ML12171A520

Pre-Filed Hearing Exhibit NYS000380, Gorman, Et Al., Companion Guide to ASME Boiler & Pressure Vessel Code, Chapter 44, PWR Reactor Vessel Alloy 600 Issues (Dec. 19, 2009)

ML12171A520
Person / Time
Site:	Indian Point
Issue date:	12/19/2009
From:	Gorman J, Hunt S, Riccardella P, White G American Society of Mechanical Engineers (ASME)
To:	Atomic Safety and Licensing Board Panel
SECY RAS
Shared Package
ML12171A508	List: ML12171A511 ML12171A512 ML12171A513 ML12171A514 ML12171A516 ML12171A517 ML12171A518 ML12171A519 ML12171A520 ML12171A521 ML12171A524 ML12171A525 ML12171A526 ML12171A527 ML12171A528 ML12171A529 ML12171A530 ML12171A533 ML12171A534 ML12171A536 ML12171A537
References
RAS 22624, 50-247-LR, 50-286-LR, ASLBP 07-858-03-LR-BD01
Download: ML12171A520 (25)

Text

NYS000380 Submitted: June 19, 2012 ASME_Ch44_p001-026.qxd 12/19/09 7:36 AM Page 1 CHAPTER 44 PWR REACTOR VESSEL ALLOY 600 ISSUES Jeff Gorman, Steve Hunt, Pete Riccardella, and Glenn A. White

44.1 INTRODUCTION

44.2.1 Alloy 600 Base Metal Alloy 600 is a nickel-based alloy (72% Ni minimum, 14-17%

Primary water stress corrosion cracking (PWSCC) of alloy 600 Cr, 6-10% Fe) with high general corrosion resistance that has nickel-chromium-iron base metal and related alloys 82 and 182 been widely used in light water reactor (LWR) power plants, i.e.,

weld metal has become an increasing concern for commercial in PWRs and boiling water reactors (BWRs). In PWR plants, pressurized water reactor (PWR) plants. Cracks and leaks have alloy 600 has been used for steam generator tubes, CRDM been discovered in alloys 600/82/182 materials at numerous PWR nozzles, pressurizer heater sleeves, instrument nozzles, and simi-plant primary coolant system locations, including at several loca- lar applications. The alloy was originally developed by the tions in the reactor vessels. The reactor vessel locations include top International Nickel Corporation (INCO) and is also known as head control rod drive mechanism (CRDM) nozzles, top head ther- Inconel 600, which is a trademark now held by the Special Metals mocouple nozzles, bottom head instrument nozzles, and reactor Corporation [1]. The reasons that alloy 600 was selected for use vessel outlet and inlet nozzle butt welds. The consequences of this in LWRs in the 1950s and 1960s include the following [2-7]:

PWSCC have been significant worldwide with 72 leaks through May 2004 (56 CRDM nozzles, 13 reactor vessel closure head (a) It has good mechanical properties, similar to those of thermocouple nozzles, 2 reactor pressure vessel bottom-mounted austenitic stainless steels.

instrument nozzles, and 1 piping butt weld), many cracked noz- (b) It can be formed into tubes, pipes, bars, forgings, and cast-zles and welds, expensive inspections, more than 60 heads ings suitable for use in power plant equipment.

replaced, several plants with several-month outage extensions to (c) It is weldable to itself and can also be welded to carbon, repair leaks, and a plant shutdown for more than 2 years due to low-alloy, and austenitic stainless steels.

extensive corrosion of the vessel head resulting from leak-age (d) It is a single-phase alloy that does not require postweld heat from a PWSCC crack in a CRDM nozzle. This chapter addresses treatment. Also, when subjected to postweld heat treatments alloys 600/82/182 material locations in reactor vessels, operating that are required for low-alloy steel parts to which it is weld-experience, causes of PWSCC, inspection methods and findings, ed, the resulting sensitization (decreased chromium levels at safety considerations, degradation predictions, repair methods, grain boundaries associated with deposition of chromium remedial measures, and strategic planning to address PWSCC at carbides at the boundaries) does not result in the high sus-the lowest possible net present value cost. ceptibility to chloride attack exhibited by austenitic stain-Several example cases of PWSCC, and resulting boric acid cor- less steels that are exposed to such heat treatments.

rosion, are described in the following paragraphs of this chapter (e) It has good general corrosion resistance in high temperature and, in some cases, the remedial or repair measures are described. water environments, resulting in low levels of corrosion It is important to note that the repairs and remedial measures products entering the coolant and resulting in low rates of described may not apply to all situations. Accordingly, it is wall thinning.

important to review each new incident on a case-by-case basis to (f) It is highly resistant to chloride stress corrosion cracking ensure that the appropriate corrective measures are applied, (SCC), and has better resistance to caustic SCC than including the need for inspections of other similar locations that austenitic stainless steels.

may also be affected. (g) Its thermal expansion properties lie between those of car-bon/low-alloy steels and austenitic stainless steels, making it a good transition metal between these materials.

44.2 ALLOY 600 APPLICATIONS It was alloy 600s high resistance to SCC, especially chloride-Figure 44.1 shows locations where alloy 600 base metal and induced SCC, that led to its selection for steam generator tubing alloy 82 or 182 weld metal are used in PWR plant reactor ves- in PWRs in the 1950s and 1960s. Several early PWRs had experi-sels. It should be noted that not all PWR reactor vessels have enced SCC of austenitic stainless steel steam generator tubing, alloys 600/82/182 materials at each of the locations shown in variously attributed to chlorides and caustics, and this had led to a Fig. 44.1. desire to use a tubing alloy with increased resistance to these

ASME_Ch44_p001-026.qxd 12/19/09 7:36 AM Page 2 2

• Chapter 44 FIG. 44.1 LOCATIONS WITH ALLOYS 600/82/182 MATERIALS IN TYPICAL PWR VESSEL environments. Similarly, some early cases of SCC of stainless susceptibility in noncontaminated PWR primary coolant environ-steel nozzle materials in BWRs during initial plant construction ments. However, by the early 1970s, it had been confirmed by sever-and startup, which was attributed to exposure to chlorides and al organizations in addition to Coriou that PWSCC of highly fluorides, led to the wide-scale adoption of alloy 600 and its relat- stressed alloy 600 could occur in noncontaminated high-temperature ed weld materials for use in BWR vessel nozzles and similar pure and primary water environments after long periods of time applications [8]. [13-15]. Starting with Siemens in the late 1960s, some designers The first report of SCC of alloy 600 in high-temperature pure or began to move away from use of alloy 600 to other alloys, such primary water environments was that of Coriou and colleagues in as alloy 800 for steam generator tubes and austenitic stainless 1959 [9] at a test temperature of 350C (662F). This type of crack- steels for structural applications [15]. By the mid-1980s, alloy 690, ing came to be known as pure water or primary water SCC an alternate nickel-based alloy with about twice as much chromium (PWSCC) or, more recently, as low potential SCC (LPSCC). In as alloy 600 (~30% vs. ~15%), had been developed and began to response to Corious 1959 report of PWSCC, research was conduct- be used in lieu of alloy 600 for steam generator tubing [16]. By the ed to assess alloy 600s susceptibility to SCC in high-temperature early 1990s, alloy 690 began to be used for structural applications pure and primary water. Most of the results of this research in the such as CRDM nozzles and steam generator divider plates.

1960s indicated that alloy 600 was not susceptible unless specific contaminants were present [10-12]. The conditions leading to sus- 44.2.2 Alloys 82 and 182 Weld Metal ceptibility included the presence of crevices and the presence of Weld alloys 82 and 182 have been commonly used to weld oxygen. Most of the test results of the 1960s did not indicate alloy 600 to itself and to other materials. These alloys are also

ASME_Ch44_p001-026.qxd 12/19/09 7:36 AM Page 3 COMPANION GUIDE TO THE ASME BOILER & PRESSURE VESSEL CODE

• 3 used for nickel-based alloy weld deposit (buttering) on weld vessels had eight 1.0-in. outside diameter alloy 600 thermocouple preparations and for cladding on areas such as the insides of reac- nozzles welded to the periphery of the head by J-groove welds.

tor vessel nozzles and steam generator tubesheets. Alloy 82 is bare Most of the Combustion Engineering vessels have alloy 600 electrode material and is used for gas tungsten arc welding incore instrument (ICI) nozzles welded to the periphery of the top (GTAW), also known as tungsten inert gas (TIG) welding. Alloy head by J-groove welds. These ICI nozzles are similar to CEDM 182 is a coated electrode material and is used in shielded metal arc nozzles except that they range from 4.5 to 6.6 in. outside diame-welding (SMAW). The compositions of the two alloys are some- ter. Several Westinghouse plants have 3.5 to 5.4 in. outside diame-what different, leading to different susceptibilities to PWSCC. ter alloy 600 auxiliary head adapters and de-gas line nozzles Alloy 182 has lower chromium (13-17%) than alloy 82 (18-22%) attached to the top head by J-groove welds. Several Westinghouse and has higher susceptibility to PWSCC, apparently as a result of plants have 5.3 to 6.5 in. outside diameter internals support hous-the lower chromium content. Most welds, even if initiated or com- ings and auxiliary head adapters attached to the vessel top head pleted with alloy 82 material, have some alloy 182 material. surface by alloy 82/182 butt welds.

In recent years, alloys 52 and 152, which have about 30% In summary, PWR reactor vessels have 38 to 102 alloy 600 noz-chromium and are thus highly resistant to PWSCC, have been zles welded to the top head, with most of these attached to the used in lieu of alloys 82 and 182, respectively, for repairs and for heads after stress relief of the head by alloy 82/182 J-groove welds.

new parts such as replacement reactor vessel heads.

44.2.4 BMI Penetrations 44.2.3 RPV Top-Head Penetrations All of the Westinghouse and Babcock & Wilcox-designed reac-CRDMs in Westinghouse- and Babcock & Wilcox-designed tor vessels in the United States and three of the Combustion PWR plants and control element drive mechanisms (CEDMs) in Engineering-designed reactor vessels in the United States have Combustion Engineering-designed PWR plants are mounted on alloy 600 instrument nozzles mounted to the vessel bottom heads.

the top surface of the removable reactor vessel head. Figure 44.2 These are often referred to as bottom-mounted instrument (BMI) shows a typical CRDM nozzle in a Babcock & Wilcox-designed nozzles. These nozzles range from 1.5 to 3.5 in. outside diameter.

plant. Early vintage Westinghouse PWR plants have as few as 37 As shown in Fig. 44.3, a typical BMI nozzle is welded to the bot-CRDM nozzles and later vintage Combustion Engineering plants tom head by a J-groove weld. In the case of the Westinghouse and have as many as 97 CEDM nozzles. These nozzles are machined Combustion Engineering plants, the J-groove welds were made from alloy 600 base metal with finished outside diameters ranging after stress relieving the vessel. In the case of the Babcock &

from 3.5 to 4.3 in. and with wall thicknesses ranging from about Wilcox-designed plants, the J-groove welds were made prior to 0.4 to 0.8 in. In some cases, a stainless steel flange is welded to vessel stress relief. Early test experience at a Babcock & Wilcox-the alloy 600 nozzle with an alloy 82/182 butt weld. The nozzles designed plant showed a flow vibration concern with the portions are installed in the reactor vessel head with a small clearance or of the BMI nozzles inside the bottom head plenum. Accordingly, interference fit (0.004 in. maximum interference on the diameter) all of the Babcock & Wilcock plant BMI nozzles were modified and are then welded to the vessel head by an alloy 82/182 after initial installation to increase the diameter of the portion of J-groove weld. The surface of the J-groove weld preparation is the nozzle extending into the lower plenum. The new extension coated with a thin butter layer of alloy 182 weld metal before was alloy 600 and the modification weld was made using alloy stress relieving the vessel head so that the nozzles can be installed 82/182 weld metal, with no subsequent stress relief heat treatment.

and the final J-groove weld can be made after vessel stress relief.

This avoids possible distortion that could occur if the CRDM noz- 44.2.5 Butt Welds zles were welded into the vessel head before vessel stress relief. Many Westinghouse reactor vessels have alloy 82/182 butt Most vessels have a single 1.0-1.3 in. outside diameter alloy welds between the low-alloy steel reactor vessel inlet and outlet 600 head vent nozzle welded to a point near the top of the head by nozzles and the stainless steel reactor coolant pipe, as shown in a J-groove weld. Two of the early Babcock & Wilcox-designed Fig. 44.4. In most cases, these welds include alloy 182 cladding on the inside of the nozzle and an alloy 182 butter layer applied to the end of the low-alloy steel nozzle prior to vessel stress relief.

FIG. 44.2 TYPICAL CONTROL ROD DRIVE MECHANISM FIG. 44.3 TYPICAL BOTTOM-MOUNTED INSTRUMENT (CRDM) NOZZLE (BMI) NOZZLE

ASME_Ch44_p001-026.qxd 12/19/09 7:36 AM Page 4 4

• Chapter 44 82/182 welds. In most cases, the vessel cladding in the area of the lugs is also alloy 182 weld metal.

44.2.7 Miscellaneous Alloy 600 Parts Most reactor vessel lower closure flanges have alloy 600 leak-age monitor tubes welded to the flange surface by alloys 82/182 weld metal. These are not discussed further since the leakage monitor tubes are not normally filled with water and, therefore, are not normally subjected to conditions that contribute to PWSCC.

44.3 PWSCC 44.3.1 Description of PWSCC PWSCC is the initiation and propagation of intergranular cracks through the material in a seemingly brittle manner, with little or no plastic deformation of the bulk material and without FIG. 44.4 TYPICAL REACTOR VESSEL INLET/OUTLET the need for cyclic loading. It generally occurs at stress levels NOZZLE close to the yield strength of the bulk material, but does not involve significant material yielding.

PWSCC occurs when three controlling factors, material sus-Babcock & Wilcox-designed plants, and all but one ceptibility, tensile stress, and the environment, are sufficiently Combustion-Engineering-designed plant, do not have alloy 82/182 severe. Increasing the severity of any one or two of the three butt welds at reactor vessel inlet and outlet nozzles since the reac- factors can result in PWSCC occurring, even if the severity of the tor coolant piping is low-alloy steel as opposed to stainless steel. remaining factor or factors is not especially high. The three Reactor vessel core flood line nozzles in Babcock & Wilcox- factors are discussed separately in the following sections.

designed plants have alloy 182 cladding and alloy 82/182 butt While mechanistic theories for PWSCC have been proposed, a welds between the low-alloy steel nozzle and stainless steel core firm understanding of the underlying mechanism of PWSCC has flood pipe. not been developed. Accordingly, the influence of material susceptibility, stresses, and environment must be treated on an 44.2.6 Core Support Attachments empirical basis, without much support from theoretical models.

Most PWR vessels have alloy 600 lugs attached to the inside surface of the vessel, as shown in Fig. 44.5, to guide the reactor 44.3.2 Causes of PWSCC: Material Susceptibility internals laterally or to support the reactor internals in the event of Based on laboratory test data and plant experience, the follow-structural failure of the internals. These lugs are attached to ing main factors influence the susceptibility of alloy 600 base cladding on the inside of the vessel by full penetration alloy metal and its weld alloys to PWSCC:

(a) Microstructure. Resistance to PWSCC tends to increase as the fraction of the grain boundaries that are decorated by chromium carbides increases. Various models have been proposed to explain this effect such as one where the car-bides act as dislocation sources and enhance plastic defor-mation at crack tips, thereby blunting the cracks and imped-ing their growth [17]. The absence of carbides in the matrix of grains also correlates with higher resistance to PWSCC, as does larger grain size [18].

(b) Yield Strength. Susceptibility to PWSCC appears to increase as the yield strength increases. However, this is considered to be a result of higher yield strength material supporting high-er residual stress levels and is, therefore, more of a stress than a material effect. As discussed in para. 44.3.3, tests indi-cate that the time to PWSCC initiation varies inversely with the fourth to seventh power of the total (applied plus resid-ual) tensile stress [19-21].

(c) Chromium Concentration. Tests of wrought materials and weld materials in the nickel-chromium-iron alloy group of materials consistently indicate that susceptibility to PWSCC decreases as the chromium content increases [22,23].

Materials with 30% chromium or more are highly resistant to PWSCC. The improved resistance of alloy 82 vs. alloy FIG. 44.5 TYPICAL CORE SUPPORT LUG 182 weld metal is attributed to the higher chromium

ASME_Ch44_p001-026.qxd 12/19/09 7:36 AM Page 5 COMPANION GUIDE TO THE ASME BOILER & PRESSURE VESSEL CODE

• 5 concentration of alloy 82 (18-22%) vs. that of alloy 182 However, axial stresses can also be high and circumferential (13-17%). Alloy 690 base metal and alloys, 52 and 152 cracks have occurred in a few cases.

weld metal, with about 30% chromium, have been found to For the case of butt welds, the weld shrinkage that occurs as be highly resistant to PWSCC in numerous tests. progressive passes are applied from the outside surface produces (d) Concentrations of Other Species and Weld Flaws. No clear tensile hoop stresses throughout the weld, axial tensile stresses on trends in PWSCC susceptibility have been observed as a the outside weld surface (and often also the inside weld surface),

function of the concentration of other species in the alloy and a region of axial compressive stress near midwall thickness.

such as carbon, boron, sulfur, phosphorous, or niobium. The hoop stresses can contribute to axial PWSCC cracks in the However, to the extent that these species, in combination weld and the axial stresses can contribute to circumferential with the thermomechanical processing to which the part is cracks. Finite element analyses show that the hoop stresses on the subjected, affect the carbide microstructure, they can have wetted inside surface of a butt weld are typically higher than the an indirect influence on susceptibility to PWSCC. Also, hot axial stresses at high stress locations, such that cracks are predict-cracks caused by some of these species (e.g., sulfur and ed to be primarily axial in orientation. However, if welds are phosphorous) can act as PWSCC initiators and, thus, repaired on the inside surface, or subjected to deep repairs from increase PWSCC susceptibility. the outside surface, the residual hoop and axial stresses on the wetted inside surface can both approach the yield strength of the 44.3.3 Causes of PWSCC: Tensile Stresses weld metal and can cause circumferential as well as axial cracks.

Industry design requirements, such as ASME BPVC Section III, specify the allowable stresses for reactor vessel components 44.3.4 Causes of PWSCC: Environment and attachments. The requirements typically apply to operating Several environmental parameters affect the rate of PWSCC condition loadings such as internal pressure, differential thermal initiation and growth. Temperature has a very strong effect. The expansion, dead weight, and seismic conditions. However, the effects of water chemistry variations are not very strong, assum-industry design standards do not typically address residual stress-ing that the range of chemistry variables is limited to those that es that can be induced in the parts during fabrication. These resid-are practical for PWR primary coolant, i.e., with the coolant con-ual stresses are often much higher than the operating condition taining an alkali to raise pH above neutral and hydrogen to scav-stresses and are ignored by the standards since they are secondary enge oxygen.

(self-relieving) in nature. It is the combination of operating condi-tion stresses and residual stresses that lead to PWSCC. (a) Temperature. PWSCC is strongly temperature dependent.

For the case of penetrations attached to the vessel heads by par- The activation energy for crack initiation is about 44 tial penetration J-groove welds, high residual stresses are caused kcal/mole for thick section nozzle materials [24] and 50 by two main factors. Firstly, the surfaces of nozzles are typically kcal/mole for thinner cold-worked steam generator tubing machined prior to installation in the vessel. This machining cold material [25]. The activation energy for crack growth is works a thin layer (up to about 0.005 in. thick) on the surface, about 31 kcal/mole [26]. Using these values, the relative thereby significantly increasing the material yield and tensile factors for crack initiation and growth at typical pressuriz-strength near the surface. Secondly, weld shrinkage, which occurs er and cold leg temperatures of 653F and 555F relative to when welding the nozzle into the high restraint vessel shell, pulls an assumed hot leg temperature of 600F are given in the nozzle wall outward, thereby creating yield strength level Table 44.1.

residual hoop stresses in the nozzle base metal and higher (b) Hydrogen Concentration. Tests using crack growth rate strength cold-worked surface layers. These high residual hoop specimens have shown that crack growth tends to be a max-stresses contribute to the initiation of axial PWSCC cracks in the imum when the hydrogen concentration results in the elec-cold-worked surface layer and to the subsequent growth of the trochemical potential being at or close to the potential where axial cracks in the lower strength nozzle base material. The lower the Ni/NiO phase transition occurs [27]. Higher or lower frequency of cracking in weld metal relative to base metal may values of hydrogen decrease crack growth rates. This effect result from the fact that welds tend not to be cold worked and can be substantial, with peak crack growth rates in some then subjected to high strains after the cold work. cases being up to four times faster when the hydrogen con-Residual stresses in the nozzles and welds can lead to crack ini- centration is at the value causing peak growth rate as com-tiation from the inside surface of the nozzle opposite from the pared to conditions with hydrogen values well away from weld, from the outside surface of the nozzle near the J-groove the peak growth rate value, as shown in Fig. 44.6 [27]. Tests weld, or from the surface of the J-groove weld. at various temperatures show that the hydrogen concentra-Most PWSCC cracks have been axially oriented. This is consis- tion for the Ni/NiO transition varies systematically with tent with results of finite element stress analyses, which predict temperature, and that the hydrogen concentration causing that the hoop stresses exceed the axial stresses at most locations. the peak growth rate exhibits a similar trend, with the

ASME_Ch44_p001-026.qxd 12/19/09 7:36 AM Page 6 6

• Chapter 44 FIG. 44.6 ALLOY 600 CRACK GROWTH RATE AT 338°C PLOTTED VS.

HYDROGEN CONCENTRATION [27]

concentration causing the peak crack growth rate becoming While tests of crack growth rate indicate increases in pH and lower as temperature decreases (e.g., 10 cc/kg at 320C, 17 lithium concentration within the normal ranges used for PWRs cc/kg at 330ºC, 24 cc/kg at 338C, and 27.5 cc/kg at 360C). have minimal effects on crack growth rate, some evaluations of Crack initiation may depend on hydrogen concentration in a crack initiation data indicate that increases in pH and lithium similar manner. However, enough testing to determine the cause moderate increases in the rate of crack initiation, e.g., in the effect of hydrogen on time to crack initiation has only been range of 10-15% for increases in cycle pHT from 6.9 to 7.2 [29].

performed at 330C, where it resulted in the most rapid However, recent tests sponsored by the Westinghouse Owners crack initiation in alloy 600 tubing at about 32 cc/kg vs. Group (WOG) indicate that the effect may be stronger, such as an about 17 cc/kg for peak crack growth rate. Reported data increase by a factor of two for an increase in cycle pHT from 6.9 regarding effects of hydrogen concentration on PWSCC ini- to 7.2. Further tests under EPRI sponsorship are underway (as of tiation and growth are shown in Fig. 44.7 [28]. The reasons 2004) to clarify this situation.

that the hydrogen concentration for peak aggressivity appears to be about twice as high for crack initiation vs.

crack growth rate (32 cc/kg vs. 17 cc/kg) are not known; the 44.4 OPERATING EXPERIENCE difference may be real or may be an artifact of data scatter or imprecision. 44.4.1 Precursor PWSCC at Other RCS Locations (c) Lithium Concentration and pH. Tests indicate that the PWSCC of alloy 600 material has been an increasing concern effects of changes in pH on crack growth rate, once the pH in PWR plants since cracks were discovered in the U-bend region is well above neutral, are minimal and cannot be distin- of the original Obrigheim steam generators in 1971. The history guished from the effects of data scatter [28]. However, when of PWSCC occurrences around the full reactor coolant system up considering the full pH range from acid to neutral to caus- though 1993, i.e., not limited to the reactor vessel, is documented tic, several tests indicate that crack growth rates decrease as in an EPRI report [31]. Between 1971 and 1981, PWSCC cracks pH is lowered to the neutral range and below, but is essen- were detected at additional locations in steam generator tubes tially constant for pHT of about 6 to 8 [29,30]. (e.g., at dents and at roll transitions), and in an increasing number of tubes. This experience showed that alloy 600 in the metallurgi-cal condition used for steam generator tubes was quite susceptible to PWSCC, with susceptibility increasing as stress, cold work, and temperature increase. It was found that susceptibility was also strongly affected by the microstructure of the material, with sus-ceptibility tending to decrease as the density of carbides on the grain boundaries increases.

The first case of PWSCC of alloy 600 in a non-steam generator tube application was reported in 1982. This incident involved PWSCC of an alloy 600 pressurizer heater sleeve [31]. Swelling of a failed electric heater element inside this sleeve was identified as a contributing cause. Subsequent to this occurrence, an increas-ing number of alloy 600 instrument nozzles and heater sleeves in pres-surizers have been detected with PWSCC. Also, increasing numbers of instrument nozzles in reactor coolant system hot legs and steam generator heads have also been detected with PWSCC.

FIG. 44.7 HYDROGEN CONCENTRATION VS. TEMPERA- Many of the susceptible nozzles and sleeves have (as of May TURE FOR N2/N2O PHASE TRANSITION, PEAK PWSCC 2005) been repaired or replaced on a corrective or preventive SUSCEPTIBILITY, AND PEAK CRACK GROWTH RATE [28] basis [31].

ASME_Ch44_p001-026.qxd 12/19/09 7:36 AM Page 7 COMPANION GUIDE TO THE ASME BOILER & PRESSURE VESSEL CODE

• 7 PWSCC in alloys 182 and 82 weld metals was first detected in The cracking discussed above was mainly related to PWSCC of October 2000 in a reactor vessel hot leg nozzle weld [32]. This alloy 600 base materials. Starting in November 2000, some plants was only a month before the first detection of PWSCC in a reac- found PWSCC primarily in the J-groove weld metal, e.g., in tor vessel head penetration weld, as discussed in para. 44.4.2. CRDM nozzle-to-vessel alloy 182 J-groove welds [37]. Since that time, several other cases of PWSCC of CRDM nozzle-to-head 44.4.2 RPV Top-Head Penetrations welds have been detected. Also, detection of PWSCC in alloys The first reported occurrence of PWSCC in a PWR reactor 182 and 82 welds appears to be increasing in frequency at other vessel application involved a leak from a CRDM nozzle at Bugey non-reactor vessel locations around the reactor coolant system.

3 in France that was detected during a 10-year inservice inspec- However, the frequency of PWSCC in welds remains lower than tion program hydrostatic test conducted in 1991 [33]. This initial in alloy 600 base material. For example, after the detection of occurrence, and the occurrences detected during the next few PWSCC in the weld metal of a CRDM nozzle at a PWR in the years, involved PWSCC of alloy 600 base material at locations United States in November 2000, and the detection of PWSCC in with high residual stresses resulting from fabrication. The high the alloy 182 weld metal at reactor vessel outlet nozzles in the residual stresses were mainly the result of weld-induced defor- United States and Sweden in late 2000, EDF inspected 754 welds mation being imposed on nozzles with cold-worked machined in 11 replaced reactor vessel heads without detecting any cracks surfaces. [24].

Subsequent to the initial detection of PWSCC in a CRDM nozzle in 1991, increasing numbers of plants detected similar 44.4.3 RPV Nozzle Butt Welds types of PWSCC, typically resulting in small volumes of leak- In October 2000, a visual inspection showed a leak from an age and boric acid deposits on the head surface as shown in alloys 82/182 butt weld between a low-alloy steel reactor vessel Fig. 44.8. In 2000, circumferential cracks were detected on the hot-leg outlet nozzle and stainless steel hot-leg pipe at the V.C.

outside diameter of some CRDM nozzles. In 2002, significant Summer plant. Destructive failure analysis showed that the leak wastage of the low-alloy steel Davis-Besse reactor vessel head was from a through-wall axial crack in the alloys 82/182 butt occurred adjacent to an axial PWSCC crack in an alloy 600 weld, as shown in Fig. 44.10. The axial crack arrested when it CRDM nozzle. The wastage was attributed to corrosion by boric reached the low-alloy steel nozzle on one side and stainless steel acid in the leaking primary coolant that concentrated on the pipe on the other side, since PWSCC does not occur in these vessel head. Figure 44.9 shows a photograph of the corroded materials. The axial crack can propagate into the low-alloy steel surface at Davis-Besse. The Davis-Besse plant was shut down and stainless steel by fatigue, but the fatigue crack growth rates for approximately 2 years for installation of a new head and will be low due to the small number of fatigue cycles. The incorporation of changes to preclude similar corrosion in the destructive examination also showed a short-shallow circumferen-future. The NRC issued several bulletins describing these events tial crack intersecting the through-wall axial crack that grew and requiring utilities to document their inspection plans for this through alloy 182 cladding and terminated when it reached the type of cracking [34-36]. low-alloy steel nozzle base metal. Examination of fabrication FIG. 44.8 TYPICAL SMALL VOLUME OF LEAKAGE FROM CRDM NOZZLE

ASME_Ch44_p001-026.qxd 12/19/09 7:36 AM Page 8 8

• Chapter 44 FIG. 44.9 LARGE VOLUME OF WASTAGE ON DAVIS-BESSE REACTOR VESSEL HEAD FIG. 44.10 THROUGH-WALL CRACK AND PART-DEPTH CIRCUMFERENTIAL CRACK IN V.C. SUMMER REACTOR VESSEL HOT-LEG OUTLET NOZZLE records showed that the leaking butt weld had been extensively In the 2005-2008 time period, the industry has begun imple-repaired during fabrication, including repairs made from the menting a massive inspection program for PWSCC in primary inside surface. Nondestructive examinations of other reactor ves- coolant loop Alloy 82/182 butt welds (In accordance with sel outlet and inlet nozzles at V.C. Summer showed some addi- Industry Guideline MRP-139 [58] - see Section 44.5.6 below tional shallow axial cracks. for complete discussion). Considering the temperature sensitivi-Shortly before the leak was discovered at V.C. Summer, part- ty of the PWSCC phenomenon discussed above, this program depth axial cracks were discovered in alloys 82/182 reactor vessel started with the highest temperature welds in the system: those outlet nozzle butt welds at Ringhals 3 and 4. Some of these cracks at pressurizer nozzles. To date, essentially all pressurizer nozzle were removed and two were left in place to allow a determination dissimilar metal butt welds (typically five or six per plant) have of the crack growth rate. The crack growth rate is discussed in been inspected, mitigated, or both. Approximately 50 nozzles para. 44.7.2. were inspected (many more were mitigated using weld overlays In addition to the PWSCC cracks in alloys 82 and 182 weld with no pre-inspections), resulting in PWSCC-like indications metal in reactor vessel CRDM nozzles and inlet and outlet nozzle being detected in nine nozzles, as documented in Table 44.2 butt welds, a leak was found from a pressurizer nozzle butt weld below.

at Tsuruga 2 in Japan and a part-depth crack was detected in a Through mid-2008, inspections of reactor vessel nozzle butt hot-leg pressurizer surge line nozzle butt weld at TMI-1. Both of welds have not yet been performed; hot leg nozzle inspections these cases occurred in 2003. Cracks were also detected in alloys under MRP-139 are slated to begin in Fall 2008. Given the above 82 and 182 cladding in steam generator heads that had been ham- pressurizer nozzle experience, it would not be surprising if at least mered and cold worked by a loose part [24]. some welds with PWSCC-like indications are discovered.

ASME_Ch44_p001-026.qxd 12/19/09 7:36 AM Page 9 COMPANION GUIDE TO THE ASME BOILER & PRESSURE VESSEL CODE

• 9 TABLE 44.2 CRACKING INDICATIONS DETECTED IN REACTOR COOLANT LOOP ALLOY 82/182 BUTT WELDS, 2005 THROUGH MID-2008 Inspection Type of Indication OD Indication a/ l/

Plant Date Nozzle Indication Depth (a, in) Length (l, in) thickness circumference Calvert Cliffs 2 2005 CL Drain Circ 0.056 0.628 10% 10%

Calvert Cliffs 2 2005 HL Drain Axial 0.392 0.000 70% 0%

DC Cook 2005 Safety Axial 1.232 0.000 88% 0%

Calvert Cliffs 1 2006 HL Drain Circ 0.100 0.450 19% 5%

Calvert Cliffs 1 2006 Relief Axial 0.100 0.000 8% 0%

Calvert Cliffs 1 2006 Surge Circ 0.400 2.400 25% 6%

Davis Besse 2006 CL Drain Axial 0.056 0.000 7% 0%

San Onofre 2 2006 Safety Axial 0.420 0.000 30% 0%

San Onofre 2 2006 Safety Axial 0.420 0.000 30% 0%

Wolf Creek 2006 Relief Circ 0.340 11.500 25.8% 46%

Wolf Creek 2006 Safety Circ 0.297 2.500 22.5% 10%

Wolf Creek 2006 Surge Circ 0.465 8.750 32.1% 19%

Farley 2 2007 Surge Circ 0.500 3.000 33% 6%

Davis Besse 2008 Axial Crystal River 3 2008 Circ 44.4.4 RPV Bottom-Head Penetrations PWSCC in BMI nozzles at South Texas 1 may be related to a com-In 2003, bare metal visual inspections of the reactor vessel bot- bination of high material susceptibility and welding flaws.

tom head at South Texas 1 showed small leaks from two BMI noz-zles, as shown in Fig. 44.11. These leaks were traced to PWSCC cracks in the nozzles that initiated at small regions of lack- 44.5 INSPECTION METHODS AND of-fusion in the J-groove welds between the nozzles and vessel REQUIREMENTS head [38]. The nozzles were repaired. Examinations of the other BMI nozzles at South Texas 1 showed no additional cracks. As a result of the increasing frequency of PWSCC cracks and Essentially all other U.S. plants have performed bare metal visual leaks identified in important PWR reactor vessel alloys 600, 82, inspections of RPV bottom-head nozzles without any evidence of and 182 materials since 2000, significant efforts are in progress by leaks. At least a dozen U.S. plants have completed volumetric the nuclear industry and the NRC to improve inspection capabilities examinations of the BMI nozzles, representing more than 20% of and develop appropriate long-term inspection requirements. The the total population of RPV bottom-head nozzles in the U.S., with following summarizes the status of inspection methods and require-no reported cracking. Similarly, no indications of in-service degra- ments as of May 2005. It is recommended that users check with the dation have been identified in volumetric inspections of RPV bot- NRC and industry programs to remain abreast of the latest changes tom-head nozzles performed in other countries. PWSCC of BMI in inspection methods and requirements.

nozzles that operate at the plant cold-leg temperature is generally considered to be less likely than PWSCC at locations operating at 44.5.1 Visual Inspections hot-leg or pressurizer temperatures. The earlier-than-expected Bare metal visual inspections have proven to be an effective way of detecting very small leaks, as shown by Figs. 44.8 and 44.11, and, therefore, should play an important role in any inspec-tion program. A key prerequisite for these inspections is that the surface should be free of preexisting boric acid deposits from other sources, because the presence of preexisting boric acid deposits can mask the small volumes of deposits shown in Figs. 44.8 and 44.11. Visual inspections with insulation in place can provide a useful backup to bare metal visual inspections but will be inca-pable of detecting small volumes of leakage, as shown in Figs.

44.8 and 44.11.

In many cases, it has been necessary to modify insulation pack-ages on the vessel top and bottom heads to facilitate performing bare metal visual inspections. As of May 2005, most of these modifications have been completed for PWR plants in the United States.

ASME Code Case N-722, Additional Examinations for PWR Pressure-Retaining Welds in Class 1 Components Fabricated with Alloys 600/82/182 Materials,Section XI, Division 1, was approved in 2005 to provide for increased visual inspections of FIG. 44.11 LEAK FROM SOUTH TEXAS 1 BMI NOZZLE potentially susceptible welds for boric acid leakage.

ASME_Ch44_p001-026.qxd 12/19/09 7:36 AM Page 10 10

• Chapter 44 44.5.2 Nondestructive Examinations Nozzle-to-safe end butt welds less than NPS 4 must be exam-Technology exists as of May 2005 to nondestructively examine ined by surface methods every inspection interval. Nozzle-all of the alloys 600, 82, and 182 locations in the reactor vessel. to-safe end butt welds NPS 4 and larger must be examined by Partial penetration nozzles (CRDM, CEDM, ICI) are typically volumetric and surface examination methods every inspection examined using one of two methods. The nozzle base metal can interval. Some deferrals of these inspections are permitted.

be examined volumetrically from the inside surface by ultrasonics (e) As of May 2005, the ASME Code did not require nonde-to confirm that the nozzle base material is free of internal axial or structive examination of the partial penetration welds for the circumferential cracks. Alternatively, the wetted surfaces of the CRDM and BMI nozzles. However, Code Case N-729-1 alloy 600 base metal and alloys 82 and 182 weld metal can be [63] was published later in 2005 that contained alternative examined by eddy current probes to ensure that there are no sur- examination requirements for PWR closure heads with noz-face cracks. If there are no surface cracks on wetted alloy 600 sur- zles having pressure-retaining partial-penetration welds.

faces, then it can be inferred that there will also be no internal This Code Case included visual, surface and volumetric cracks. Nozzles in the reactor vessel top head can be examined examinations for PWR closure heads with Alloy 600 noz-when the head is on the storage stand during refueling. Nozzles in zles and Alloy 82/182 partial-penetration welds at inspec-the reactor vessel bottom head can be examined ultrasonically or tion intervals that are based on the temperature dependence by eddy current when the lower internals are removed from the of the PWSCC phenomenon described in para. 44.3.4.

vessel during a 10-year in-service inspection outage. In some (Since RPV closure heads operate at varying temperatures, cases, the inside surfaces of BMI instrument nozzles can be there are significant head-to-head temperature differences examined by tooling inserted through holes in the lower internals. between plants.) Code Case N-729-1 also contains inspec-Reactor vessel inlet and outlet nozzle butt welds are normally tion requirements for PWR closure head with nozzles and inspected ultrasonically from the inside surface using automated partial-penetration welds of PWSCC resistant materials to equipment. These inspections are typically performed during address new and replacement heads.

10-year in-service inspection outages when the lower internals are (f) As noted in para. 44.5.1, Code Case N-722 [64] for visual removed from the reactor vessel. Eddy current methods are also inspections of alloys 82/182 welds was approved in 2005.

being used in some cases for examining the inside surfaces of (g) As of May 2008, the ASME Code is working on a new these welds for cracks, although eddy current inspection sensitivi- Section XI Code Case that contains alternate inspection ty is a function of the condition of the weld surface. For example, requirements Alloys 82/182 welds butt welds. ASME Code discontinuities in the weld profile can cause the eddy current actions are also in progress addressing various repair and probes to lift off of the surface being examined and, thereby, mitigation options for dealing with PWSCC. These are adversely affect the inspection sensitivity. discussed below in para. 44.9.

CRDM nozzle butt welds can be examined from the outside surface by standard ultrasonic methods.

A key to obtaining good nondestructive examinations is to have 44.5.4 NRC Inspection Requirements for RPV the process and the operators qualified on mockups containing Top-Head Nozzles prototypical axial and circumferential flaws. The EPRI NDE Subsequent to the discovery of significant corrosion to the Center in Charlotte, NC, is coordinating qualification efforts for Davis-Besse reactor vessel head, the NRC issued NRC Order inspection methods and inspectors in the United States. EA-03-009 [39]. This order specifies inspection requirements for RPV head nozzles based on the effective degradation years of 44.5.3 ASME BPVC Reactor Vessel Inspection operation. Effective degradation years (EDYs) are the effective Requirements full-power years (EFPYs) adjusted to a common 600F tempera-ASME BPVC Section XI specifies inservice inspection require- ture using an activation energy model. For plants with 600F head ments for operating nuclear power plants in the United States. temperatures, the EDYs are the same as the EFPYs. For plants Portions of these requirements that apply to PWSCC susceptible with head temperatures, greater than 600F, the EDYs are greater components in the RPV are summarized as follows: than the EFPYs. For plants with head temperatures less than 600F, the EDYs are less than the EFPYs. The NRC order (a) Table IWB-2500-1, Examination Category B-P, requires a specifies two types of inspections:

VT-2 visual examination of the reactor vessel pressure-retaining boundary during the system leak test after every (a) bare metal visual inspections of the RPV head surface refueling outage. No leakage is permitted. including 360 around each RPV head penetration nozzle (b) Table IWB-2500-1, Examination Category B-O, requires (b) nondestructive examinations of the RPV nozzles by one of that 10% of the CRDM nozzle-to-flange welds be inspected the two following methods:

by volumetric or surface methods each inspection interval.

(1) ultrasonic testing of each RPV head penetration nozzle (c) Table IWB-2500-1, Examination Category B-N-1, requires (i.e., base metal material) from 2 in. above the J-groove that attachment welds to the inside surface of the reactor weld to the bottom of the nozzle plus an assessment to vessel be examined visually each inspection interval. Welds determine if leakage has occurred through the interfer-in the beltline region must be inspected by VT-1 methods ence fit zone while welds outside the beltline region must be inspected by (2) eddy current testing or dye penetrant testing of the wetted VT-3 methods.

surface of each J-groove weld and RPV head penetration (d) Table IWB-2500-1, Examination Category B-F, specifies nozzle base material to at least 2 in. above the J-groove weld examination requirements for dissimilar metal welds in reactor vessels. Nozzle-to-safe end socket welds must be The first of the nondestructive examinations is to show that examined by surface methods every inspection interval. there are no axial or circumferential cracks in the nozzle base

ASME_Ch44_p001-026.qxd 12/19/09 7:36 AM Page 11 COMPANION GUIDE TO THE ASME BOILER & PRESSURE VESSEL CODE

• 11 metal or leak paths past the J-groove weld. The second of the categories requiring augmented inspection intervals and/or sample nondestructive examinations is to show that there are no axial or size. Category A is the lowest category, consisting of piping that circumferential cracks in the nozzle base metal by confirming the has been replaced (or originally fabricated) with PWSCC resistant absence of surface breaking indications on the nozzle and weld material. These weldments are to be inspected at their normal wetted surfaces. ASME Code frequency, as defined in ASME Section XI, Table The order specifies inspection intervals for three categories of IWB-2500-1. Category D refers to unmitigated PWSCC suscepti-plants: high susceptibility plants with greater than 12 EDY or ble weld in high temperature locations (e.g. pressurizer or hot leg where PWSCC cracks have already been detected, moderate sus- nozzles). These require an early initial inspection (before end of ceptibility plants less than or equal to 12 EDY and greater than or 2008 for pressurizer nozzles and before 2010 for hot leg nozzles),

equal to 8 EDY, and low susceptibility plants with less than 8 EDY. followed by more frequent inspections if they are not treated with As of June 2008, the U.S. NRC is expected shortly to transition some form of mitigation. Other categories (thru Category K) the requirements for inspection of RPV top-head nozzles based on address susceptible welds that have been mitigated (B and C),

NRC Order EA-03-009 [39] to a set based on ASME Code Case welds that have been inspected and found cracked, with or with-N-729-1 [63], with caveats. The inspection schedules in this code out mitigation, and welds for which geometric or material condi-case are generally based on the RIY (reinspection years) concept, tions limit volumetric inspectability. For the latter group, by the which normalizes operating time between inspections for the time the examination is due, plant owners are required to have a effect of head operating temperature using the thermal activation plan in place to address either the susceptibility of the weld or the energy appropriate to crack growth in thick-wall alloy 600 material inspectability of the weld.

(31 kcal/mol (130 kJ/mol)). The basis for this approach to nor- At the time of this writing, inspections are well under the malizing for the effect of head temperature is that the time for a MRP-139 guidelines are well underway in U.S. plants. Essentially flaw just below detectable size to grow to through-wall (and leak- all pressurizer nozzles have been inspected and or mitigated, and age) is dependent on the crack growth process. The requirements plans are in place to perform the other initial inspections required in ASME Code Case N-729-1 [63] were developed by ASME, by MRP-169. Plans include mitigation of most susceptible weld-with extensive technical input provided by a U.S. industry group ments in high temperature locations, thus moving the weldments (Materials Reliability Program) managed by EPRI [68]. into Categories A, B or C. Work is also currently underway to develop an ASME Section XI Code Case (N-790, alternative 44.5.5 NRC Inspection Requirements for examination requirements for PWSCC pressure-retaining butt RPV BMI Nozzles welds in PWRs) which will eventually replace MRP-139 and NRC Bulletin 2003-02, Leakage from Reactor Pressure Vessel place the augmented examination requirements for PWSCC sus-Lower Head Penetrations and Reactor Coolant Pressure ceptible butt welds back under the ASME Section XI Code.

Boundary Integrity [40], summarizes the leakage from BMI noz-zles at South Texas 1 and requires utilities to describe the results of BMI nozzle inspections that have been performed at their 44.6 SAFETY CONSIDERATIONS plants in the past and that will be performed during the next and following refueling outages. If it is not possible to perform bare 44.6.1 Small Leaks metal visual examinations, utilities should describe actions that Small leaks due to axial cracks such as shown in Figs. 44.8 and are being made to allow bare metal visual inspections during sub- 44.11 do not pose significant safety risk. The leak rates are low sequent outages. If no plans are being made for bare metal visual enough that the leaking primary coolant water will quickly evapo-or nonvisual surface or volumetric examinations, then utilities rate leaving behind a residue of dry boric acid. Most of the leaks must provide the bases for concluding that the inspections that detected to date have resulted in these relatively benign condi-have been performed will ensure that applicable regulatory tions. As shown in the figures, very small leaks are easily detected requirements are met and will continue to be met. On September by visual inspections of the bare metal surfaces provided that the 5, 2003, the NRC issued Temporary Instruction 2515/152 [41], surfaces are free from boric acid deposits from other sources. One which provides guidance for NRC staff in reviewing utility sub- explanation for the extremely low leak rates is that short tight mittals relative to Bulletin 2003-02. While the Temporary PWSCC cracks can become plugged with crud in the primary Instruction does not represent NRC requirements, it does indicate coolant, thereby preventing leakage under normal operating con-the type of information that the NRC is expecting to receive in ditions. It is hypothesized that distortions, which occur during response to the bulletin. plant transients, allow small amounts of leakage through the crack before it becomes plugged again. Regardless, these small leaks do 44.5.6 Industry Inspection Requirements for not pose a significant safety concern.

Dissimilar Metal Butt Welds The industry in the United States has developed a set of manda- 44.6.2 Rupture of Critical Size Flaws tory inspection guidelines for PWSCC susceptible. Alloy 82/182 Initially, leaking RPV top-head nozzles were thought to be butt welds, which are documented in the report MRP-139 [58]. exclusively the result of axial cracks in the nozzles, and it was MRP-139 defines examination requirements in terms of categories thus believed that they did not represent a significant safety con-of weldments that are based on 1) the IGSCC resistance of the cern. However, as more examinations were performed, findings materials in the original weldment, 2) whether or not mitigation arose that called this hypothesis into question.

has been performed on the original weldment, 3) whether or not a pre-mitigation UT examination has been performed, 4) the exis- (a) Relatively long circumferential cracks were observed in two tence (or not) of cracking in the original weldment, and 5) the nozzles in the Oconee Unit 2 RPV head, and several other likelihood of undetected cracking in the original weldment. The plants also discovered shorter circumferentially oriented categories range from A through K, with the higher letter cracks.

ASME_Ch44_p001-026.qxd 12/19/09 7:36 AM Page 12 12

• Chapter 44 Because of the concern for PWSCC in PWR piping dissimilar metal butt welds, methods for predicting the critical crack size for rupture in such welds have received recent attention [59]. Axial PWSCC flaws in these welds are limited to the width of the alloy 82/182/132 weld material. Experience has confirmed that the PWSCC cracks arrest when they reach the PWSCC-resistant low-alloy steel and stainless steel materials [50]. Therefore, the maxi-mum axial crack lengths are limited to a few inches at most (much less than the critical axial flaw length), except for the small number of cases involving alloy 600 safe ends or alloy 600 pipe/tube (CRDM and BMI nozzles), where axial cracks initiating in the weld could potentially propagate into the alloy 600 base metal. Thus, critical crack size calculations for PWR piping dis-similar metal butt welds typically assume one or more circumfer-entially oriented PWSCC flaws.

In 2007, EPRI sponsored a detailed investigation of the growth of circumferential PWSCC flaws in PWR pressurizer nozzle dis-similar metal butt welds [59]. Using finite-element methods, this study examined the effect of an arbitrary crack profile on crack FIG. 44.12 SCHEMATIC OF RPV TOP-HEAD NOZZLE growth and subsequent crack stability and leak rate versus the GEOMETRY AND NATURE OF OBSERVED CRACKING standard assumption of a semi-elliptical crack profile. The crack stability (i.e., critical crack size) modeling of the EPRI study was (b) Circumferential NDE indications were discovered in the based on a standard limit load (i.e., net section collapse)

North Anna Unit 2 head in nozzles that showed no apparent approach as applied to an arbitrary crack profile around the weld signs of boric acid deposits due to leakage. circumference [65]. The potential for an EPFM failure mode was considered using a Z-factor approach specific to piping dissimilar Figure 44.12 presents a schematic of a top-head CRDM nozzle metal welds [66]. Finally, the role of secondary piping thermal and J-groove weld and the nature of the cracking that has been constraint stresses in the rupture process was investigated on the observed. There is some uncertainty as to whether circumferential basis of available experimental pipe rupture data [67], elastic-cracks arise as a result of axial cracks growing through the weld plastic finite-element analyses of a pipe with an idealized or nozzle and causing leakage into the annular region between the through-thickness crack [59], and pressurizer surge line piping nozzle and head, ultimately leading to reinitiation of circumferen- models applied to evaluate the maximum capacity of the tial cracking on the outside surface of the tube, or if they are due secondary loads to produce rotation at a cracked pressurizer to the axial cracks branching and reorienting themselves in a surge nozzle [59].

circumferential direction, as depicted on the right-hand side of Fig. 44.12. A destructive examination program has been per- 44.6.3 Boric Acid Wastage Due to Larger Leaks formed on several of the North Anna Unit 2 nozzles, indicating Small concentrations of boron are added to the primary coolant that the circumferential nozzle defects found there were in fact the water in PWR plants in the form of boric acid to aid in controlling result of grinding during fabrication rather than service-related core reactivity. At the start of an operating cycle with new fuel, cracking. Nevertheless, the occurrence of circumferential crack- the boron concentration is typically about 2,000 ppm or less. The ing adds a new safety perspective to the RPV top-head nozzle concentration of boron is reduced with fuel burnup to about cracking problem, because of the potential for such cracks to 0 ppm at the end of an operating cycle when fuel is ready to be grow to a critical length and ultimately lead to ejection of a nozzle replaced. Work by EPRI and others to determine the probable rate from the vessel, although a large circumferential flaw covering of corrosion of low-alloy steel by boric acid is documented in the more than 90% of the wall cross section is typically calculated for EPRI Boric Acid Corrosion Guidebook [43]. This document nozzle ejection to occur because of the relatively thick wall typical shows that the corrosion rate of low-alloy steel by deareated pri-of RPV top-head nozzles. mary coolant (inside the pressure vessel and piping) with 2,000 PWSCC in PWR RPV inlet/outlet nozzles could also potentially ppm boron is negligible. The corrosion rate for low concentration develop circumferentially oriented flaws, which could lead to pipe (2,000 ppm) aerated boric acid is also very low. However, when rupture. To date, observed cracking has been primarily axial with high-temperature borated water leaks onto a hot surface, the water only very small circumferential components. With time, however, can boil off leaving concentrated aerated boric acid. The corro-PWSCC in large piping butt welds might be expected to follow sion rate of low-alloy steel by hot concentrated aerated boric acid trends similar to the IGSCC cracking issue in BWRs [42]. In the can be as high as 10 in./year under some conditions.

BWR case, cracking and leakage were initially seen only as axial- As evidenced by the significant volume of material corroded ly oriented cracks in smaller diameter piping. With time, however, from the Davis-Besse reactor vessel head, boric acid corrosion axial and circumferential cracking were observed in pipe sizes up has the potential to create significant safety risk. Figure 44.13 to and including the largest diameter pipes in the system. shows cross-section and plan views of the corroded region of the Considering the potential existence of weld repairs during initial Davis-Besse head shown in Fig. 44.9. As indicated, a large vol-construction of the plants and the associated high residual stresses ume of the low-alloy head material was corroded, leaving the that they produce in both axial and circumferential directions, stainless steel cladding on the inside of the vessel head to resist significant circumferential cracking may eventually be observed the internal pressure. Part-depth cracks were discovered in the in large-diameter PWR pipe-to-nozzle butt welds. unsupported section of cladding.

ASME_Ch44_p001-026.qxd 12/19/09 7:37 AM Page 13 COMPANION GUIDE TO THE ASME BOILER & PRESSURE VESSEL CODE

• 13 FIG. 44.13 PLAN AND CROSS-SECTION THROUGH CORRODED PART OF DAVIS-BESSE REACTOR VESSEL HEAD Based on available evidence, it was determined that the leakage of the nozzles frequently enough to catch PWSCC cracks before that caused the corrosion had been occurring for at least 6 years. they grow through wall. Secondly, clean the external surfaces of While it was known that boric acid deposits were accumulating preexisting boric acid deposits from other sources and perform bare on the vessel top head surface, the utility attributed the accumula- metal visual inspections at frequent enough intervals to detect leaks tions to leakage from spiral-wound gaskets at the flanged joints at an early benign stage. Thirdly, if the risk is believed high or between the CRDM nozzles and the CRDMs. The accumulations of boric acid had not been removed due to poor access to the enclosed plenum between the top of the vessel head and the bot-tom of the insulation, as shown in Fig. 44.14.

The transition from relatively benign conditions, such as shown in Figs. 44.8 and 44.11, to severe conditions, which created the cav-ity shown in Figs. 44.9 and 44.13, is believed to be a function of the leakage rate. A PWSCC crack that first breaks through the nozzle wall or weld will initially be small (short), resulting in a low leak rate. It is believed that the small leak rate will not lower the metal surface temperature enough to allow liquid conditions to exist. As the crack grows in length above the J-groove weld, the leak rate is expected to increase to the point where boric acid on the surface near the leak remains moist or where the leaking borated water locally cools the low-alloy steel material to the point where the sur-face will remain wetted and allow boric acid to concentrate.

Preliminary models of these conditions have been developed, and test work was started by EPRI in 2004 to more accurately deter-mine the conditions where the leakage produces wetted conditions that can cause high boric acid corrosion rates and where the leakage results in essentially benign dry boric acid deposits.

Conditions such as occurred at Davis-Besse can be prevented by FIG. 44.14 CROSS-SECTION THROUGH DAVIS-BESSE a three-step approach. Firstly, perform nondestructive examinations REACTOR VESSEL HEAD

ASME_Ch44_p001-026.qxd 12/19/09 7:37 AM Page 14 14

• Chapter 44 inspections are difficult or costly, replace the susceptible parts or R  universal gas constant apply a remedial measure to reduce the risk of PWSCC leaks.  8.314  10-3 kJ/mole

• K (1.103  10-3 kcal/mole

• R)

T  absolute operating temperature at location of crack, K (or R) 44.7 DEGRADATION PREDICTIONS Tref  absolute reference temperature used to normalize data

 325C  598.15 K (617F  1076.67 R) 44.7.1 Crack Initiation   crack growth amplitude Initiation of PWSCC in laboratory test samples and in PWR K  crack tip stress intensity factor, Mpam (or ksiin) steam generator tubing has been found to follow standard statisti- Kth  crack tip stress intensity factor threshold cal distributions such as Weibull and log-normal distributions  9 Mpam (8.19 ksiin)

[44-47]. These distributions have been widely used for modeling

  exponent and predicting the occurrence of PWSCC in PWRs since about

 1.16 1988, and continue to be used for this purpose.

The parameters of a statistical distribution used to model a Temperature dependence is modeled in this crack growth rate given mode of PWSCC, such as axial cracks in CRDM nozzles, equation via an Arrhenius-type relationship using the aforemen-only apply to the homogeneous set of similar items that are tioned activation energy of 31 kcal/mole. The stress intensity exposed to the same environmental and stress conditions, and factor dependence is of power law form with exponent 1.16.

only to the given crack orientation being modeled. For example, Figure 44.15 presents the distribution of the coefficient () in the axial and circumferential cracking are modeled separately since power law relationship at constant temperature (617F). The data the stresses acting on the two crack orientations are different. in this figure exhibit considerable scatter, with the highest and In general, two parameter Weibull or log-normal models are used lowest data points deviating by more than an order of magnitude to model and predict the future occurrence of PWSCC. An initia- from the mean. The 75th percentile curve (see Figure 44.15a) was tion time, which sometimes is used as a third parameter, is not gen- recommended for use in deterministic crack growth analyses erally modeled, because use of a third parameter has been found to [26,48], and this curve is now included in Section XI for disposi-result in too much flexibility and uncertainty in the predictions. tion of PWSCC flaws in RPV top-head nozzles. In addition, prob-PWSCC predictions are most reliable when the mode of crack- abilistic crack growth rate studies have been performed of top ing is well developed with results for detected cracking available head nozzles using the complete distribution [49]. An additional for three or more inspections. In this situation, the fitted parameters factor of 2 has been applied to the 75th percentile value when to the inspection data are used to project into the future. When no analyzing crack growth exposed to leakage in the annular gap cracking has been detected in a plant, rough predictions can still be between the nozzle and the head, to allow for possible abnormal developed using industry data. This is generally done using a two- water chemistry conditions that might exist there [26,48].

step process. The first step involves developing a statistical distribu- Similar crack growth rate testing has been conducted for tion of times to occurrence of PWSCC at a selected threshold level alloys 82 and 182 weld metals. The weld metal crack growth (such as 0.1%) for a set of plants with similar designs. Data for data are sparser and exhibit similar statistical variability. A plants with different temperatures are adjusted to a common tem- review of weld metal PWSCC crack growth data has also been perature using the Arrhenius equation (see Table 44.1). The distrib- completed under EPRI sponsorship [61,62]. This study (MRP-ution of times to the threshold level is used to determine a best esti- 115) showed that Alloy 182/132 weld metal crack growth obeys mate time for the plant being modeled to develop PWSCC at that a similar relationship to that shown above for alloy 600 base threshold level. Techniques are available to adjust the prediction to metal, but with crack growth rates about four times higher than account for the time already passed at the plant without detecting the alloy 600 curve for stress intensity factors greater than about the mode being evaluated. Once the best estimate time for occur- 20 ksiin (see Figure 44.15a). Similar to the heat-by-heat analy-rence at the threshold level is determined, future cracking is pro- sis for the wrought material, a weld-by-weld analysis was per-jected from that point forward using the median rate of increase formed on the available worldwide laboratory crack growth rate (Weibull slope or log-normal standard deviation) in the industry for data for the weld materials (see Figure 44.15b). The EPRI study the mode of PWSCC being evaluated. (MRP-115) concluded that PWSCC crack growth rates for alloy 82/182/132 weld metal behave in accordance with the following 44.7.2 Crack Growth relationship, where no credit for a stress intensity factor thresh-Numerous PWSCC crack growth studies have been performed old greater than zero was taken because of insufficient data on on thick-wall alloy 600 material in PWR environments at test tem- this parameter:

peratures that span the range of typical PWR operating tempera-tures. In 2002, these tests were reviewed and summarized under Qg 1 a = exp c- a - b da falloy forient K b sponsorship of EPRI [26,48]. The EPRI study (MRP-55) conclud- . 1 ed that PWSCC crack growth rates for thick-wall alloy 600 base R T Tref metal behave in accordance with the following relationship:

where:

Qg 1 .

a  crack growth rate at temperature T in m/s (or in/h) a = exp c- a - b da(K - K th)b

. 1 R T Tref Qg  thermal activation energy for crack growth where  130 kJ/mole (31.0 kcal/mole)

. R  universal gas constant a  crack growth rate at temperature T in m/sec (or in./hr)  8.314  10-3 kJ/mole-K (1.103  10-3 kcal/mole-°R)

Qg  thermal activation energy for crack growth T  absolute operating temperature at location of crack, K

 130 kJ/mole (31.0 kcal/mole) (or °R)

ASME_Ch44_p001-026.qxd 12/19/09 7:37 AM Page 15 COMPANION GUIDE TO THE ASME BOILER & PRESSURE VESSEL CODE

• 15 FIGURE 44.15A DETERMINISTIC CRACK GROWTH RATE CURVES FOR THICK-WALL ALLOY 600 WROUGHT MATERIAL AND FOR ALLOY 182/132 AND ALLOY 82 WELD MATERIALS [61,62]

FIGURE 44.15B LOG-NORMAL FIT TO 19 WELD FACTORS FOR SCREENED MRP DATABASE OF CGR DATA FOR ALLOY 82/182/132 [61,62]

Tref  absolute reference temperature used to normalize data then inserted into the appropriate crack growth relationship (alloy

 598.15 K (1076.67°R) 600, 82, or 182) at the component operating temperature and inte-

  power-law constant grated with time to predict crack size versus operating time at the

 1.5  10-12 at 325°C for a in units of m/s and K in applicable temperature.

. Figure 44.16 shows typical crack growth predictions for a cir-units of MPa m (2.47  10-7 at 617°F for a in units cumferential crack in a steep angle RPV top-head (CRDM) noz-of in/h and K in units of ksi in) zle. (Nozzles in the outer rings of vessel heads having the steepest falloy  1.0 for Alloy 182 or 132 and 1/2.6  0.385 for Alloy 82 angles between the nozzle and the head have been found to be forient  1.0 except 0.5 for crack propagation that is clearly controlling in terms of predicted growth rates for circumferential perpendicular to the dendrite solidification direction cracks). The analysis depicted in Fig. 44.16 assumed a through-K  crack-tip stress intensity factor, MPam (or ksiin) wall, 30 of circumference crack in the most limiting azimuthal

  exponent location of the nozzle at time zero, and predicted the operating time

 1.6 for it to grow to a size that would violate ASME Section XI flaw evaluation margins with respect to nozzle ejection (~300). It is Deterministic crack growth rate predictions have been per- seen that, even for relatively high RPV temperatures, operating formed for axial and circumferential cracking in RPV top- and times on the order of 8 years or greater are predicted for circumfer-bottom-head nozzles and in large-diameter butt welds [49,50]. ential nozzle cracks to propagate to a size that would violate Welding residual stresses are a primary factor contributing to ASME Section XI safety margins.

crack growth in all these analyses. Stress intensity factors versus Figure 44.17 shows similar crack growth predictions for a crack size, considering residual stresses plus operating pressure postulated circumferential crack in a large-diameter nozzle butt and thermal stresses are first computed in these studies. These are weld. Stress intensity factors were computed in this analysis for

ASME_Ch44_p001-026.qxd 12/19/09 7:37 AM Page 16 16

• Chapter 44 FIG. 44.16 CRACK GROWTH RATE PREDICTIONS FOR CIRCUMFER-ENTIAL CRACKS IN RPV TOP-HEAD NOZZLE AT VARIOUS ASSUMED OPERATING TEMPERATURES INITIAL CRACK ASSUMPTION 30 THROUGH-WALL CRACK AT MAXIMUM STRESS AZIMUTH IN HIGH ANGLE NOZZLE.

a 6-to-1 aspect ratio crack in a large-diameter RPV inlet/outlet repair were assumed, little or no crack growth would be predict-nozzle, ranging in depths from 0.1 in. to 2.2 in. The nozzle was ed over the plant lifetime. For this same crack, including the conservatively assumed to have a large, inside surface repair, effect of the repair, the predicted time for a 0.1 in. deep crack to and the crack was assumed to reside in the center of that repair grow to 75% through-wall at a typical inlet nozzle temperature (i.e., in the most unfavorable residual stress region of the weld). (555F) is about 11 years.

The predicted crack growth in this case is fairly rapid for a typi- The strong effect of operating temperature is apparent in both cal outlet nozzle temperature, 602F, propagating to 75% crack growth analyses. The outlet nozzle analysis also demon-through-wall (the upper bound of ASME Section XI allowable strates the detrimental effect of weld repairs that were performed flaw sizes in piping) in about 3 years. Conversely, if no weld during construction at some plants.

FIG. 44.17 CRACK GROWTH RATE PREDICTIONS FOR CIRCUMFERENTIAL CRACKS IN RPV MAIN COOLANT LOOP DISSIMILAR METAL NOZZLE BUTT WELD AT OPERATING TEMPERATURES TYPICAL OF REACTOR INLET AND OUTLET NOZZLES INITIAL CRACK ASSUMPTION 0.1 0.6 INSIDE SURFACE CRACK AT MAXIMUM STRESS AZIMUTH IN NOZZLE WITH ASSUMED INSIDE SURFACE FIELD REPAIR.

ASME_Ch44_p001-026.qxd 12/19/09 7:37 AM Page 17 COMPANION GUIDE TO THE ASME BOILER & PRESSURE VESSEL CODE

• 17 FIG. 44.18 PROBABILITY OF NOZZLE FAILURE (NSC) AS A FUNCTION OF VARIATIONS IN TOP-HEAD TEMPERATURE AND INSPECTION INTERVALS 44.7.3 Probabilistic Analysis (e) modeling of the effects of inspections, including inspection Because of the large degree of statistical scatter in both the type, frequency, and effectiveness crack initiation and crack growth behavior of PWSCC in alloy A series of PFM analysis results is presented in [49], which cov-600 base metal and associated weld metals, probabilistic fracture ers a wide variety of conditions and assumptions. These include mechanics (PFM) analyses have been used to characterize the base cases, with and without inspections, and sensitivity studies to phenomenon in terms of the probabilities of leakage and failure evaluate the effects of various statistical and deterministic assump-

[49] for RPV top head nozzles. The analysis incorporates the fol-tions. The model was benchmarked with respect to field experience, lowing major elements:

considering the occurrence of cracking and leakage and of circum-(a) computation of applied stress intensity factors for circum- ferential cracks of various sizes. The benchmarked parameters were ferential cracks in various nozzle geometries as a function then used to evaluate the effects of various assumed inspection pro-of crack length and stresses grams on probability of nozzle failure and leakage in actual plants.

(b) determination of critical circumferential flaw sizes for noz- A sample of the results is presented in Figs. 44.18 and 44.19.

zle failure Figure 44.18 shows the effect of inspections on probability of (c) an empirical (Weibull) analysis of the probability of nozzle nozzle failure (Net Section Collapse, or ejection of a nozzle) for cracking or leakage as a function of operating time and tem- head operating temperatures ranging from 580F to 600F. A no-perature of the RPV head inspection curve is shown for each temperature. Runs were then (d) statistical analysis of PWSCC crack growth rates in the made assuming NDE inspections of the nozzles. Inspections were PWR primary water environment as a function of applied assumed to be performed at intervals related to head operating tem-stress intensity factor and service temperature perature (more frequent inspections for higher head temperatures, FIG. 44.19 PROBABILITY OF NOZZLE LEAKAGE AS A FUNCTION OF VARIATIONS IN TOP-HEAD TEMPERATURE AND INSPECTION INTERVALS

ASME_Ch44_p001-026.qxd 12/19/09 7:37 AM Page 18 18

• Chapter 44 FIGURE 44.19A PRESSURIZER DISSIMILAR METAL BUTT WELD FLAW INDICATIONS COMPARED TO CRITICAL FLAW SIZE PROBABILITY ESTIMATES less frequent for lower temperatures). It is seen from the figure seen from this figure that all of the flaw indications detected were that the assumed inspection regimen is sufficient to maintain the far short of the sizes needed to cause a rupture. The probabilistic nozzle failure probability (per plant year) below a generally analysis also addressed the small but finite probability that larger accepted target value of 1  103 for loss of coolant accidents flaws may exist in uninspected nozzles, plus the potential for crack due to nozzle ejection. growth during future operating time until all the nozzles are Figure 44.19 shows similar results for the probability of leak- inspected (or mitigated) under MRP-139 [58] guidelines.

age from a top-head nozzle. It is seen from this figure that the same assumed inspection regimen maintains the probability of leakage at or about 6% for the cases analyzed. Analyses similar to 44.8 REPAIRS those reported in Figs. 44.18 and 44.19 have been used, in conjunc-tion with deterministic analyses, to define an industry-recommended When cracking or leakage is detected in operating nuclear inspection and corrective action program for PWR top heads that power plant pressure boundary components, including the reactor results in acceptable probabilities of leakage and failure. This vessel, repair or replacement may be performed in accordance work also constituted the basis for the inspection requirements with ASME BPVC Section XI [51].Section XI specifies that the incorporated in ASME Code Case N-729-1 [63]. flaws must be removed or reduced to an acceptable size in accor-Similar probabilistic analyses have been performed for PWSCC dance with Code-accepted procedures. For PWSCC in RPV alloy susceptible butt welds in pressurizer nozzles, as part of the effort 600 components, several approaches have been used.

documented in MRP-216 [59]. Analyses established the current expected flaw distribution based on pressurizer nozzle DMW 44.8.1 Flaw Removal inspections to date, (Table 44.1), estimates were made of the prob- For relatively shallow or minor cracking, flaws may be ability of cracking versus flaw size, and of crack growth rate ver- removed by minor machining or grinding. This approach is per-sus time. A plot of the flaw indications found to date, in terms of mitted by the ASME Code to eliminate flaws and return the com-crack length as percentage of circumference (abscissa) versus ponent to ASME Code compliance. However, this approach gen-crack depth as percentage of wall thickness (ordinate) is illustrated erally does not eliminate the underlying cause of the cracking.

in Figure 44.19a. Axial indications plot along the vertical axis There will still be susceptible material exposed to the PWR envi-(l/circumference = 0) in this plot, with leaking flaws plotted at a/t ronment that caused the cracking originally, and in some cases the

= 100%. Circumferential indications plot at non-zero values of susceptibility might be aggravated by surface residual stresses l/circumference, at the appropriate a/t. Clean inspections are plot- caused by the machining or grinding process. Simple flaw ted randomly in a 10% box near the origin, to give some indication removal is thus not considered to be a long-term repair, unless of inspection uncertainty. Also shown on this plot are loci of criti- combined with some other form of mitigation. However, in the cal flaw sizes based on an evaluation of critical flaw sizes present- short term, for example, where future component replacement is ed in Ref. [59]. 50th and 99.9th percentile plots are shown. It is planned, it may be a viable approach for interim operation.

ASME_Ch44_p001-026.qxd 12/19/09 7:37 AM Page 19 COMPANION GUIDE TO THE ASME BOILER & PRESSURE VESSEL CODE

• 19 FIG. 44.21 SCHEMATIC OF WELD OVERLAY REPAIR APPLIED TO RPV OUTLET NOZZLE problem. Although WOLs, shown in Fig. 44.21, do not eliminate the PWSCC environment from the flaw as in the flaw embedment process, the repair has been shown to offer multiple improve-ments to the original pipe welds, including the following:

(a) structural reinforcement FIG. 44.20 SCHEMATIC OF RPV TOP-HEAD NOZZLE (b) resistant material FLAW EMBEDMENT REPAIR (c) favorable residual stress reversal Weld overlays also offer a significant improvement in inspec-44.8.2 Flaw Embedment tion capability, because once a weld overlay is applied, the Surface flaws are much more significant than embedded flaws required inspection coverage reduces to just the weld overlay from a PWSCC perspective, because they continue to be exposed material plus the outer 25% of the original pipe wall, often a to the PWR primary water environment that caused the crack and much easier inspection than the original dissimilar metal weld that can lead to continued PWSCC flaw growth after initiation. (DMW) inspection.

Accordingly, one form of repair is to embed the flaw under a Weld overlay repairs have been recognized as a Code-accept-PWSCC-resistant material. Figure 44.20 shows an embedment able repair in an ASME Section XI Code Case [52] and accepted approach used by one vendor to repair PWSCC cracks or leaks in by the U.S. NRC as a long-term repair. They have also been used, top-head nozzles and welds. The PWSCC-susceptible material, albeit less extensively, to repair dissimilar metal welds at nozzles shown as the cross-hatched region in the figure, is assumed to be in BWRs.

entirely cracked (or just about to crack). PWSCC-resistant material, The weld overlay repair process was first applied to a PWR typically alloy 52 weld metal, is deposited over the susceptible large-diameter pipe weld (on the Three Mile Island 1 pressurizer material. The assumed crack is shown to satisfy all ASME BPVC to hot-leg nozzle) in the fall of 2003. Since that time, as part of Section XI flaw evaluation requirements, in the absence of any the MRP-139 inspection effort described in para. 44.5.6, over 200 growth due to PWSCC, since the crack is completely isolated weld overlays have been applied to pressurizer nozzle dissimilar from the PWR environment by the resistant material. Note that metal butt welds. Part of the reason for this trend is that many the resistant material in this repair must overlap the susceptible pressurizer nozzles were unable to be volumetrically inspected to material by enough length in all directions to preclude new cracks achieve the required examination coverage in their original con-growing around the repair and causing the original crack to be figuration. By applying weld overlays, in addition to mitigating reexposed to the PWR environment. Although this repair the welds, their inspectability was enhanced such that post over-approach has been used successfully in several plants, there have lay ultrasonic exams could be performed in accordance with been many incidents in which nozzles repaired by this approach applicable requirements. Technical justification for the WOL during one refueling outage have been found to be leaking at the process as a long-term repair is documented in Ref. [53].

subsequent outage. These occurrences were attributed to lack of Requirements for weld overlays in PWR systems, including their sufficient overlap of the repair, because it is sometimes difficult to use as mitigation as well as repair, is documented in Ref. [60].

distinguish the exact point at which the susceptible material ends (for instance the end of the J-groove weld butter and the begin- 44.8.4 Weld Replacement ning of the RPV cladding in Fig. 44.20). Finally, the flawed weld may be replaced in its entirety. In PWR top-head nozzles, this process has been implemented extensively by 44.8.3 Weld Overlay relocating the pressure boundary from the original PWSCC-Another form of repair that has been used extensively to repair susceptible J-groove weld at the inside surface to a new weld at the cracked and leaking pipe welds is the weld overlay (WOL). midwall of the RPV head (see Fig. 44.22). With this repair Illustrated schematically in Fig. 44.21, WOLs were first con- approach, the PWSCC-susceptible portion of the original J-groove ceived in the early 1970s as a repair for IGSCC cracking and weld and buttering is left in the vessel, but it is no longer part of leakage in BWR main coolant piping. Over 500 such repairs have the pressure-retaining load path for the nozzle. The lower portion of been applied in BWR piping ranging from 4 in. to 28 in. in diam- the original nozzle is first removed by machining to a horizontal ele-eter, and some weld overlay repairs have been in service for over vation above the J-groove weld (left-hand side of Fig. 44.22). A 20 years, with no evidence of any resumption of the IGSCC weld prep is produced on the bottom of the remaining portion of

ASME_Ch44_p001-026.qxd 12/19/09 7:37 AM Page 20 20

• Chapter 44 FIG. 44.22 SCHEMATIC OF RPV TOP-HEAD NOZZLE WELD REPLACEMENT REPAIR the nozzle, and a new, horizontal weld is made between the original (a) Zinc Additions to Reactor Coolant. Laboratory tests indicate nozzle and the bore of the RPV head (righthand side of Fig. 44.22). that the addition of zinc to reactor coolant significantly slows The new weld is made with PWSCC-resistant material (generally down the rate of PWSCC initiation, with the improvement alloy 52 weld metal), and the surface of the weld is machined for factor increasing as the zinc concentration increases [29].

NDE. The repair process still leaves some portion of the original The improvement factor (slowdown in rate of new crack ini-PWSCC-susceptible alloy 600 nozzle in place, potentially in a high tiation) shown by tests varies from a factor of two for 20 ppb residual stress region at the interface with the new weld. However, a zinc in the coolant to over a factor of ten for 120 ppb zinc.

surface treatment process, such as roll peening or burnishing, has The effect of zinc on crack growth rate is not as certain, with been applied to this interface in many applications to reduce poten- some tests indicating a significant reduction in crack growth tial PWSCC concerns. Experience with this repair process has been rate but others indicating no change. Further testing is under-good, in terms of subsequent leakage from repaired nozzles, and in way under EPRI sponsorship (as of 2004) to clarify the most cases the repair need only survive for one or two fuel cycles, effects of zinc on crack growth rate. As of mid-2004, evalu-because, once PWSCC leakage is detected in an RPV head, com- ation of plant data, especially the data for a two-unit station mon industry practice has been to schedule a future head replace- with PWSCC at dented steam generator tube support plates, ment (not because of the repaired nozzle, but because of concerns is encouraging but not conclusive with regard to whether use that other nozzles are likely to be affected by the problem leading to of zinc is reducing the rate of PWSCC. The uncertainty is the costly future inspections, repairs, and outage extensions). result of changes in inspection methods simultaneously with changes in zinc concentration.

(b) Adjustments of Hydrogen Concentration. The EPRI PWR 44.9 REMEDIAL MEASURES Primary Water Chemistry Guidelines require the hydrogen concentration in the primary coolant to be kept between 25 44.9.1 Water Chemistry Changes and 50 cc/kg [28]. As discussed in the Guidelines and sum-Three types of water chemistry changes that could affect the marized above in para. 44.3.4, the rate of PWSCC initiation rate of PWSCC are zinc additions to the reactor coolant, adjust- and rate of PWSCC crack growth both seem to be affected ments to hydrogen concentration, and adjustments to lithium by the hydrogen concentration, with lower concentrations concentration and pH. The factors are described below. being more aggressive at lower temperature and higher

ASME_Ch44_p001-026.qxd 12/19/09 7:37 AM Page 21 COMPANION GUIDE TO THE ASME BOILER & PRESSURE VESSEL CODE

• 21 concentrations at higher temperature. Depending on the not all of the specimens were fabricated from the same heat of plant situation as far as which parts are at most risk of material. Therefore, there were differences in material PWSCC PWSCC, and depending on the temperature at those parts, susceptibility in addition to differences in remedial measure effec-there may be some benefit, such as an improvement factor tiveness. The methods used to correct for differences in specimen of about 1.2, in operating at hydrogen concentrations at PWSCC susceptibility are discussed in the paper.

either end of the allowed range. In the longer term, The remedial measures fell into three main effectiveness groups.

increased benefit may be achieved by operating slightly outside of the allowed range (e.g., at 60 cc/kg), although (a) most effective this will require confirmation that the change does not (1) waterjet conditioning result in some other undesirable effects. (2) electro mechanical nickel brush plating (c) Adjustments of Lithium Concentration and pH. As dis- (3) shot peening cussed in para. 44.3.4, some tests indicate that the rate of PWSCC initiation is increased by increases in lithium con- (b) intermediate effectiveness centration and pH, e.g., by factors ranging from about 1.15 (1) electroless nickel plating to 2.0. On the other hand, increases in lithium and pH pro- (2) GTAW weld repair vide proven benefits for reducing the potential harmful (3) laser weld repair deposit buildup on fuel cladding surfaces and for reducing shutdown dose rates [28]. Based on these opposing trends, (c) least effective plants can select a lithium/pH regime that best suits their (1) EDM skim cutting needs, i.e., does not involve substantial risks of aggravating (2) laser cladding PWSCC, while still providing benefits for reducing fuel (3) flapper wheel surface polishing deposits and shutdown dose rates. When evaluating the pos-sible risks to PWSCC of increasing lithium and pH, it As of May 2005, it is not believed that any of these remedial should be noted that crack growth rate tests show no harm- measures had actually been applied to a reactor vessel in the field.

ful effect while crack initiation tests do. The data from crack growth rate tests are considered to be more reliable, and it is 44.9.4 Stress Improvement recommended that they be given greater weight than the To mitigate against the IGSCC problem in BWR piping, many results from crack initiation tests. An additional considera- plants implemented residual stress improvement processes. These tion is that the use of zinc can provide a stronger benefit were performed both thermally (induction heating stress improve-than the possible deficit associated with increases in lithium ment or IHSI) and by mechanical means (mechanical stress and pH, and, thus, can make use of a combined zinc adjust- improvement process or MSIP). As described above, residual ment and increase in lithium and pH attractive. stresses play a major role in susceptibility to both IGSCC and PWSCC, because large piping butt welds tend to leave significant 44.9.2 Temperature Reduction residual stresses at the inside surfaces of the pipes, especially when field repairs were performed during construction. Both To date, a main remedial measure applied in the field for RPV stress improvement processes have been demonstrated to reverse top-head PWSCC has been modification of the reactor internals the unfavorable residual stresses, leaving compressive stresses on package to increase bypass flow through the internals flange the inside surface of the pipe, which is exposed to the reactor region and, thereby, reduce the head temperature. The lower head environment. MSIP has also been applied to PWSCC-susceptible temperature is predicted to reduce the rates of crack initiation and butt welds in PWR piping, primarily dissimilar metal welds at growth based on the thermal activation energy model, as shown in vessel nozzles, such as the V.C. Summer outlet nozzle cracking Table 44.1. However, experience in France suggests that PWSCC problem described above. As long as the stress improvement may occur at head temperatures close to the reactor cold-leg tem-process is applied relatively early in life, when cracking has not perature. This is especially significant given PWSCC of two initiated or grown to significant depths, it clearly constitutes a South Texas Project Unit 1 BMI nozzles at a temperature of about useful remedial measure that can be applied to vessel nozzles, 565F. The South Texas Project experience shows that materials eliminating one of the major factors that contribute to PWSCC.

and fabrication-related factors can result in PWSCC at tempera-One of the benefits of the weld overlay process described above tures lower than otherwise expected.

to repair PWSCC-cracked butt welds is that it reverses the resid-ual stress pattern in the weld, resulting in compressive stresses on 44.9.3 Surface Treatment the inside surface. Thus, a novel mitigation approach that is being EPRI has sponsored tests of a range of mechanical remedial explored at several plants is the application of weld overlays pre-measures for PWSCC of alloy 600 nozzles. Results of these tests emptively, before cracking is discovered. Applying a preemptive were reported by Rao at the Fontevraud 5 Symposium [54]. The WOL in this manner produces the same remedial benefits remedial measures test program consisted of soliciting remedial described above for the stress improvement processes, but also measures from vendors, fabricating full-diameter and wall-thickness places a PWSCC-resistant structural reinforcement on the outer ring specimens from archive CRDM nozzle material, installing surface of the pipe. So, if the favorable residual stresses were to specimens in rings that locked in high residual stresses on the relax in service, or for some reason be ineffective in arresting the specimen inside surface, applying the remedial measures to the PWSCC phenomenon, the PWSCC-resistant overlay would still stressed surface, and then testing the specimens in doped steam provide an effective barrier against leakage and potential pipe with hydrogen overpressure at 400C (750F). The specimens rupture. Moreover, the revised inspection coverage requirements were removed from the autoclave at intervals and inspected for specified for WOLs apply to such preemptive overlays, providing SCC. A complicating factor in interpreting the test results is that the added benefit of enhanced inspectability [52].

ASME_Ch44_p001-026.qxd 12/19/09 7:37 AM Page 22 22

• Chapter 44 44.9.5 Head Replacement 44.10.3 Assessing Risk of Rupture and Core Damage The most obvious way to address RPV top-head cracking Due to Nozzle Ejection issues is head replacement. Approximately one-third of operating The risk of nozzle ejection (net section collapse) is determined PWRs in the United States have replaced their heads or have using methods such as described in para. 44.6.2.

scheduled head replacements in the near future. Such head replacements take advantage of the lessons learned to date regard- 44.10.4 Assessing Risk of Rupture and Core Damage ing the PWSCC phenomenon, and the new heads are generally Due to Boric Acid Wastage produced so as to eliminate all PWSCC-susceptible materials, The risk of failure of the carbon or low-alloy steel reactor ves-replacing them with resistant materials (alloy 690 and associated sel head by boric acid wastage is determined using methods such weld metals alloys 52 and 152). RPV head replacement is a key as described in para. 44.6.3.

aspect of strategic planning to address the alloy 600 problem in PWRs, and is performed as part of a coordinated alloy 600 main- 44.10.5 Identifying Alternative Life Cycle tenance program that addresses steam generator, pressurizer, and Management Approaches piping issues as well as the RPV. An important step in developing a life cycle management plan is to identify the alternative approaches that can be considered.

These alternatives can include the following:

44.10 STRATEGIC PLANNING (a) continue to inspect and repair indefinitely without applying Within constraints posed by regulatory requirements, utilities remedial measures.

are free to develop a strategic plan that ensures a low risk of leak- (b) apply remedial measures, such as lowering the vessel head age, ensures an extremely low risk of core damage, and results in temperature by increasing bypass flow through the vessel the lowest net present value (NPV) cost consistent with the first internals flange, adding zinc to the primary coolant, and two criteria. Development of a strategic plan for RPV top-head water-jet conditioning the wetted surface of nozzles and nozzles was described by White, Hunt, and Nordmann at the 2004 welds to remove small flaws and leave the material surface ICONE-12 conference [55]. The strategic planning process was with a compressive residual stress.

based on life cycle management approaches and NPV economic (c) replace the vessel head as quickly as possible.

modeling software developed by EPRI [56,57]. (d) replace the vessel head shortly after detecting the first The main steps in the strategic planning process are as follows: PWSCC cracks.

(e) use other approaches identified.

(a) predicting time to PWSCC (b) assessing risk of leaks Each of these alternatives must be studied to determine the (c) assessing risk of rupture and core damage due to nozzle difficulty of application, the likely effectiveness, and the effect of ejection the change on required inspections. For example, head replace-(d) assessing risk of rupture and core damage due to boric acid ment may involve the need to cut an access opening in the con-wastage tainment structure or to procure a new set of CRDMs to allow the (e) identifying alternative life cycle management approaches head changeout to be performed quickly, so as to not adversely (f) evaluating economically the alternative management affect the refueling outage time. If openings must be cut in con-approaches tainment, consideration should also be given to the possible need While the paper and following discussion are based on RPV to cut other openings in the future, such as for steam generator or top-head nozzles, the same basic approach can be applied to BMI pressurizer replacements. Consideration must also be given to the nozzles and butt welds. disposal of a head after it is replaced.

44.10.1 Predicting Time to PWSCC 44.10.6 Economic Evaluations of Alternative Predictions of the time to PWSCC crack initiation are Management Approaches described in para. 44.7.1. The predictions are typically based on a Most life cycle management evaluations include economic statistical distribution such as a two-parameter Weibull or log- analyses to determine the NPV cost of each alternative. The NPV normal model. Predictions are most accurate if based on plant- cost is the amount of money that is required today to pay all pre-specific repeat inspections showing increasing numbers of dicted future costs, including the effects of inflation and the dis-cracked nozzles. If such data are not available, then predictions count rate. Inputs to an LCM economic analysis typically include are typically based on data for other similar plants (e.g., design, the following:

material, operating conditions) with corrections for differences in operating time and temperature. (a) costs of planned preventive activities including inspections, remedial measures, and replacements.

44.10.2 Assessing Risk of Leaks (b) predicted failure mechanisms (e.g., cracks, leaks, and rup-The risk of leakage at a particular point in time (typically refu- ture) and failure rates.

eling outage number) is typically determined by a probabilistic (c) costs for corrective maintenance in the event of a failure (Monte-Carlo) analysis using the distribution of predicted time to including the cost to make the repair, the estimated value of crack initiation (para. 44.7.1), crack growth (para. 44.7.2), and lost production, and an allowance for consequential costs such other probabilistic modeling techniques (para. 44.7.3). The proba- as increased regulatory scrutiny. Consideration should be bilistic analysis should include a sensitivity study to identify the given to the fact that a major incident such as the Davis-Besse most important analysis input parameters, and these important RPV head wastage can result in lost production and conse-parameters should be reviewed to ensure that they can be substan- quential costs far higher than the cost to replace the affected tiated by available data. component.

ASME_Ch44_p001-026.qxd 12/19/09 7:37 AM Page 23 COMPANION GUIDE TO THE ASME BOILER & PRESSURE VESSEL CODE

• 23 FIG. 44.23 TYPICAL RESULTS OF STRATEGIC PLANNING ECONOMIC ANALYSIS FOR RPV HEAD NOZZLES Figure 44.23 shows typical results of a strategic planning (Seche et Aqueuse), Organisé a Saclay les 29-s30 juin et 1er juillet 1959, analysis with economic modeling. North Holland Publishing Cy, Amsterdam, Pays-Bas, 1960.

The final step in the economic evaluation is to review the pre- 10. Copson HR, Berry WE. Corrosion of Inconel Nickel-Chromium dictions in light of other plant constraints, such as planned plant Alloy in Primary Coolants of Pressurized Water Reactors. Corrosion life, potential power uprates, budget constraints, and the availability 1962;18:21t-26t.

of replacement heads. In many cases, the alternative with the low- 11. Copson HR, Dean SW. Effect of Contaminants on Resistance to Stress est predicted NPV cost may not represent the best choice. Corrosion Cracking of Ni-Cr Alloy 600 in Pressurized Water.

Corrosion 1965;21(1):1-8.

12. Copson HR, Economy G. Effect of Some Environmental Variables on 44.11 REFERENCES Stress Corrosion Behavior of Ni-Cr-Fe Alloys in Pressurized Water.

Corrosion 1968;24(3):55-65.

1. SMC 027, Inconel Alloy 600. In: Special Metals Corporation Handbook. 2000. 13. Rentler RM, Welinsky IH. Effect of HN03-HF Pickling on Stress Corrosion Cracking of Ni-Cr-Fe Alloy 600 in High Purity Water at

2. White DE. Evaluation of Materials for Steam Generator Tubing. 660F (WAPD-TM-944). Bettis Atomic Power Laboratory; 1970.

Bettis Technical Review, report WAPD-BT-16, December 1959.

14. Hübner W, Johansson B, de Pourbaix M. Studies of the Tendency to

3. Howells E, McNary TA, White DE. Boiler Model Tests of Materials Intergranular Stress Corrosion Cracking of Austenitic Fe-Cr-Ni for Steam Generators in Pressurized Water Reactors. Corrosion Alloys in High Purity Water at 300C (Studsvik report AE-437).

1960;16:241t-245t. Nykoping, Sweden; 1971.

4. Copson HR, Berry WE. Qualification of Inconel for Nuclear Power 15. Debray W, Stieding L. Materials in the Primary Circuit of Water-Plant Applications. Corrosion 1960;16:79t-85t. Cooled Power Reactors. International Nickel Power Conference,

5. Copson HR. Effect of Nickel Content on the Resistance to Stress- Lausanne, Switzerland, May 1972, Paper No. 3.

Corrosion Cracking of Iron-Nickel-Chromium Alloys in Chloride 16. Shoemaker C. Proceedings: Workshop on Thermally Treated Alloy Environments. First International Congress on Metallic Corrosion 690 Tubes for Nuclear Steam Generators (NP-4665S-SR). Palo Alto, London, 1961, p328-333; Butterworths, 1962. CA: Electric Power Research Institute; 1986.

6. LaQue FL, Cordovi MA. The Corrosion of Pressure Circuit Materials 17. Bruemmer SM, et al. Microstructure and Microdeformation Effects in Boiling and Pressurized-Water Reactors (Special Report 69). on IGSCC of Alloy 600 Steam Generator Tubing. Corrosion 87, Paper London: The Iron and Steel Institute; 1961: 157-178. No. 88, NACE, 1987.

7. Copson HR, Berry WE. Corrosion of Inconel Nickel-Chromium 18. Cattant F. Metallurgical Investigations of CRDM Nozzles From Bugey Alloy in Primary Coolants of Pressurized Water Reactors. Corrosion and Other Plants. Proceedings: 1992 EPRI Workshop on PWSCC of 1962;18:21t-26t. Alloy 600 in PWRs, Orlando, FL, December 1-3, 1992; Paper B5 (TR-

8. Bush SH, Dillon RL. Stress Corrosion in Nuclear Systems. Stress 103345), Palo Alto, CA: Electric Power Research Institute; 1993.

Corrosion Cracking and Hydrogen Embrittlement of Iron Base Alloys, 19. Bandy R, van Rooyen D. Stress Corrosion Cracking of Inconel Alloy Conference held at Unieux-Firminy, France, June 12-16, 1973, pp. 600 in High Temperature Water: An Update. Corrosion 83, Paper No.

61--79, Case 3, NACE, 1977. 139, NACE, 1983.

9. Coriou MM, et al. Corrosion Fissurante suos Contrainte de LInconel 20. Yonezawa T, et al. Effect of Cold Working on the Stress Corrosion dans LEau a Haute Température. 3e Colloque de Métallurgie Corrosion Cracking Resistance of Nickel-Chromium-Iron Alloy. Conference:

ASME_Ch44_p001-026.qxd 12/19/09 7:37 AM Page 24 24

• Chapter 44 Materials for Nuclear Reactor Core Applications, Vol. 2, Bristol, UK, 37. Fyfitch S, Whitaker DE, Arey ML. CRDM and Thermocouple Nozzle October 27-29, 1987; London: Thomas Telford House; 1987. Inspections at the Oconee Nuclear Station. Proceedings of the 10th International Symposium on Environmental Degradation of Materials

21. Seman DJ, Webb GL, Parrington RJ. Primary Water Stress Corrosion in Nuclear Power Systems-Water Reactors, NACE, 2001.

Cracking of Alloy 600: Effects of Processing Parameters (TR-100852).

Proceedings: 1991 EPRI Workshop on PWSCC of Alloy 600 in PWRs, 38. Thomas S. PWSCC of Bottom-Mounted Instrument Nozzles at South Palo Alto, CA: Electric Power Research Institute; 1992: 1-18. Texas Project (Paper 49521). Proceedings of 12th International Conference on Nuclear Engineering, Arlington, VA, April 25-29,

22. Yonezawa T, Sasaguri N, Onimura K. Effects of Metallurgical Factors 2004.

on Stress Corrosion Cracking of Ni-Based Alloys in High Temperature Water. Proceedings of the 1988 JAIF International Conference on 39. U.S. NRC Issuance of Order Establishing Interim Inspection Water Chemistry in Nuclear Power Plants, 1988, p. 490. Requirements for Reactor Pressure Vessel Heads at Pressurized Water Reactors (EA-03-009). Washington, DC: U.S. Nuclear Regulatory

23. Buisine D, et al. PWSCC Resistance of Nickel-Based Weld Metals Commission; 2003.

With Various Chromium Contents (EPRI TR-105406). Proceedings:

1994 EPRI Workshop on PWSCC of Alloy 600 in PWRs. Palo Alto, 40. U.S. NRC Leakage from Reactor Pressure Vessel Lower Head CA: Electric Power Research Institute; 1995. Penetrations and Reactor Coolant Pressure Boundary Integrity

24. Amzallag C, et al. Stress Corrosion Life Assessment of 182 and 82 (Bulletin 2003-02). Washington, DC: U.S. Nuclear Regulatory Welds Used in PWR Components. Proceedings of the 10th Commission; 2003.

International Symposium on Environmental Degradation of Materials 41. U.S. NRC Reactor Pressure Vessel Lower Head Penetration Nozzles in Nuclear Power Systems-Water Reactors, NACE, 2001. (Bulletin 2003-02), Temporary Instruction 2515/152. Washington,

25. Hunt ES, et al. Primary Water Stress Corrosion Cracking (TR-103824). DC: U.S. Nuclear Regulatory Commission; 2003.

In: Steam Generator Reference Book, Revision 1. Palo Alto, CA: 42. U.S. NRC Technical Report on Material Selection and Processing Electric Power Research Institute; 1994. Guidelines for BWR Coolant Pressure Boundary Piping (NUREG-

26. White GA, Hickling J, Mathews LK. Crack Growth Rates for 0313, Rev. 2). Washington, DC: U.S. Nuclear Regulatory Evaluating PWSCC of Thick-Wall Alloy 600 Material. Proceedings Commission; 1988.

of the 11th International Conference on Environmental Degradation of 43. Managing Boric Acid Corrosion Issues at PWR Power Stations. In:

Materials in Nuclear Power Systems-Water Reactors, ANS, 2003. Boric Acid Corrosion Guidebook, Rev. 1. Palo Alto, CA: Electric

27. Attanasio S, Morton D, Ando M. Measurement and Calculation of Power Research Institute; 2001.

Electrochemical Potentials in Hydrogenated High Temperature Water, 44. Staehle RW, Stavropoulos KD, Gorman JA. Statistical Analysis of Including an Evaluation of the Yttria-Stabilized Zirconia/Iron-Iron Steam Generator Tube Degradation (NP-7493). Palo Alto, CA:

Oxide (Fe/Fe3O4) Probe as a Reference Electrode. Corrosion 2002, Electric Power Research Institute; 1991.

Paper 02517, NACE, 2002.

45. Turner APL, Gorman JA, et al. Statistical Analysis of Steam

28. Pressurized Water Reactor Primary Water Chemistry Guidelines, Generator Tube Degradation: Additional Topics (TR-103566). Palo Revision 5, Section 2.3. Palo Alto, CA: Electric Power Research Alto, CA: Electric Power Research Institute; 1994.

Institute; 2003.

46. Stavropoulos KD, Gorman JA, et al. Selection of Statistical

29. Morton DS, Hansen M. The Effect of pH on Nickel Alloy SCC and Distributions for Prediction of Steam Generator Tube Degradation.

Corrosion Performance. Corrosion 2003, Paper 03675, NACE, 2003. Proceedings of the 5th International Symposium on Environmental 30 Rebak RB, McIlree AR, Szklarska-Smialowska Z. Effects of pH and Degradation of Materials in Nuclear Power Systems - Water Stress Intensity on Crack Growth Rate in Alloy 600 in Lithiated and Reactors, pp. 731-738, ANS, 1992.

Borated Water at High Temperature. Proceedings of the 5th 47. Gorman JA, et al. PWSCC Prediction Guidelines (TR-104030). Palo International Symposium on Environmental Degradation of Materials Alto, CA: Electric Power Research Institute; 1994.

in Nuclear Power Systems - Water Reactors, pp. 511-517, ANS, 1992. 48. Materials Reliability Program (MRP) Crack Growth Rates for Evaluating Primary Water Stress Corrosion Cracking (PWSCC) of

31. Hunt ES, Gross DJ. PWSCC of Alloy 600 Materials in PWR Primary Thick-Wall Alloy 600 Materials (MRP-55NP) Revision 1, EPRI, Palo System Penetrations (TR-103696). Palo Alto, CA: Electric Power Alto, CA: 2002. 1006695-NP.

Research Institute; 1994.

49. Riccardella P, Cofie N, Miessi A, Tang S, Templeton B.

32. U.S. NRC Crack in Weld Area of Reactor Coolant System Hot Leg Probabilistic Fracture Mechanics Analysis to Support Inspection Piping at V. C. Summer (Information Notice 2000-017, 2000; Intervals for RPV Top Head Nozzles. U.S. NRC/Argonne National Supplement 1, 2000; Supplement 2, 2001). Washington, DC: U.S. Laboratory Conference on Vessel Head Penetration Inspection, Nuclear Regulatory Commission. Cracking, and Repairs, September 29-October 2, 2003, Gaithersburg,

33. Hunt ES, Gross DJ. PWSCC of Alloy 600 Materials in PWR Primary Maryland.

System Penetrations (TR-103696). Palo Alto, CA: Electric Power 50. Materials Reliability Program (MRP-113NP): Alloy 82/182 Pipe Butt Research Institute; 1994. Weld Safety Assessment for U.S. PWR Plant Designs (1007029-NP).

34. U.S. NRC Circumferential Cracking of Reactor Vessel Head Palo Alto, CA: Electric Power Research Institute; 2004.

Penetration Nozzles (Bulletin 2001-01). Washington, DC: U.S. 51. ASME BPVC Section XI, Rules for Inservice Inspection of Nuclear Nuclear Regulatory Commission; 2001. Power Plant Components. In: ASME Boiler and Pressure Vessel

35. U.S. NRC Reactor Pressure Vessel Head Degradation and Reactor Code. New York: American Society of Mechanical Engineers; Coolant Pressure Boundary Integrity (Bulletin 2002-01). Washington, 2002.

DC: U.S. Nuclear Regulatory Commission; 2002. 52. ASME BPVC Code Case N-504-2, Alternative Rules for Repair of

36. U.S. NRC Reactor Pressure Vessel Head and Vessel Head Penetration Classes 1, 2, and 3 Austenitic Stainless Steel Piping,Section XI, Nozzle Inspection Programs (Bulletin 2002-02). Washington, DC: Division 1. In: ASME Boiler and Pressure Vessel Code. New York:

U.S. Nuclear Regulatory Commission; 2002. American Society of Mechanical Engineers; 1997.

ASME_Ch44_p001-026.qxd 12/19/09 7:37 AM Page 25 COMPANION GUIDE TO THE ASME BOILER & PRESSURE VESSEL CODE

• 25

53. Riccardella PC, Pitcairn DR, Giannuzzi AJ, Gerber TL. Weld Overlay 62. G. A. White, N. S. Nordmann, J. Hickling, and C. D. Harrington, Repairs From Conception to Long-Term Qualification. International Development of Crack Growth Rate Disposition Curves for Primary Journal of Pressure Vessels and Piping 1988;34:59-82. Water Stress Corrosion Cracking (PWSCC) of Alloy 82, 182, and 132 Weldments, Proceedings of the 12th International Conference on

54. Rao GV, Jacko RJ, McIlree AR. An Assessment of the CRDM Alloy Environmental Degradation of Materials in Nuclear Power Systems -

600 Reactor Vessel Head Penetration PWSCC Remedial Techniques.

Water Reactors, TMS, 2005.

Proceedings of Fontevraud 5 International Symposium, September 23-27, 2002. 63. ASME Code Case N-729-1,Section XI, Division 1, Alternative Examination Requirements for PWR Reactor Vessel Upper Heads

55. White GA, Hunt ES, Nordmann NS. Strategic Planning for RPV Head With Nozzles Having Pressure-Retaining Partial-Penetration Welds, Nozzle PWSCC. Proceedings of ICONE12, 12th International Conference approved March 28, 2006.

on Nuclear Engineering, April 25-29, 2004, Arlington, Virginia.

64. ASME Code Case N-722,Section XI, Division 1, Additional

56. Demonstration of Life Cycle Management Planning for Systems, Inspections for PWR Pressure Retaining Welds in Class 1 Pressure Structures and Components: With Applications at Oconee and Prairie Boundary Components Fabricated with Alloy 60/82/182 Materials, Island Nuclear Stations, Palo Alto, CA: Electric Power Research approved July 5, 2005.

Institute; Charlotte, NC: Duke Power; East Welch, MN: Northern States Power (now Xcel Energy); 2001. 65. S. Rahman and G. Wilkowski, Net-Section-Collapse Analysis of Circumferentially Cracked CylindersPart I: Arbitrary-Shaped

57. Demonstration of Life Cycle Management Planning for Systems, Cracks and Generalized Equations, Engineering Fracture Mechanics, Structures and Components - Lcm VALUE User Manual and Tutorial. Vol. 61, pp. 191-211, 1998.

Palo Alto, CA: Electric Power Research Institute; 2000.

66. G. Wilkowski, H. Xu, D.-J. Shim, and D. Rudland, Determination of

58. Materials Reliability Program: Primary System Piping Butt Weld the Elastic-Plastic Fracture Mechanics Z-factor for Alloy 82/182 Weld Inspection and Evaluation Guidelines (MRP-139), EPRI, Palo Alto, Metal Flaws for Use in the ASME Section XI Appendix C Flaw CA: 2005. 1010087. Evaluation Procedures, PVP2007 26733, Proceedings of ASME-

59. Materials Reliability Program: Advanced FEA Evaluation of Growth PVP 2007: 2007 ASME Pressure Vessels and Piping Division of Postulated Circumferential PWSCC Flaws in Pressurizer Nozzle Conference, San Antonio, TX, 2007.

Dissimilar Metal Welds (MRP-216, Rev. 1), EPRI, Palo Alto, CA: 67. G. M. Wilkowski, et al., Degraded Piping Program-Phase II, 2007. 1015400.s Summary of Technical Results and Their Significance to Leak-

60. Materials Reliability Program: Technical Basis for Preemptive Weld Before-Break and In-Service Flaw Acceptance Criteria, NUREG/CR-Overlays for Alloy 82/182 Butt Welds in PWRs (MRP-169), EPRI, 4082, Vol. 1-8, January 1985 to March 1989.

Palo Alto, CA: 2005. 1012843. 68. Materials Reliability Program Reactor Vessel Closure Head

61. Materials Reliability Program Crack Growth Rates for Evaluating Penetration Safety Assessment for U.S. PWR Plants (MRP-110NP):

Primary Water Stress Corrosion Cracking (PWSCC) of Alloy 82, 182, Evaluations Supporting the MRP Inspection Plan, EPRI, Palo Alto, and 132 Welds (MRP-115NP), EPRI, Palo Alto, CA: 2004. 1006696-NP. CA: 2004. 1009807-NP.