ML18064A828

SE Re Cracking of Inconel 600 Components at Plant

ML18064A828
Person / Time
Site:	Palisades
Issue date:	06/27/1995
From:	NRC (Affiliation Not Assigned)
To:
Shared Package
ML18064A826	List: ML18064A825 ML18064A828
References
NUDOCS 9507060109
Download: ML18064A828 (17)

Text

..

~p.R Rf Gu{

~"\\;

"9>-

f¥,

~*1

(

I UNITED STATES NUCLEAR REGULATORY COMMISSION

~

~

WASHINGTON, D.C. 20555-0001

~

~

z,-? * * * * ~ ~

0

=SA....,F'-=E=--"T__._Y--'E=V'"'-A=LU=A'"""'T-=I=ON=-=B-=-Y--'T""'"'H=E--'O"-"F-'-F-=-I C=E.___O..... F __ N.___U=C L"""E"'""A"""'R_,R=E.,_,A-=-CT-'-'O"-'-R'--'"""'RE"'""'G""'"U-=LA..... T

...... I~ON CRACKING OF INCONEL 600 COMPONENTS CONSUMERS POWER COMPANY PALISADES PLANT DOCKET NO. 50-255 1

• 0 I NTRODUCTU\\JN 1.1 Purpose The purpose of this safety evaluation (SE) is to assess the following licensee submittals:

0 Ailloy 600 Project Plan Palisades Nuclear Plant," Revision 1, dated February 27, 1995; the supplemental "Response to Request for Additional Information to Support NRC Review of the Project Plan for Addressing Alloy 600 Issues," dated April 24, 1995; "Justification for Continued Operation for the Repaired Pressurizer Instrument Nozzles," dated April 28, 1995; "Additional Information to S11;pport ConUnued Operation for the Repaired Pressurizer Instrumentatiomi 'ffozzle, 11 dated June 7, 1995; and the "Project Pl an for Addressing Alloy 600 Issues - Revised Inspection Scope," dated June 15, 1995.

1.2 Background

During plant startup from the 1993 refueling outage, the licensee discovered a leak in the pl1'\\e.ssurizer pil~r-operated relief valve (PORV) nozzle safe end.

The licensee repaired the line. Then it discovered and repaired cracks in two temperature element (TE) penetrations on the pressurizer.

An evaluation determined tha!t J>'rimary waiter stress-corrosion cracking (PWSCC) was the most probable cause of the cracking of these Alloy 600 components.

The details are described in the staff's SE of January 18, 1994.

As a result of these leaks,. the licensee in its letter dated October 7, 1993, committed to developing a p:lan to address PWSCC of Inconel 600 components in the Palisades Plant primary coolant system (PCS).

This plan was to include the technical bases for planned inspections, modifications, repairs, and replacements, particularly for the.PORV, surge line, and spray line nozzles.

The licensee was to submit this plan 3 months before the next refueling outage. This plan was submitted to the NRC on February 27, 1995, and revised June 15, 1995. ITihe plan was revised because about 3 weeks into the outage, the licensee excee:dled its predkted radiation dose for inspections by four to five times.

9507060109 950627 PDR ADOCK 05000255

• P*

PDR l

• J 2.0 DISCUSSION AND EVALUATION 2.1 Licensee's Submittal 2.1.1 Alloy 600 Project Plan In developing its plan, CPCo has been keeping up to date with industry experience by participating in Combustion Engineering Owners Group (CEOG),

Nuclear Energy Institute (NEI), Electric Power Research Institute (EPRI),

American Society of Mechanical Engineers (ASME), and American Nuclear Society (ANS) meetings and other industry forums, consulting with the major vendors, and actively interacting with other utilities planning Alloy 600 inspections.

Incorporating inputs from all these sources, CPCo developed a plan that identifies all Alloy 600 components in the PCS, ranks them according to PWSCC susceptibility, and establishes a program for inspection, repairs, and mitigation.

2.1.1.1 Inspection Program Development The licensee's plan affects 251 Alloy 600 penetration nozzles and safe ends in the PCS.

The licensee used four criteria to prioritize the inspection of the components:

Susceptibility: considers the material heat treatment temperature, carbon content, fabrication process, yield strength, weld configuration, postweld heat treatment, and service temperature.

Consequence: postulated failures bounded by existing large and small break LOCA [loss-of-coolant accident] analyses; control rod ejection also bounded by analysis. Consequences ranged from permanent plant shutdown to simple repair during an unforced outage.

Contributing factors included the potential for core damage, plant conditions required for repair, whether the cracking was circumferential or axial in nature, location of the affected components, difficulty of post-incident cleanup before restart, and public and regulatory perception.

Detectability: criteria based upon leak detection margin using Leak Before Break concepts for circumferential cracking in Alloy 600 girth butt-welded components.

The licensee states that axial cracks at J-groove welded penetrations can be detected effectively by the boric acid walkdown program before the cracks become critical.

ALARA [as low as is reasonably achievable]: considered indepth inspections of components of low PWSCC susceptibility and low consequence profiles to cause unnecessary exposure in high dose areas.

These criteria resulted in a classification scheme consisting of the following three groups of components.

They are ranked from high to low priority for inspection:

GROUP I Pressurizer PORV Nozzle Safe End Pressurizer and Hot Leg Surge Nozzle Safe Ends Pressurizer Temperature Element (TE) Nozzles Primary Relief Valve (RV) Mounting Flanges Pressurizer Heater (EH) Sleeves with 44 degree < Setup Angles < 58 degrees Pressurizer Heater (EH) Sleeves with Setup Angles < 44 degrees Hot Leg Shutdown Cooling (SOC) Outlet Nozzle Safe End Pressurizer Level Indicator Tap Nozzles GROUP II Pressurizer Spray Safe End Reactor Head CROM [control rod drive mechanism] Nozzles with Setup Angles > 45 degrees Reactor Head Incore Instrumentation (ICI) Nozzles Cold Leg Safety Injection/Shutdown Cooling (SOC) Tap Penetrations Reactor Head CROM Nozzles with 22.5 degree < Setup Angles < 45 degrees Reactor Head Gas Vent Nozzle GROUP III Hot Leg Drain Penetration Hot Leg RTD [resistance temperature detector] Nozzles Reactor Head CROM Nozzles with Setup Angles < 22.5 degrees Cold Leg RTD Nozzles Cold Leg Drain, Charging, Letdown and Spray Penetrations Cold Leg Pressure (DPT) and Sampling (SX) Tap Penetrations The licensee identified examination methods for each component based on configuration, expected flaw type, and critical flaw size. The intent of the inspection program is to establish baselines for the most susceptible penetrations during the 1995 refueling outage.

The licensee will use the 1995 inspection results together with the susceptibility study to develop a long-term maintenance and inspection schedule.

Fracture Mechanics Assessment of Alloy 600 Components The licensee submitted a fracture mechanics analysis of the Palisades Alloy, 600 components performed by Babcock and Wilcox.

The licensee evaluated postulated internal axial and circumferential flaws under the fracture toughness requirements of the ASME Code,Section XI, IWRr-3612, considering the potential for crack growth and failure by net secti<m collapse (limit load).

Its objective was to develop a set of curves from which the allowable time for continued service can be determined for a given flaw size. This was done by evaluating both fatigue crack growth due to design cyclic loading and stress corrosion crack growth due to steady-state stresses for each of the Alloy 600 components.

• The licensee calculated the fatigue crack growth due to desfgn cyclic loading using the Paris Law, similar to that given in Article A-4000 of Section XI of the code and using rate data for Alloy 600 in a pressurized-water reactor (PWR) environment.

for stress-corrosion cracking the licensee stated:

the corrosion crack growth was calculated using Scott's model with an activation energy of 33 Kcal/mole.

This crack growth rate model was

• developed based on industry data for the stress corrosion cracking of Alloy 600 steam generator tubing. This model is considered the most conservative model available for Alloy 600 PWSCC on steam generator (SG) tubing.

An appropriate temperature correction is incorporated to account for the difference in operating temperature between the different Alloy 600 components and SG tubes.

The results showed that the amount of fatigue crack growth compared to that from stress-corrosion cracking is extremely small.

The licensee found that:

nozzles attached to PCS components by J-groove type welds (pressurizer TE nozzles and PCS loop RTD nozzles) are not connected to any type of external piping. This in turn virtually eliminates the potential for circumferential PWSCC, since there is no axial stress to either initiate or drive a crack.

On the other hand, circumferential PWSCC may very well be the primary mode of failure for full penetration type nozzles with girth butt welded safe ends.

Specific results were as follows:

For the pressurizer temperature element nozzles, the licensee states:

it would take about 7.5 years for an axial flaw and 13 years for a circumferential flaw to grow to a critical size.

For the spray nozzle safe end if the temperature ts assumed to be that of the cold leg, the component should be able to remain in service without failure for about 35 years.

However, if the pressurizer temperature is used the component could only be expected to last less than 3 years.

All other components can be expected to remain in service for 40 years without failure.

• 2.1.1.2 Resolution of Existing Pressurizer Penetration Concerns PORV Safe End After repairing the PORV safe end, the licensee reviewed its design and appropriate modifications to ensure a suitable lifetime. It decided to replace the safe end with a Type 316 stainless steel transition piece not susceptible to PWSCC.

It will weld this replacement to the stainless steel clad carbon steel nozzle with Alloy 690 weld material. This replacement with resistant material should eliminate the potential for PWSCC.

An analysis showed that the stresses on the existing safe end to piping are allowable so the replacement will not require modifications to the PORV downstream piping.

TE-0101 and TE-0102 Penetrations The licensee had detected an axial through-wall crack on each of two TE nozzles, TE-101 and TE-102, in October 1993.

The TE-101 nozzle is connected to the upper head by a partial penetration weld at the inner diameter (ID) and leaked from an annular space between the Inconel 600 nozzle and the carbon steel head.

As a solution, the licensee decided to modify the nozzle design by welding a pad to the exterior surface of the pressurizer to re-establish the structural support and pressure boundary.

It also modified TE-102 with an external weld pad.

To stop the leaks, the licensee replaced the original pressure boundary weld on the interior surface of the shell, the "J" weld, with a weld pad around each nozzle on its exterior surface. This exterior weld re-establishes the pressure boundary downstream of the original 11J 11 welds, replaces the structural support that may have been weakened by the cracks, and bypasses the cracked part of the nozzles.

Welding the nozzles to the shell at both the ID and outer surface can cause high stresses in the nozzle because of the differences in thermal expansion between the Inconel 600 nozzle and the carbon steel plate p*ressurizer.

The residual stress, caused by having the nozzle fixed at two locations to the shell, was evaluated. The evaluation showed that the nozzle for TE-101 needed to be severed within the thickness of the upper head to prevent the stresses caused by differences in thermal expansion during heatup and cooldown from exceeding the ASME allowable stresses. Severed, the outer part of the nozzle can expand at the same rate as the exterior of the upper head.

The nozzle for TE-102, located in the lower section liquid environment, was not severed since it does not experience as large a thermal gradient during heatup and cooldown.

For this nozzle, an axial residual stress field between the new weld (outside) and the J-weld (inside) may exist due to the existing restraint during the welding process. The axial stress developed on this nozzle as a result of accident transients is still within ASME Code stress and fatigue allowables even with the nozzle constrained at the inner and outer surface of the pressurizer.

The licensee intended the weld pad repairs to be for one fuel cycle but has since determined through analysis that these nozzles could be left in their current configuration.

CPCo submitted a letter dated April 28, 1995, justifying continued operation for the life of the plant. The justification included an ASME Code fatigue analysis and assessments of PWSCC at the new weld and corrosion of the carbon steel pressurizer by borated primary water.

The l i censee justified continued operation on the basis of the f o 11 owing:

The 1993 weld pad modification met ASHE III requirements for Class 1 compDnent fatigue.

CPCo and B&W analyses as well as industry experience and recent inspection results from other utilities of instrumentation nozzles show that pressurizer base metal exposed to primary coolant at existing cracks will corrode very slowly.

Thus further base metal corrosion inspections of repairs to these nozzles are not necessary.

The material at the weld pad (Alloy 182/82) is less susceptible to PWSCC than that near the original J-weld (Alloy 600). This indicates that the time for ID initiated PWSCC to develop leakage near the weld

-pad will be longer than the time it took to develop the axial cracks found in the 1993 outage. Moreover, an axial through-wall crack in the nozzle body is more likely to occur before a crack through the interface between the nozzle body and the weld pad. (Analysis showed that~ 0.010-inch flaw at the nozzle ID would grow to critical size in 7.5 years for an axial flaw and 13 years for a circumferential one.)

AxiaHy 'oriented PWSCC in the nozzle body is not a safety concern since tbe crack length is limited and will not result in an ejection failure,-of the nozzle. A crack would result in small leakage that can be easHy detected by visual inspection well before failure.

During the 1995 refueling outage, the licensee will perform a VT-2 inspection and some baseline inspections of the TE-101 nozzle and a VT-2 of the TE-102 nozzle.

It will use the results to determine the frequencies of future inspections.

CPCo plans to perform a VT-2 examination of the TE-101 nozzle at every refueling outage to identify any leakage from the nozzle in a timely way.

Since there is no significant short-term safety concern, discovering leakage by VT-2 inspections is CPCo's main line of defense.

The licensee does not intend to examine TE-102 every outage because TE-101 is the limiting case.

On TE-102, in contrast to TE-101, the licensee stated:

no electro-discharge machining (EDM) cut was made for the repair since the the.rmal fatigue loading conditions at TE-102 are less than the conditions.of the TE-101. Since the TE-102 nozzle body is still intact, both J-weld and pad are structural welds.

Therefore, even if the pad weld were completely severed, the J weld is still present to prevent the TE nozzle from any possible ejection. Thus, such a failure results in small leakage from TE-102 nozzle.

• e e Spray and Surge Nozzles The licensee assessed other pressurizer nozzles.

Its letter of April 28, 1995, addressed the acceptability of the condition of the spray and surge lines for operation.

CPCo calculated leakage rates from critical through-wall flaws based on results of a PICEP analysis performed for the spray and surge lines using normal operating loads and critical crack size for normal operation.

The critical through-wall circumferential crack length for the spray line was 6.7 inches and the corresponding leak rate 77.3 gpm.

For the surge line the corresponding values were 22.1 inches and 110 gpm.

The calculation is set forth in the licensee's submittal of November 30, 1993.

This leakage should be easily detectable as shown by CPCo's discovering the crack in the PORV nozzle through the leak detection system. That leakage rate was about 0.20 gpm.

CPCo considers the use of a constant through-wall stress distribution for axial residual weld stresses conservative for evaluating circumferential flaws in girth butt-welded components.

Therefore, it also applied an alternate method, that specified in NUREG-0313, "Technical Report on Material Selection and Processing Guidelines for BWR Coolant Pressure Boundary Piping."

In contrast to the first method, this method applies a nonlinear stress distribution. The licensee used this method to analyze the pressurizer surge and spray safe ends to attached piping welds.

Spray Nozzle The licensee calculated the remaining life of the spray nozzle safe end using two methods, the constant through-wall stress method (under the ASME Code) and by the NUREG-0313 method at two temperatures, 540°F and 640°F.

It calculated the time to failure for both temperatures since the operating temperature depends on the flow rate of coolant through the safe end.

It assumed a high flow rate, 50-70 gpm, leading to an operating temperature approaching that of the cold leg piping, 540°F.

The licensee believes this nozzle was reworked and therefore has high residual stresses.

(For the first analysis it set the constant through-wall axial stress equal to the material yield strength at operating temperature.

For the second analysis, it set the inner surface residual stress equal to the yield strength at operating temperature and determined the remaining through-wall distribution according to NUREG-0313).

Using the first method, the licensee calculated the lifetime to be 40 years for a circumferential flaw with an initial size of 0.010 inch and a 6:1 aspect ratio at 540°F.

The initial flaw size of 0.010 inch was chosen because quantitative measurements of surface residual stresses caused by machining processes were found at depths up to 0.010 inch.

At 640°F, the lifetime is much shorter, 3.6 years. The maximum allowable flaw size, 0.015 inch, calculated for 640°F, is below the ultrasonic examination (UT) detection limit (2 mm (0.079 inch)).

Using the NUREG-0313 method, the licensee calculated the time to failure to be 40 years and 5.36 years, respectively, for operating temperatures of 540°F and 640°F.

The licensee also stated that the critical leak rate is more than 70 times the plant technical specifications (TS) limit for PCS unidentified leakage.

The staff's previous SE stated that the inaccessible spray nozzle ID surface and outside diameter (OD) geometry prevent effective UT.

The licensee, to address this problem, fabricated a mockup to show that the critical flaw size can be detected by the technique.

As part of the IO-year inservice inspection (ISI} program, it will also examine the OD surface with PT.

The licensee states:

The Alloy 600 project is planning to perform a bare metal VT-2 and an internal video inspection of the spray nozzle safe-end to pipe weld during the 1995 refueling outage to confirm that no weld rework was performed during original installation of the nozzle.

Information concerning weld rework and condition of the spray nozzle safe-end discovered during this refueling outage will be used to evaluate the need for future inspections, repair and/or replacement of the spray nozzle safe-end.

Based on analysis results to date for a worst case scenario, we have shown that assuming a worst case temperature of 640°F, the pressurizer spray nozzle safe-end has an estimated service life of 5.36 years.

Future PWSCC inspection intervals of every other refueling outage will be sufficient to assure the long-term, safe operation of the pressurizer spray nozzle.

Surge Nozzle Safe End The licensee has detenrrined that with a postulated initial PWSCC of 0.010-inch deep, the time to failerire for this safe end would exceed 40 years.

The critical leak rate at failure would be more than 100 times the TS limit for unidentified leakage. The licensee justifies continued service for these reasons and on the basis of expected inspection results.

Because a leak at this location could result in a possible LOCA, the licensee determined that an initial inspection of this safe end be done to identify any cracks.

The licensee plans to perform follow-up inspections at its discretion, based on the results of previous inspections. It will procure and store Alloy 690 material for a possible replacement nozzle.

Further, because this nozzle was probably reworked, it is highly susceptible to cracking and would affect safety should it fail, CPCo will apply the mechanical stress improvement process (MSIP) at both the pressurizer and hot leg surge nozzle safe ends.

Safety Valve Flanges These flanges are of the same heat as the PORV and spray nozzle safe ends.

The nozzle girth butt welds were shop welded and post weld stress relieved, so the flanges are less susceptible to PWSCC than the safe ends.

Moreover, when examining these flanges {in 1993) by dye penetrant (PT) {internal and external), radiography (RT), and UT to detect intergranular stress corrosion

- - cracking {IGSCC), EPRI-qualified examiners did not find any indications. The licensee considers the leak detection margin of the safety valve nozzles to be as large as that of the PORV and spray safe ends {70 times TS limit).

The licensee states that if indications are discovered and continuing operation cannot be justified, stainless steel flanges can be attached in the same way as the replacement PORV piece.

Heater Sleeve Service Life Evaluation The licensee cites industry experience and analyses to show that PWSCC in heater sleeves is typically axial since cracks are controlled by residual hoop stresses a short distance from the J-welds.

The licensee considers circumferential cracking unlikely. The low axial stress and the fact that the nozzles are shrink fit (and thus resist crack opening) indicate that PWSCC at the J-welds is not a safety concern. Additionally, a CEOG evaluation determined that the Palisades heater sleeves were among the least susceptible sleeves of all CE plants. The licensee also contends that since Palisades operates with heaters full time, various benefits accrue such as better pressure and chemistry equalization in the PCS, less thermal cycling of heater elements, and a lower spray inlet nozzle safe~end temperature.

The licensee has committed to plugging leaking heater sleeves unless the number of sleeves exceeds the allowable number plugged, at which time it will repair or replace the sleeves. Alloy 690 nozzle plugs and replacement sleeve material are available.

Level Tap Safe Ends These are of the same heat as the TE nozzles but are considerably lower in PWSCC susceptibility because of a different design.

The safe ends are butt welded onto carbon steel nozzles, and residual stresses should be less than for the TE nozzle J-welds. Destructive mockup testing has confirmed acceptable residual stress levels in this design. Additionally, the welds are cooler than pressurizer internal temperatures since they are several pipe diameters outside the pressurizer inner wall and since the fluid is essentially stagnant inside the taps during normal operation.

The safe ends were radiographed in the 1993 outage in addition to the external surface being examined by dye penetrant.

No indications were found.

The licensee has Alloy 690 material available for replacements.

2.1.1.3 Assessment of Reactor Head Penetrations The licensee states that except for Alloy 600 PWSCC, no other industry issue exists that warrants considering internal inspection. The licensee, citing NUREG/CR-6245 "Assessment of Pressurized Water Reactor Control Rod Drive Mechanism Nozzle Cracking," further states that reactor head nozzle Alloy 600 PWSCC is not a safety issue.

Incore Instrumentation CIC!) Nozzle The licensee states that the ICI nozzle is similar in geometry and size to that of CROM penetrations but that the ICI nozzles are more susceptible to PWSCC due to the larger setup angle (slope of the vessel head at the penetration).

Control Rod Drive Mechanism CCRDMl Nozzle The licensee has determined that the risks associated with internal inspection of CROM penetrations outweigh near-term benefits for the following reasons:

1)

The CROM penetrations are a rack and pinion design which would require an extensive amount of work and planning to inspect.

2)

The CROM internals are highly contaminated and would cause considerable radiation exposure to workers, and be expensive.

3}

The CROM nozzles are restrained from ejection due to interconnection beneath the reactor head.

Each CROM nozzle is welded to an extension sleeve, and all 45 sleeves are welded to a common latticework approximately 17 in. above the head flange attachment plate.

Reactor Head Gas Vent Penetration Evaluation The licensee states that the reactor head vent line is an Alloy 600 nozzle welded to an Alloy 600 pipe approximately 6 feet long with two bends. The licensee states that it cannot be internally accessed from the top and that the bottom access is limited by CROM extension tubes and latticework.

External access is also severely limited due to the location of the CROM penetrations. The licensee determined that the vent penetration is among the least susceptible head penetrations because of lower yield material, near-center installation, and no postweld cold work.

Leakage risk is less than for Tf-0101 and TE-0102 due to lower operating temperature and lower yield strength material.

2.1.1.4 PCS Loop Penetrations and Safe Ends The licensee assessed PCS loop penetrations and safe ends.

It states that PWSCC growth rate in Alloy 600 is highly temperature dependent and that Palisades operates at relatively low temperatures compared to other PWRs, the primary system cold and hot legs being cooler than the reactor and

Pressurizer.

The licensee further states that since all hot and cold RTO

.penetration, drain, charging, spray, pressure, and sampling nozzles are among the least susceptible to PWSCC in the PCS, intensive examination in high radiation areas is not warranted.

11 -

Safe End and Full Penetration Welded Nozzle The licensee states that the field welds of the hot leg surge, shutdown cooling outlet, and safety injection safe ends are the most susceptible of the PCS loop components to PWSCC (in descending order). They are similar in size (12 in. OD and 1 in. thick). The time to failure was conservatively estimated to be over 40 years for a postulated 0.010-inch deep PWSCC (initial size).

Critical leak rate at failure is more than 100 times the plant TS for PCS unidentified leakage for these nozzles.

Some field welds were repaired during construction, but the inside surfaces of the welds are inaccessible and cannot be inspected for evidence of root repair. The licensee has committed to performing an UT inspection of these welds to detect possible PWSCC and has scheduled the MSIP for the surge and shutdown cooling outlet nozzle safe ends to delete high tensile stresses at the weld root area.

PCS Loop J-Welded Nozzle The licensee states that the PCS loop RTD nozzles belong to the lowest susceptibility group.

Potential cracks are axially oriented, and so are not an immediate safety concern.

CPCo will perform VT-2 inspections.

The licensee also states that any repair might require PCS drainage below the loop centerline.

2.1.1.5 Inspection Acceptance Criteria and Contingency Planning l?ressurizer The licensee, having performed fracture mechanics analyses with conservative PWSCC and fatigue crack growth data and loading conditions, generated service life curves for various flaw sizes and shapes, including allowable flaw sizes for one fuel cycle.

The licensee has committed to procuring Alloy 690 replacement material for any Alloy 600 safe end that cannot be justified for continued service.

Reactor Vessel Head The licensee will use the NRC-approved part of the NEI flaw acceptance criteria for CRDMs for eddy current testing {ECT) characterized and UT-sized indications from inside the ICI nozzles. These criteria are described in the staff's safety evaluation of March 9, 1994, "Acceptance Criteria for Control Rod Drive Mechanism Penetration Inspections at Point Beach Nuclear Plant, Unit 1.

11 Penetrations exceeding flaw acceptance criteria will be repaired.

Leaking CROM and head vent penetrations will be also be repaired.

PCS loops The licensee performed fracture mechanics analyses for all PCS loop penetrations and nozzle safe ends with satisfactory results. The licensee expects that inspection results will justify continued operation of existing PCS loop nozzle safe ends and penetrations but has replacement material if needed.

2.1.1.6 Mitigation Methods The licensee has evaluated three mitigation schemes for this outage:

Weld Overlay The licensee will use weld overlay techniques for either repair or stress reduction without further additional development.

Mechanical Stress Improvement Process The licensee decided to perform MSIP on the safe ends and pipe at the pressurizer and hot leg surge nozzles and the hot leg SDC nozzle.

Records of field welds showed that the welds were reworked, and therefore, probably under high residual stress. The nozzles connect to large diameter pipes (12 in. NPS

[nominal pipe size]) and are subjected to high temperatures of the hot leg and pressurizer and ranked as highly susceptible to PWSCC.

Failures would have significant safety consequences.

The licensee will include these locations in future inspections.

Axial MSIP for reactor head penetrations is still under development and will not be available for near-term use at Palisades.

Zinc Addition to PCS The licensee is evaluating the efficacy of zinc additions for possible future use.

2.1.1.7 Nondestructive Examination (NDE) for Each Type of Alloy 600 Penetration The licensee intends as its primary method UT with diffracted longitudinal wave transducers at various angles.

To develop and verify the procedures, it used mockups with implanted thermal fatigue cracks.

It stated that it demonstrated accurate flaw detection and sizing and can detect flaws of the size that would be allowed to exist for one operating cycle.

The licensee determined that the following groups of penetrations may be examined by the NDE methods described below.

"Enhanced UT" refers to techniques that exceed the standard ASME Section XI methods.

Specifically, it refers to the use of automated data collection and analysis, longitudinal

• • waves, focused search units, increased scanning sensitivity, tip diffraction sizing and optimization by the use of mockups of similar geometry with implanted thermal fatigue cracks.

Penetrations at the Primary Loops The following penetrations will be examined by enhanced UT and dye penetrant of the exterior surface:

Shutdown cooling outlet nozzle safe ends Surge nozzle safe ends on the hot legs Safety injection penetrations (4} to the cold legs The following were originally scheduled for examinations by UT and PT.

Letdown and drain nozzles Pressure measurement and sampling nozzles Spray nozzles on the primary loops Charging inlet nozzles However, the licensee revised its inspection plan (June 15, 1995 letter}

stating that it will not perform UT on the following:

On the hot leg:

Pressure, sampling, and drain welds On the cold leg:

Charging and spray nozzles Pressure and sampling nozzles Drain nozzles Safety injection nozzles nor PT on the spray, safety injection, charging nozzles on the cold leg, thus reducing the original 79 ultrasonic and 77 PT inspections to 27 ultrasonic and 61 PT examinations.

RTD nozzles on the hot legs and cold legs are welded at the loop ID with a J-weld and will receive a VT-2 only.

Penetrations of the Pressurizer The new PORV nozzle safe end will be examined by UT, RT, and dye penetrant examination of the exterior surface.

The spray nozzle safe end will be examined by enhanced UT and a dye penetrant examination of the exterior surface.

Remote weld root inspection (video) is also planned.

Safety valve nozzle flanges will be examined by enhanced UT and a dye penetrant examination of the exterior surface.

Temperature element TE-101 nozzle weld pad will have its weld root geometry established by UT techniques for use in future assessments.

Pressurizer base metal corrosion will be examined by UT (within TE-101 penetration) and TE-102 will be examined by VT-2.

The surge nozzle safe end at the pressurizer will receive the same examination as the surge nozzle safe end at the hot leg (same configuration).

Heater penetrations on the pressurizer will have a VT-2 examination.

Penetrations on the Reactor Head ICI flanges will be examined by ECT to identify indications.

UT and dye penetrant examinations will be used to size the depth of found indications.

CRD penetrations will be examined by VT-I for leaks by remote video (if possible). Otherwise, direct visual examination after insulation removal will be performed.

The licensee will evaluate the inspection results from this outage to determine future inspection intervals and methods.

2.2 NRC Evaluation Quality of CPCo's Plan The licensee's plan for managing PWSCC in Alloy 600 components was to include the technical bases for planned inspections, modifications, repairs, and replacements, particularly for the PORV, surge line, and spray line nozzles.

The licensee drew up a comprehensive plan that included all these elements.

The licensee's plan reflected a good understanding of the issues and sound approach to managing PWSCC in Alloy 600.

The contributing factors to PWSCC (stress, susceptible material, environment) are fairly well understood from considerable worldwide studies and field experiences.

The licensee's plan analyzed all these factors for each component.

The technical bases and assessments were consistent with current studies, experiences, and industry practice. The licensee drew on the expertise of the major vendors, CE, and Babcock & Wilcox (B&W).

It reviewed the current literature and PWSCC events worldwide in developing its plan.

CE and B&W supplied detailed analyses.

CPCo stated that there were no specific or unique factors contributing to PWSCC at Palisades to affect the analyses.

Inputs, such as crack growth rates, were up to date and generally used by the industry.

(The crack growth rates used were those developed by Peter Scott and published in "An Analysis of Primary Water Stress Corrosion Cracking in PWR Steam Generators," in Proceedings, Specialists Meeting on Operating Experience with Steam Generators, Brussels, Belgium, September 16-20, 1991.)

The crack growth rate model with its use of a low activation energy was conservative.

In general, the staff concurs with the methodology used by the licensee to rank Alloy 600 components for susceptibility to PWSCC.

However, the ranking methodology relies heavily on temperature and the resulting Arhennius plots.

The staff agrees in principle that temperature is a factor, and qualitatively useful, but using it quantitatively for life predictions for components of Alloy 600 has been shown by recent events to be questionable.

A possible weakness is the exclusion of microstructural evaluation as a criteria. The licensee did not have information on microstructure available.

The microstructure has a marked effect on a given heat's resistance to PWSCC.

Inter-and intra-heat variability can be quite diverse as shown at St. Lucie SG tube plug inspections. However, the licensee did consider the effects of the recognized important factors of heat treatments and carbon content on microstructures.

PORV. Surge and Spray Lines The NRC staff agrees that replacing the PORV safe end with one of resistant materi~l is a sensible course of action to ensure safe plant operation.

For the spray line, the staff finds the licensee's treatment valid. The licensee performed thorough fracture mechanics by two standard methods.

The inputs, such as the postulated initial crack sizes and growth rates, into the analyses were reasonable. Although 540°F is probably a realistic temperature, the licensee also conservatively calculated growth rates at 640°F, the worst case.

The use of enhanced UT, which surpasses code requirements, realistic detection limits (0.079 inch), and a critical leak rate at failure of 70 times the TS limit should ensure early detection of a crack.

The licensee's commitment to repair any detected cracks with PWSCC-resistant material is an accepted way to manage PWSCC.

Similarly, the staff finds the licensee's treatment of the surge line sound.

The licensee's analyses, similar to those for the spray line, had acceptable results. Calculations predicted a long life of 40 years.

They also showed that the acceptable flaw size for one cycle is greater than the detectable flaw size.

The use of enhanced UT and dye penetrant examinations and a critical leak rate at failure of 100 times the TS limit should ensure early crack detection.

As a precaution, the licensee has stocked a more resistant material, Alloy 690, in case replacement is needed.

As an additional mitigating measure, the licensee will perform MSIP on the surge line safe ends to mitigate the residual stress contributing to PWSCC.

e e TE Nozzles For the TE nozzles, the staff finds the licensee's justification for future operation sufficient. The licensee determined that industry experience and analysis of corrosion of the carbon steel pressurizer shell indicate no structural or safety concern for TE-101 and TE-102 nozzles.

If a leak does develop, visual inspection should easily identify it before failure.

Regarding the inspection of TE nozzles, the staff believes that although the stresses and temperatures are lower in the repair than the original weld, the short lifetimes predicted suggest the prudence of conducting a visual inspection during each refueling.

Inspecting TE-101 each refueling is reasonable since it is the limiting case.

Other Components The staff concurs with the analyses for the other components.

Analyses consistently followed the same methodology as for the components described abov;e..Management by enhanced examination, leak detection, and replacement with resistant material is logical and in accordance with the most current industry practice.

Cracking in CRDMs has been a current problem in many plants but has been determined to not be a major safety issue.

The main safety concern relating to CRDMs has been the ejection of control rods because of circumferential cracking.

The unique configuration at Palisades (welded connections to a common latticework) that restrains these nozzles from ejection is the major favorable factor.

Moreover, CE Owners' Groups have studied the effects of borated water on the low alloy steel head material. According to the CEN-614 report, "Safety Evaluation of the Potential for and Consequence of Reactor Vessel Head Penetration Alloy 600 OD-Initiated Nozzle Cracking," (December 1993), low level undetected leakage could persist for 8.8 years without degrading the integrity of the head, and borated water trapped within a seal crevice would result in only minor corrosion of the low alloy steel. It is unlikely, though, that boric acid corrosion following leakage from a through-wall crack could continue undetected.

Walkdown inspections required by Generic Letter 88-05, "Boric Acid Corrosion of Carbon Steel Reactor Pressure Boundary Components in PWR Plants, 11 would reveal evidence of leakage before ASME Code structural limits are challenged due to material loss from wastage.

The staff finds the NOE plans for this outage, inspection acceptance criteria, and contingency plans reasonable and consistent with analysis and industry practice. The licensee's plan to use current inspection results as feedback to determine future inspection plans is a logical approach.

The revised scope of the inspections, though reduced, assures that all the most susceptible Group 1 components and the more susceptible Group II components will receive either an ultrasonic or a dye penetrant inspection along with a visual inspection and that all 251 locations in the PCS will receive a visual inspection. The examinations that have been dropped from the scope were all scheduled for the lower susceptible locations in Group II and those in Group III, the least susceptible group.

Moreover, no PWSCC has been found by the inspections completed to date (according to the June 15, 1995, letter.) However, in view of the greatly reduced scope of the inspections, the licensee should consider a sampling plan for the next outage of the components missed during the current inspections.

Neither the staff nor the licensee can cite experience with MSIP as a method to mitigate stresses that cause PWSCC.

The technique is experimental for PWSCC, and no test data exists. At least one study has proposed it as a mitigation method for CRDMs.

This process has been used in boiling-water reactors (BWRs) to alleviate residual stresses that cause intergranular stress corrosion cracking, a similar phenomena to PWSCC and should similarly alleviate the residual stress factor contributing to PWSCC.

In view of the successful application of MSIP to BWRs, the use of this technique appears valid. Considering the susceptibility and safety consequences, it appears applying MSIP to these highly susceptible components would be better than doing nothing. However, inherent to the technique is that the area of high tensile stress that undergoes MSIP is moved to another location and the component remains susceptible. Therefore the licensee has committed to including subject components in its future inspection plans.

The staff finds these provisions acceptable.

3. 0 CONCLUSIONS Based.on the licensee's submittals, the staff finds that the licensee's plan to manage PWSCC in Alloy 600 components in the PCS at the Palisades Plant is

.adequate to ensure safe operation.

However, in view of the greatly reduced scope of the inspections, the licensee should consider for the next outage a sampling plan of the components missed during the current inspections.

The staff also finds the justification for the continued operation of the repaired instrument nozzles acceptable.

Principal Contributor:

M. Banic Date:

June 27, 1995