a sample and review to determine whether the repair meets the Flowserve Part 21 recommendations.

8. In all cases, if operability or functionality of the susceptible valves is challenged, the licensee shall follow their operability or functionality process to provide reasonable assurance that the valves will be able to perform their specified safety functions until corrective actions are implemented, independent of the valves category. The diagnostics and stem rotation checks described in Revision 4 to the BWROG guidance Issue Date: DRAFT 16 2515/195

required to be performed every 2 years can provide reasonable expectation of operability over the short term until corrective actions are implemented.

2515/195-05 REPORTING REQUIREMENTS The inspection results of this TI should be included in the integrated quarterly inspection report.

A copy of the report containing the results should be forwarded to NRR/DE/EMIB, Attention Michael Farnan, via e-mail at Michael.Farnan@nrc.gov. Mr. Farnan can also be reached at (301) 415-1486. The inspection results from this TI will be used to evaluate the licensees promptness in addressing Flowserve Part 21 notification Stem-Wedge Separation of an Anchor/Darling Double Disc Gate Valve at Exelon, LaSalle County Station, Unit 2, February 2017. Inspectors should contact NRR/DE/EMIB with any questions related to the scope of this TI or with questions related to other inspector concerns identified while implementing this TI.

2515/195-06 COMPLETION SCHEDULE This TI is to be completed by December 31, 2019.

2515/195-07 EXPIRATION The TI will expire on June 30, 2020 2515/195-08 CONTACT(S)

Any technical questions regarding this TI should be addressed to Michael Farnan at (301) 415-1486 and/or email Michael.Farnan@nrc.gov.

2515/195-09 STATISTICAL DATA REPORTING Charge all direct inspection and information collection efforts to TI 2515/195 using IPE code TI (CAC 001462). Charge all preparation and documentation time to activity code TPD (CAC 000989).

2515/195-10 RESOURCE ESTIMATE Estimated time to complete this TI is 80 hours9.259259e-4 days <br />0.0222 hours <br />1.322751e-4 weeks <br />3.044e-5 months <br /> per site for direct inspection (two inspectors for one week) and information collection efforts and 24-32 hours for preparation and documentation.

2515/195-11 TRAINING Conducted at the regions during Spring/Summer 2018 Issue Date: DRAFT 17 2515/195

2515/195-12 REFERENCES

1) Tennessee Valley Authority (TVA) 10 CFR Part 21 notification letter to the NRC subject titled Anti-Rotation Pin Failure in (Flowserve) Anchor/Darling Double Disc Gate Valve 1/24/2013 (ADAMS Accession No. ML13008A186)

2) Flowserve 10 CFR Part 21 notification, Wedge Pin Failure of an Anchor/Darling Double Disc Gate Valve at Browns Ferry Nuclear Plant Unit 1 2/25/2013 (ADAMS Accession No. ML13064A012)

3) Flowserve updated Part 21 notification Stem-Wedge Separation of an Anchor/Darling Double Disc Gate Valve at Exelon, LaSalle County Station, Unit 2, 7/11/2017 (ADAMS Accession No. ML17199F890)

4) Information Notice IN 2017-03 ANCHOR/DARLING DOUBLE DISC GATE VALVE WEDGE PIN AND STEM-DISC SEPARATION FAILURES (ADAMS Accession No. ML17153A053)

5) BWROG Guidance Document TP16-1-112, Revision 4, "Recommendations to Resolve Flowserve 10CFR Part 21 Notification Affecting Anchor Darling Double Disc Gate Valve Wedge Pin Failures (Revision 4)," (ADAMS Accession No. ML17243A137)

6) ASME Code Case OMN-3 Requirements for Safety Significance Categorization of Components Using Risk Insights for Inservice Testing of LWR Power Plants with the regulatory positions discussed in Regulatory Guide 1.192

7) Regulatory Guide 1.192 Operation and Maintenance Code Case Acceptability, ASME OM Code Revision 2

8) BWROG Topical Report NEDC-32264 Revision 1 Application of Probabilistic Safety Assessment to GL 89-10 Implementation 9/10/1994 (ADAMS Legacy Microfiche 81307:

008 to 81307: 225)

9) NRC safety evaluation of BWROG Topical Report NEDC-32264 Revision 1 issued 2/27/1996 (ADAMS Legacy Microfiche 87458: 122 to 87458: 131)

10) WOG Topical Report V-EC-1658 Revision 1 Risk Ranking Approach for MOVs in Response to GL 96-05 12/21/1997 (ADAMS Legacy Microfiche A1626: 274 to A1626:

322)

11) NRC safety evaluation of WOG Topical Report V-EC-1658 Revision 1 issued 4/14/1998 (ADAMS Legacy Microfiche A3718: 321 to A3718: 331)

12) NUREG/CR-6750, Performance of MOV Stem Lubricants at Elevated Temperatures (ADAMS Accession No. ML020150282)

13) EPRI Topical Report TR-103237-R1 EPRI MOV Performance Prediction Program Revision 1, (Can be viewed at NRR/EMIB share point site http://fusion.nrc.gov/nrr/team/de/emib/default.aspx)

Issue Date: DRAFT 18 2515/195

14) EPRI Topical Report TR-106563-V1, Application Guide for Motor-Operated Valves in Nuclear Power Plants Revision 1: Gate and Globe Valves (Can be viewed at NRR/EMIB share point site http://fusion.nrc.gov/nrr/team/de/emib/default.aspx)

15) Limitorque Selection Procedure Guides SEL 1 through 16 (Can be viewed at NRR/EMIB share point site http://fusion.nrc.gov/nrr/team/de/emib/default.aspx)

16) ASME B1.1-2003, Unified Inch Screw Threads (UN and UNR Thread Form), American Society of Mechanical Engineers, 2003 List of Attachments: - Assessment of Anchor Darling Double Disc Gate Valves Flowchart - Limitorque SEL-7 Gate and Globe Valve Efficiency Chart - Limitorque SEL-8 Sluice Gates Efficiency Chart Issue Date: DRAFT 19 2515/195

Issue Date: 20 2515/195 Attachment 2 Limitorque SEL-7 Gate and Globe Valve Efficiency Chart UNIT EFFICIENCIES UNIT EFFICIENCIES UNIT RATIO 1500/1800 MOTOR SPEEDS 3000/3600 MOTOR SPEEDS SIZE RANGE PULLOUT RUN APPROX. PULLOUT RUN APPROX.

STALL STALL 12.5030.60 60 60 80 60 65 80 000 33.50100.00 40 50 50 40 50 55 102.00136.00 35 45 45 35 50 50 9.7022.04 60 60 80 65 65 90 23.0041.00 40 50 60 45 60 65 00 43.60109.00 40 50 50 40 55 60 114.00183.90 35 45 45 35 50 50 11.8017.50 65 65 85 65 70 95 18.5026.10 60 65 80 65 70 90 26.4241.33 40 55 55 45 55 60 0

43.6996.20 40 50 50 40 50 55 102.60150.80 35 45 45 40 50 50 158.30247.00 30 40 40 35 45 45 11.6017.12 60 65 85 65 70 95 18.1325.65 60 65 85 65 65 90 27.2040.15 40 50 55 45 60 60 1

42.5088.40 40 50 50 40 55 55 92.40171.60 35 45 45 35 50 50 191.70234.00 30 40 40 35 45 45 10.6017.77 60 65 85 65 70 95 18.8525.55 60 60 85 60 65 90 26.2441.51 40 55 55 45 60 60 2

43.9982.50 40 50 50 40 55 55 84.84150.00 35 45 45 35 50 50 153.00212.50 30 40 40 35 45 45 11.0524.11 65 70 90 70 75 95 25.7637.28 60 70 90 65 70 95 43.8757.40 40 55 60 45 60 60 3

61.5095.53 40 50 60 40 55 60 98.61132.81 38 45 50 38 50 55 138.40186.40 33 45 50 35 50 50 10.1332.30 65 70 90 70 75 95 33.6048.45 60 60 90 65 70 95 51.79124.95 40 55 55 40 60 60 4

131.78147.90 35 50 50 38 55 55 152.13219.30 33 45 45 35 50 50 61.4296.40 40 55 60 45 60 65 5

101.12230.17 40 50 60 40 55 65 Shaded Area = Non-Locking Gearing Issue Date: 21 2515/195

Attachment 3 Limitorque SEL-8 Sluice Gates Efficiency Chart UNIT EFFICIENCIES UNIT EFFICIENCIES UNIT RATIO 1500/1800 MOTOR SPEEDS 3000/3600 MOTOR SPEEDS SIZE RANGE PULLOUT RUN APPROX. PULLOUT RUN APPROX.

STALL STALL 12.5030.60 55 60 70 55 65 75 000 33.50100.00 35 50 50 40 50 55 102.00136.00 30 45 45 35 50 50 9.7022.04 50 60 70 55 65 80 23.0041.00 38 50 55 40 60 60 00 43.60109.00 35 50 50 35 55 55 114.00183.90 30 40 45 30 50 50 11.8017.50 55 65 75 60 70 85 18.5026.10 50 65 75 55 70 80 26.4241.33 40 55 55 40 55 60 0

43.6996.20 35 50 50 40 50 55 102.60150.80 35 45 45 35 50 50 158.30247.00 30 40 40 30 45 45 11.6017.12 55 65 75 55 70 90 18.1325.65 50 65 75 50 65 85 27.2040.15 40 50 55 40 60 65 1

42.5088.40 35 50 50 35 55 60 92.40171.60 33 45 45 35 50 55 191.70234.00 30 40 40 33 45 50 10.6017.77 55 65 75 60 70 90 18.8525.55 50 60 75 55 65 90 26.2441.51 40 55 55 40 60 60 2

43.9982.50 35 50 50 35 55 55 84.84150.00 33 45 45 35 50 50 153.00212.50 30 40 40 33 45 45 11.0524.11 60 70 80 60 75 90 25.7637.28 55 70 80 55 70 90 43.8757.40 40 55 60 40 60 60 3

61.5095.53 35 50 60 40 55 60 98.61132.81 35 45 50 35 50 55 138.40186.40 30 45 50 30 50 50 10.1332.30 60 70 80 65 75 80 33.6048.45 55 60 75 60 70 80 51.79124.95 40 55 55 40 60 60 4

131.78147.90 35 50 50 38 55 55 152.13219.30 30 45 45 33 50 50 61.4296.40 40 55 60 40 60 65 5

101.12230.17 35 50 55 33 55 60 Shaded Area = Non-Locking Gearing Issue Date: 22 2515/195

Revision History for TI 2515/195 Commitment Accession Description of Change Description of Comment Resolution Tracking Number Training Required and and Closed Feedback Number Issue Date Completion Date Form Accession Change Notice Number (Pre-Decisional, Non-Public Information )

ML18058A131 Initial issuance. Researched commitments for the ML18058A130 DRAFT last four years and found none. This TI applies to CN 18- all operating nuclear power reactor licensees that have Anchor/Darling Double Disc Gate Valves (A/D DDGVs) with threaded stem to upper wedge and a locking wedge pin design installed in safety-related systems.

Issue Date: Att1-1 2515/195