In September 1991, leakage occurred from the Bugey 3 T54 vessel head pen­etration in France. After 10 years of monitoring the leak was detected using the acoustic emission method as part of the thermal-hydraulic test and it was estimated to be approximately 1 L/h. Non-destructive inspection using dye penetrant testing, eddy-current testing and ultrasonic testing confirmed a ver­tical crack penetrating the pipe in the opposite direction to the weld. Through metallographic analysis, it was concluded that the crack was PWSCC. By the end of 1992, Non-Destructive Examination (NDE) programmes using Eddy Current Test (ECT), Ultra sound Test (UT) and Visual Test (VT) had found penetrating cracks in five 900 MW vessel heads and four 1300 MW vessel heads (after 30 000-40 000 h of operation). Due to vessel head and RPV homogeneity, EDF (the operating utility) decided to replace all of the vessel heads with Alloy 600 penetrations (54 out of 58 vessel heads are Alloy 600, the remaining four are Alloy 690 penetrations). In France, pressure ves­sel heads at Bugey Unit 5 have been replaced since 1993; in Japan, replace­ment of the pressure vessel heads started at Takahama-1 in 1996; in Spain at Almaraz-1 in 1996; in Sweden at Ringhals-2 in 1996; in Belgium at Tihange-1 in 1999; in the United States at North Anna-2; and in China at Guangdong-2 in 2003. As of 2005, the pressure vessel heads of 93 power plants across the world had been replaced with Alloy 600 penetrations.

The importance of carbide was first established for steam generator tubes, but this has also been applied to penetration of the upper heads. Interpretation was based on carbon content, as well as forging, tempera­ture of rolling and resistance strength upon hot forming. Three classes were determined in terms of grain boundary carbide decoration: Class 1 — well — developed intergranular carbide structure; Class 2 — mainly prior grain boundary carbides with recrystallized grain; Class 3 — mainly intragranular carbides with recrystallized grain. Modelling for crack initiation probability took into account the composition of materials considering the influence of other classes, angle of penetration and location (i. e. near or opposite to the weld zone) and the influence of cold working treatment.

In 2004, after over 100 000 operating hours, leakage of the 47th CRDM head penetration was detected in Ohi Unit 3 in Japan through visual inspec­tion. Ohi Unit 3 is a power plant where RV head reactor coolant system (RCS) temperature had been revised from 289°C to 310°C in 1997. In order to distinguish the leakage ratio, helium leakage testing, eddy-current test­ing, dye-penetration testing and ultrasonic testing were carried out on the

J-weld of the 47th CRDM head penetration. The leak in the J-weld was found through helium leakage testing; ECT found some indications of cracks on the J-weld. After grinding of the J-welds, dye-penetration testing was carried out. Cracks were observed on the portion of dye penetrant indication located along the grain boundary. It was ascertained by additional grinding that the long crack was connected with other cracks. The reactor vessel head (RVH) was replaced in 2007. The new head had a penetration nozzle and J-weld made from Alloy 690. There are 23 in-service PWR plants in Japan. At the present, 14 plants have replaced RVHs and seven additional plants will replace RVHs with Alloy 690 TT in the near future. One power plant has had CRDM nozzle penetrations which have been thermally treated with Alloy 690 since plant construction. Other power facilities in Japan have solved the issues of CRDM head penetrations by reducing the temperature at penetration.

In April 2003, a small amount of boric acid sediment was found in two BMI penetrations (No. 1 and No. 46 of a total of 58) in South Texas Project Unit 1 (STP Unit 1). This is the only evidence of leakage of BMI nozzle penetration reported in facilities in the United States up to now. The BMI penetrations of STP Unit 2 were built with drilled Alloy 600 bar, and con­nected to the lower head of the nuclear reactor vessel by welding Alloy 82/182 J-grooves.

In January 2003, a small cracking signal was detected on the internal sur­face area of the BMI penetration nozzle, and 50 BMI penetrations were found in Takahama Unit 1 in Japan through ECT. This indicator was within the permissible limit (<3 mm depth) but it was concluded that the facility in this unit was likely to be at the beginning of PWSCC. The utility applied water jet peening on the surface in the BMI penetration nozzle after remov­ing the crack indication. Laser and water jet peening are used for relief in other Japanese PWR plants. The peening method has been carried out in welding J-grooves in this location.

In September 2003, the thirteenth regular inspection of the PZR nozzles in Tsuruga Unit 2 in Japan was conducted. Cracks were found in the weld zone of the pressure relief line nozzle stub. This was the first case where sed­iments of boric acid were found. In ultrasonic testing on the relief line stub, two indicators were found located on the repair weld zone. In ultrasonic testing of other nozzle stubs, an indicator on the safety valve was found, but nothing was detected elsewhere. According to the observation, cracks remained in the weld zone only and developed in a circumferential direc­tion of the pipe. The cause was analysed as SCC created in nickel-based alloy (600 type), also of the same type as the material welded in the PWR first coolant environment. The welding metal for the weld zone of the pipe nozzle on the pressure relief valve, the pipe nozzle stub on the pressure relief valve and the safe end was changed to nickel-based alloy (690 type) which has resistance to SCC. At the end of 2008, most of the pressurizer nozzles would be supplemented with an Alloy 52M weld cover. Minor pres — surizers have been replaced with Alloy 690 material.

Many incidences of cracks in Alloys 182 and 82 have been found in in-service PWR power plants. In July 2000, cracks at the outlet nozzle to pipe safe end weld of Ringhals Unit 4 were found. Many small axial cracks were found and removed with a boat sample through electro discharge machining (EDM). The first cracking had in fact been discovered in Ringhals Unit 3 in June of that year, but the power plant was permitted to continue to operate without any repair because the crack seemed superficial and shallow in depth. In both cases, welding used Alloy 182 and cracks were axial. The cracks in Ringhals Unit 3 and Unit 4 were removed in 2003 and in 2004, respectively, and the cracks were repaired by welding inlay using Alloy 52 M.

The next largest incident happened in October 2000. Penetrating cracks and leakage were found in the V. C. Summer power plant in the same part as in Ringhals Unit 3. Initial UT was carried out on the internal surface of the pipe and as a result, an axial defect near the upper part of the pipe was dis­covered. The next test was conducted in spring 2001 and many defects were found. All of the defects were axial, and the largest defect was penetrated. The defects were removed, and a new spool piece was welded. The part was restored to its original condition. Alloy 52 was used for the V. C. Summer repair from the exit nozzle to the pipe weld zone; Alloy 82 was used in some parts for thickness and for the rest of the weld zone. The other V. C. Summer exit nozzle was repaired by using mechanical stress improvement process (MSIP).

The in-service inspection (ISI) programmes delivered partly by the sup­plier or developed by Hungarian institutes basically follow the ex-Soviet regulation.

Recent review and overall updating of the ISI programmes adopt state-of-the-art techniques and methodologies (e. g. AS ME Section XI). Extensive studies are ongoing to provide a solid basis for changing the rules and techniques of ISI. One practical question is the periodicity of the ISI programmes, which is four years at Paks NPP, in accordance to the ex-Soviet regulation. For practical reasons the new ISI period should be eight years. At the same time, the scope and depth of ISI programmes also have to be upgraded. This type of modification is not unique; moreover there are similar examples among the countries operating VVER-440 type NPPs (e. g. Finland); however, the change cannot be performed routinely, it requires careful justification.

The internals (flux thimble tube, lock bars, baffle bolts and re-entrant cor­ners) are made from austenitic stainless steels of type SA304, 316 or 347. The insignificantly low swelling behaviour, observed from the limited tests conducted in environments that are more severe than those of a PWR, is believed to be associated with irradiation-induced formation of very fine precipitates (such as G phase, carbides and Y phase) in high number density whose interfaces act as efficient sinks to irradiation-induced vacancies and thereby the agglomeration of the vacancies is suppressed.92

The RPV of a PWR is a cylindrical vessel with two hemispherical shells — one at the top, bolted with flange, and one at the bottom, welded — which contains the core through which pressurized light water is circulated at an average temperature and pressure of 300°C and 16 MPa, respectively. The inner surface of the cylindrical vessel is lined with a thin layer (thickness around 3-10 mm) of austenitic steel (SAE 308/309) to protect the vessel from corrosion. The vessel is made from plates of low alloy steel (typically ASTM SA302/SA533B or ASME SA508) with a thickness of about 225 mm. Although many materials are acceptable for reactor vessels accord­ing to Section III of the ASME Code, the special considerations pertain­ing to fracture toughness and radiation effects limit the basic materials for most parts of vessels to SA533 Grade B Class 1, SA508 Class 2 and

SA508 Class 3. Creep per se does not pose any safety related problem to RPVs.

A major issue with the ferritic steels used for RPV applications, as described earlier, lies in the increase in DBTT and decrease in the upper shelf energy due to radiation exposure and these factors depend on the concentration of alloying and/or impurity elements. Sensitivity of radiation embrittlement of ferritic steels to the concentration of copper is illustrated in Fig. 1.35a and 1.35b that depict the effects of copper composition on the increase in transition temperature and decrease in the upper shelf energy, respectively, with neutron radiation dose.93 Welds and HAZ are relatively more sensitive to radiation exposure than the base metal, the reasons being variations in composition, microstructures, etc. Detailed knowledge gath­ered on the effects of alloying compositions (in particular Cu, S, P and Ni) on radiation embrittlement of RPV steels makes it possible to design mate­rials for new systems to be devoid of these issues that are confronting the currently operating reactors where the fracture characteristics in the weld materials are degraded, mainly by impurities such as Cu. There have been numerous studies to understand the underlying micro-mechanisms respon­sible for the observed radiation embrittlement of RPV steels and the role of alloying and impurity elements.9495 Recent emphasis has been on atomistic modelling along with the characterization of defects using advanced micro­structural evaluation techniques such as HRTEM, atom probe microscopy, small angle neutron scattering (SANS), etc.96

In the reactor vessel surveillance programmes (RVSPs), samples taken from the base, weld and HAZ of the actual vessel material used during its construction are included in a capsule that is placed closer to the reactor core so that the samples withdrawn after different neutron dose levels can be tested (tensile, Charpy and fracture toughness). The results from these surveillance samples provide information regarding the degradation of the real structure and corrective action can be taken before any major damage occurs. While hardness and tensile tests are routinely performed to get an idea on the effects of neutron irradiation, the effect of radiation dose on RTNDT and upper and lower shelf energies through Charpy tests are sig­nificant in evaluating the radiation embrittlement. The RVSP capsules are taken out at intervals during reactor operation and changes in the proper­ties of the samples are monitored to make sure that these changes are less than those prescribed by the NRC regulation guide (10CFR50). In cases where the results reveal degradation greater than the limit prescribed by the regulation guide, the reactor vendor/utility needs to take appropriate actions to demonstrate the safety of continued operation of the reactor so that the RPV does not fail in a brittle mode. It is to be noted that the Charpy tests do not yield fracture toughness (K) data which are related to the crack length (see, e. g. Equation [1.9]) and the specimen size required for obtaining valid K1C tests, as in the case of low strength RPV steels, is too large to be practical to investigate radiation effects. Thus, efforts are being put in to correlating the CV values obtained from standard Charpy tests to fracture toughness evaluated using compact tension (CT) and/or elastic-plastic toughness tests (J1C, etc.).97 The master curve approach pro­vides an alternative transition temperature index parameter to the RTndt data measured from Charpy tests. This new parameter, defined as RTT0,98 is based on a simple addition of 19.4°C (35°F) to the value of T0 evalu­ated according to ASTM E 1921. The advantage of this approach is that RTTo can be measured directly on irradiated samples rather than having to measure initial properties and then add the transition temperature shift.99 It is also worthwhile considering ‘dynamic’ values such as dynamic frac­ture toughness (K1d) which are sensitive to applied strain-rate and which are of importance during accidents such as loss of coolant (LOCA); K1d is generally determined using pre-cracked samples by instrumented Charpy impact tests100 though these are not routinely considered in RVSP sched­ules. Further details on the reactor vessel integrity are included in a later chapter in Part II.

At the VVER-1000 plants, ageing may affect the pre-stressing of the con­tainment. Important ageing mechanisms of the pre-stressed containment resulting in loss of pre-stress are the relaxation of tendons, shrinkage, creep of steel. Requirements for the testing of the containment pre-stressing system are defined both by the designer and regulation (Orgenergostroy, 1989a; 1989b). The scope of inspection should be extended if defects are observed, and/or if average loss of tension force is more than 15%. If fur­ther testing verifies the results obtained, it is necessary to test 100% of the tendons. Tendons with force losses of more than 15% should once again be controlled after straining. If a force loss at 24 hours is more than 10%, the tendon should be replaced.

In order to enable monitoring of the level of the containment pre-stress­ing, measurement systems are installed permanently on the structure and these systems measure structural deformations and pre-stressing force in the cables. At VVER-1000 plants, detailed field investigations and analyses have been carried out for the assessment and evaluation of the condition of pre-stressing tendons. There are design solutions for the replacement of tendons. Thus, all existing defects leading to a loss of stressing force and rupture of tendons have been avoided. At some plants, new pre-stressing systems and an additional system for automatic control of stressing forces is installed in the bundles.

The containment structure of an LWR acts as both a barrier to the spread of fission products from the reactor into the environment and as a shield to protect the nuclear components within it from missiles such as from aircraft and errant turbine blades. The issues with the containment include attack of the concrete by the surrounding environment (for instance, acid rain), attack of the rebar within the containment causing the concrete to spall off the outside (or inside), and corrosive attack of the steel liner. Besides the containment structure, which is the biggest and most obvious component, there are other components that must also work as designed to control the spread of fission products outside of the containment. Among these are:

• Steam isolation valves that block the main steam lines from the reactor inside the containment to the turbines outside of the containment.

• Various heat exchangers that provide cooling of systems inside the con­tainment using cooling water from outside of the containment. One example of these is the aluminum air coolers that remove heat after a LOCA from the steam/air mixture inside the containment.

Maintaining the integrity of the containment structure and the other com­ponents that separate the reactor from the environment requires research and development in the following areas:

• Effect on the containment concrete of the external environment (for instance acid rain, bird feces, thermal cycling, etc.) as well as radiation over very long periods of time.

• Effect of high moisture levels on the containment steel liner, galvanized steels, and aluminum components found in the containment.

• Effect of age, temperature, erosive flow on large valves and piping with the associated welds throughout the system for both steam and water.

• Corrosion of stainless steel and aluminum heat exchangers both under normal operating conditions of high humidity, moderate temperatures and moderate radiation levels.

• Non-destructive examination of large structures.

• Methods to repair large concrete and steel structures.

The acceptance criteria are expressed as a limit value for the controlled parameter of the ageing. The limit value corresponds to the performance or functioning with required margin. Acceptance criteria have to be defined for each component, or commodity, for each degradation mechanism in relation to fulfilment of the intended safety function. The acceptance crite­ria can be derived from stress calculations in case of allowable wall thick­ness of piping, or fatigue calculation regarding allowable load cycles. The acceptance criteria for degradation phenomena entailing decrease of the brittle toughness are determined by the relevant TLAA analysis results.

The compliance criteria for water chemistry parameters are defined in the relevant chemistry instructions.

The steps described in the sections above can be illustrated by the exam­ples for civil structures given in Table 8.8.

Since LCF and HCF are controlled by ductility and strength respectively, and radiation results in hardening and embrittlement, we expect life in HCF to be improved and that in LCF to be degraded. Murty and Holland15 examined the fatigue characteristics of Type 304 SS from hexagonal cans of EBR-II guide tubes before and after irradiation to a fast fluence of ~8 x 1026 n/m2 (Fig. 1.12a). Tests were performed in four-point bend mode at con­stant displacements under symmetrical strain reversal fatigue at 0.1 cps and strains were varied from ~1% to 2.4% with the number of cycles to failure varying from 500 to 40 000. While a slight decrease in fatigue life is noted at high strains or low cycles, the data clearly revealed improved fatigue life at low strains or high cycles (Fig. 1.26) where the model predictions using universal slopes are correlated with experimental results. According to their model,

where f is the frequency of cycling, £ the strain-rate, D the ductility, n the work-hardening parameter, au and af are true tensile and fracture stresses and the remaining factors are as defined earlier.

Fatigue crack growth follows the same kinetics as described in section

1.2.4 but with KIC replaced by КЦЇ in Equation [1.14], similar to the case of environmental effects where KIC is replaced by KISCC.

Grain boundary sliding as a mechanism of creep is usually observed at tem­peratures higher than 0.4TM. GBS is typically a response of grain boundaries

(usually high angle) to an applied shear stress and is supposed to occur by the relative motion of grains along a common boundary or along a narrow zone immediately adjacent to the boundary. The relative motion of grains along a common boundary is known as Lifshitz GBS when the accommo­dating process is diffusional creep.44 On the other hand, when the process of accommodation occurs by glide and climb of dislocations, the GBS pro­cess is termed as Rachinger sliding.45 GBS with the deformation limited to a zone around the boundary comes under the category of Rachinger slid­ing. Accommodation is necessary to avoid the formation of voids at the grain boundary. Strain compatibility and relaxation of stress concentra­tion are only possible through the process of accommodation, usually by diffusional flow.

By ‘mechanical properties’ we essentially mean strength and ductil­ity. Strength is expressed in terms of hardness, tensile strength, burst strength, fatigue strength, etc. Ductility is likewise expressed in terms of strain-to-failure or strain-to-some limit for the various loading conditions. Fracture toughness is a combination of strength and ductility which describes the stress required to propagate a specific crack geometry under specific loading conditions. In this section we discuss various mechanical properties as affected by reactor neutron irradiation. In addition, we describe mecha­nisms and parameters which are related to mechanical properties and which affect reactor component behaviour. This section deals primarily with prop­erties which can be determined by out-of-reactor (or post-irradiation) test­ing. For instance, tensile properties (strength and ductility) of interest for in-reactor performance are mainly dependent on fluence and independent of flux. However, if the rate at which strain is applied becomes very low (<10-6 s-1), the mechanism of deformation changes, and flux becomes an important variable (Azzarto et al. , 1969). That phenomenon is dealt with more as creep, in the section on dimensional stability.

As described in Section 4.3.1, neutron irradiation dramatically alters the microstructure of zirconium alloys. Of importance for mechanical prop­erties are creation of <a> dislocation loops, and to a lesser extent, disso­lution of precipitates (SPPs). Irradiation increases strength and hardness, and decreases ductility. The effect on fatigue life (or strength) is less clear and depends on testing technique, but generally appears to be small, with some reduction of fatigue life in the low cycle region (Wisner et al, 1994). Fracture toughness is clearly reduced by irradiation in Zr-2.5Nb (Davies et al., 1994), with concurrent effects of trace elements like chlo­rine (Coleman & Theaker, 2004), and there are indications of a smaller reduction in Zircaloys, for example (Bertsch et al, 2010). The combination of irradiation and hydride effects is important; for uniformly distributed hydrides the observed reduction in ductility is mainly an irradiation, rather than a hydride effect; however concentrations of hydrides (rims, blisters) or low test temperature can overwhelm irradiation effects. Details are given in the following sub-sections.

Sipping

If a core is suspected or known to contain leaking fuel, the fuel is almost universally checked for leaks during the subsequent refuelling (or some­times forced) outage. Various techniques are used to identify leaking fuel assemblies and are collectively referred to as ‘sipping’ methods. They rely on changes in the concentration of fission products that are released from the FA or rod being tested to signify the presence of a leak; such as, changes in gamma activity, beta activity, isotopic composition of fission products or combinations of such measurements. The technique used to identify a leak­ing assembly can vary depending on the size of the leak, the background activity from tramp uranium and on the time of sipping relative to shut down. Changes in the gross (total) gamma activity of water or noble gas sam­ples representative of the fuel being tested are sometimes used to identify a leaker, particularly in cases of low background activity. In general, however, changes in the gamma or beta activity of nuclides with moderate-to-long half lives are typically used to minimize the effects of background activity and decay time; examples include Xe-133 (5.25 d), I-131 (8.04 d) or Kr-85 (10.72 y). In addition, many of the current sipping methods also involve the collection and measurement of noble gases to enable the detection of leaks that are too small to allow the release of soluble fission products in quanti­ties clearly detectable in the sipping process.

Historically, sipping methods have been classified as wet, dry or vacuum, based on the manner in which fission products are collected for measure­ment (Lin, 1996). In practice, however, sipping techniques can better be clas­sified as ‘open’ or ‘closed’ methods based on the manner in which the fuel being tested is isolated from other assemblies during the sipping process. The wet, in-reactor methods represented by the TELESCOPE, INMAST and various hood systems are open methods. They have become the pri­mary means of leak detection because of the small amount of time needed to inspect a BWR, PWR or VVER core; for example, ~16 h for a large BWR with an in-core sipping hood versus close to a week for vacuum sipping (Knecht et al, 2001). Wet sipping is also used in storage pools to test indi­vidual assemblies or fuel rods. The dry method makes use of a canister to isolate the fuel being tested and is a closed method. Dry sipping is not used today in power reactors because of issues related to handling and test time and to decay heat and cladding temperature. Vacuum sipping also makes use of a canister to isolate individual fuel assemblies and is frequently used to supplement the in-reactor methods because of its higher resolution and detection capabilities.

The wet, in-core or in-mast sipping methods are used almost exclusively in power reactors because of their speed and the need for short refuelling outages. Vacuum or similar wet canister sipping is now used to confirm or resolve the results of the in-reactor methods.