For the justification of operability of RPV and RPV internals for extended operational lifetime, the following analyses have to be performed.

PTS analyses for RPV test the structural integrity against brittle frac­ture (fast fracture) of the RPV; it is ensured if the factual ductile-brittle transition temperature (DBTT) of its critical components is less than the maximum allowable component-specific DBTT. The analysis is based on the comparison of the static fracture toughness of the material and stress intensity factor calculated from the given loading situation (Linear Elastic Fracture Mechanics (LEFM) concept). The steps in the analysis are pre­sented by Katona, Ratkai and Pammer (2011). The final conclusion of the analyses is that the RPVs at Paks NPP can be safely operated for at least 60 years. For the sake of completeness of the studies, some additional anal­yses are still ongoing regarding PTS sequences initiated by internal fires, flooding and earthquakes under shutdown conditions. The neutron fluences also have to be modified taking into account the new fuel design introduced after power up-rate.

Analysis of fracture toughness of structures within the reactor pressure vessel was undertaken. According to the preliminary results the irradiation- assisted stress corrosion cracking and void swelling may be of interest. The stud joints fixing the polygon mantle to the core basket can be critical in both ageing mechanisms. Measures may be identified after visual inspec­tion of the core basket and review of inspection procedure. The possibil­ity of implementation of a non-destructive volumetric test method for the bolts is also a consideration. With respect to void swelling, the possibility of implementation of ultrasonic measurements as well as gamma heating and a replacement programme are being investigated.

Analyses related to operational limits and conditions

Reviews of the Final Safety Analyses Report and reconstruction of design bases, which have been performed at the Paks NPP, resulted in a recognition of the need for justification of operational limits and conditions related to certain ageing phenomena via adequate thermo-hydraulic, stress and frac­ture mechanics analyses. These analyses have been included in the scope of TLAAs required for the justification of LTO of the Paks NPP. The task also includes the justification for modification of the limits and conditions in accordance with operational needs allowing rapid temperature changes in certain cases. The temperature measurements and the temperature rate control methodology have also been reviewed and amended.

The calculation methodology was based on an adaptation of ASME BPVC. For the calculation of temperature transients in the primary system, the RELAP5/mod3.3 code was used. A thermo-hydraulic model was devel­oped for accident simulation. This model consists of a detailed model of the primary system, the heat removal system and the automatic control system and it takes into account operator actions during the heat-up and cool-down processes. The thermo-hydraulic model and the calculation method have been verified via comparison of the calculated transient time histories with the measured ones.

Fuel failure occurs when the cladding barrier is degraded and breached. The fuel rod failure rate in LWRs has been significantly reduced since 1987.75 This is due, besides design improvements, to the introduction of many improved variants of Zr-base alloys over the years. With the recent developments in LWR fuels, Zr-1%Nb-Sn-Fe alloys with higher resistance to irradiation-induced growth, creep and corrosion, are being used for guide tubes and for fuel rod cladding with extended residence time (5-6 years).76 At low burnup, the pellet densifies and the external water pressure causes the clad tube to creep-down. On power ramp, the pellet expands and applies excess strain on the clad. This leads to the pellet touching the clad and results in PCI failure or hydride related cracking. The sheath should have good creep rupture property not to fracture. Further, the creep of the fuel assem­bly and guide thimble can lead to bowing of the assembly.7 7 An analysis performed at Ringhals concluded that the bowing in this reactor had been caused by a large creep deformation due to excessive compressive forces of the hold down spring on the fuel assemblies, with a decrease in lateral stiffness. The proposal to introduce advanced cladding and guide thimble materials with a low growth rate and higher creep resistance to improve the dimensional stability of assemblies is also being considered (M5 by Areva and ZIRLO of Westinghouse);78 further details can be found in Part II on Zr-alloys.

The creep behaviour of unirradiated material is taken as a benchmark to postulate its performance in reactor. Though these out-of-pile tests may not be representative of their in-reactor behaviour, they have been used success­fully to grade various materials during alloy development programmes and to gain a basic understanding of the material behaviour. It has been recog­nized that hoop strain in a clad tube (in-pile or during spent fuel storage) is a vital parameter in the breach of fuel clad and evaluation of their creep and burst behaviours is very important to assess the integrity of the tube. Biaxial creep79 and burst80 characteristics are usually studied using internally pressurized tubing over a range of pressures and temperatures. One may also use ring-creep81 tests to characterize hoop creep behaviour under uni­axial hoop loading; this might be advantageous in cases where only a limited amount of material is available and also in evaluating radiation effects that require relatively small size samples. It becomes relatively more complex for Zircaloys that exhibit distinct textures leading to anisotropic deforma­tion and creep82 that need to be accounted for, along with possible radiation effects, in predicting the dimensional changes in reactor. As demonstrated by Murty and co-workers, the relatively weak hoop direction for CWSR mate­rial became stronger following recrystallization annealing, illustrating the profound effect of heat treatment on the creep anisotropy of Zircaloys.83 The observation that under equi-biaxial stress state the secondary slip sys­tems (basal and pyramidal) are also favoured along with the easier prism slip, concur with the observation that the irradiated recrystallized Zircaloy exhibited a creep locus similar to that of isotropic material (texture reduced as all slip systems were favoured). The recent work on the thermal creep of Zr-2.5%Nb alloy by Kishore et al,84 indicates that a microstructure contain­ing a stable phase creeps faster than the one with a meta-stable phase and a phase redistribution is established. The stable в phase (80 wt%Nb) dissolves during creep deformation and re-precipitates as meta-stable в phase (~35 wt%Nb), this phase change adding to the creep strain. Moreover, transients during sudden load drops or changes in temperature need to be considered in proper predictions.85 The importance and significance of studying the tran­sitions in creep mechanisms along with microstructural characterization fol­lowing deformation were clearly outlined by Gollapudi et al,86 and the reader is referred to Chapter 3 on Creep for further details.

The steam generators in VVER-440 are horizontal (see Fig. 8.1). The advan­tages of the VVER horizontal steam generator design are the high reliabil­ity, absence of vibrations, no accumulation of sludge at the tube sheet and ease of access for maintenance. The SG design has a positive impact on safety as well, for example the design allows reliable natural circulation, effective gas removal, large water inventory and essential thickness of the heat-exchange tubes.

The heat-exchange tubes and the steam generator tube headers (col­lectors) are manufactured from austenitic stainless steel (18% Cr, 10% Ni stabilized with titanium) in VVERs, instead of the nickel-based alloys (Alloy 600 and 690) and higher chromium-containing alloys (Alloy 800) as used in PWRs. The material of the SG heat-exchange tubes in VVER-440

is equivalent to AISI 321. A comparison of the SG materials selected for VVER-440, VVER-1000 and a typical PWR is shown in Table 8.1.

The oldest VVER-440 type steam generators at Novovoronesh NPP Units 3 and 4 have been operating for 40 years. The condition of the old­est VVER-440 steam generators at the Kola and Novovoronesh plants has allowed a 15 year extension of operation for these plants. According to the operational history, the feed-water distributor inside the SG shows accel­erated ageing due to erosion. These elements were replaced at almost all VVER-440 plants (see the dark coloured new distributor in Fig. 8.1). The experience regarding the ageing of VVER steam generators is summarized in TECDOC-1577 from the International Atomic Energy Agency (IAEA, 2007b).

At VVER-440 plants, the lifetime limiting ageing mechanism of the SGs is outer diameter stress corrosion cracking (ODSCC) of the austenitic stain­less steel heat-exchanger tubes. The ODSCC indications appear typically (80%) at the grid structure supporting the tube bundle, where the secondary circuit corrosion products (with concentrated corrosive agents) are depos­ited. An eddy current inspection programme is implemented for monitor­ing the tubes. Samples have been removed from plugged tubes to facilitate investigations into the phenomenon. The rate of the ODSCC was essentially slowed down by a series of modifications and actions, implemented at dif­ferent plants and to different extents. The measures implemented are as follows: [12]

• Introducing high pH secondary water chemistry.

• Replacement of the high-pressure pre-heaters (with erosion-corrosion resistant tubes).

All of these measures have been implemented at the Paks NPP, and have completely changed the conditions and rate of ODSCC in the SGs. Consequently, a better (i. e. decreasing) plugging trend is experienced, which can also be expected in the long term. The gaps between the tubes and sup­port grid are still the critical sites, since any remaining corrosion products will accumulate there. It is therefore difficult to forecast the ODSCC rate in the gaps and the ageing process has to be closely monitored in the future. Under the new conditions, sludge may accumulate at the bottom area of the SG and an effective method for draining it must be found. The reserve in heat-exchanger surfaces of the SG is relatively large (more than 15%). Considering past experience and the recent plugging trend of the heat — exchange tubes, none of the SGs would exceed 10% of plugged tubes by the end of 50 years operation, due to measures implemented (Katona et al, 2003; Trunov et al, 2006b). The number of allowable plugged tubes became more important at the plants where the primary energy output is increased for the power up-rate. Therefore, establishing an adequate performance cri­terion for the steam generators is very important.

Recent (Fukushima) and not so recent events (Three Mile Island 2), have accelerated the quest for new cladding materials that will be much more resistant, if not totally tolerant, to current LOCA conditions (~1200°C for 400 s) as well as more extreme beyond-design basis accident conditions such as a long-term full station blackout where there is little or no sup­ply of coolant to the fuel. The most notable of these materials is the use of SiC composites. As little as is known about the behavior of zirconium alloys on a phenomenological basis under normal or accident conditions, much less is known about SiC composites or any other ceramic materi­als that could potentially replace metal alloys as fuel cladding materials. SiC cladding along with higher density or higher enrichment fuel could provide: [23]

These SiC characteristics allow simplification of safety systems, higher energy density (to reduce capital costs) and longer operating cycles with higher density or enrichment advanced fuels to reduce fuel and reactor operating costs. However, they do come at a cost — the need to develop whole new areas of understanding in behavior and manufacturing including:

• Modeling and design of a SiC composite that will be acceptable for use in reactors, that will not shatter, and can withstand normal handling and operations as well as transients and accidents. However, many dif­ficulties remain before these composite structures are understood well enough to model, including:

◦ The behavior of ceramic composites in a radiation field is not well known.

◦ The interaction between the monolithic SiC tube and the composite layer is especially difficult to model.

• The interaction between pellet and ceramic cladding will require exten­sive tests and modeling.

• The impact of manufacturing variations on the ultimate performance of the ceramic structure under operating and accident conditions will need to be understood.

• Joining an end plug to the tube to form a hermetic seal in a cost effective way is an extremely challenging task.

• Methods to produce about 15 million feet of reactor cladding per year to very exacting specifications at acceptable costs will require significant manufacturing development.

Ultimately, the understanding of all these effects for this relatively new material must be integrated into a licensable fuel performance code. The data that is needed can be gained empirically and fitted into phenomeno­logical models with empirical verification. At a minimum, the following is needed under irradiation and coolant conditions (Lahoda, 2011): [24]

• The impact of core melt on reactor internals and reactor vessel integrity.

• The effects of long-term wet and dry storage as well as environmen­tal conditions in any long-term disposal site on the integrity of the SiC cladding.

• Phenomenological models for the behavior of SiC-composite structures as a function of temperature, stress, and radiation.

The second step of the development of the AMPs is the identification of the means of preventing or controlling ageing. For example, the corrosion phenomena on the internal surfaces can be slowed down via adequate water chemistry parameters. General corrosion and soil corrosion may be reduced by coatings and ensuring the undamaged state of the coatings. The most effective way of avoiding boric acid corrosion is the timely detection and effective termination of leakages onto carbon steel elements, which are the subject of walk-down inspections.

It was mentioned in earlier sections that BCC materials such as iron and some steels show a distinct yield point (Fig. 1.13) where as FCC and HCP materials show a continuous transition from elastic to plastic range (Fig. 1.3). The distinct yield point is due to the locking of the dislocation sources by interstitial impurities such as C and N in low alloy steels that increases the stress resulting in a sudden increase in free or mobile dislocation density. The velocity of these dislocations decreases in order to maintain the imposed

1.12 ( a) Voids and precipitates in irradiated stainless steel.15

(b) Faulted Frank loops and dislocations in irradiated stainless steel.16

(c) Faulted Frank loops in irradiated aluminium, copper and iron.17

1.13 ( a) Typical stress vs strain curve depicting yield point such as is observed in steels20 and (b) extrapolation of the plastic curve to elastic line delineating the source (os) and friction (o) hardening terms.22

constant strain-rate resulting in the observed load or stress drop. The stress maximum known as the upper yield point is followed by deformation tak­ing place within a relatively small region of the specimen (Luders band), with continued elongation of the specimen by the propagation of the band along the gauge section wherein the deformation is inhomogeneous. Once the entire gauge section is traversed by the band, normal strain hardening occurs with stress increasing and further deformation taking place. This dis­tinct yield points (oy) in stress-strain curves can be represented as a sum of a non-zero source hardening term (os) and a friction hardening term which represents the resistance experienced by the mobile dislocation (o;),

ay = ai + as, [1.20a]

similar to the well-known Hall-Petch equation:
where d is the grain size and ky is the Petch unpinning coefficient. Thus, the source hardening term (os) is equivalent to the grain size dependent term which can be determined from the grain size dependence of the yield stress:

ky

0 = d. [1.20c]

Alternatively, it can be evaluated using the Makin-Minter19 method by extrapolating the work-hardening portion of the stress-strain curve in Fig. 1.13a to the elastic range (Fig. 1.13b). The intercept is interpreted as the friction stress (o) and the difference between the yield stress and the inter­cept is the source hardening (os). Murty20 demonstrated the equivalence of Hall-Petch relation and Makin-Minter method from experimental results on grain size dependence of the mechanical properties of pure iron.

Effects of neutron radiation exposure in austenitic stainless steel (FCC)21 and mild steel (BCCp are shown in Fig. 1.14a and 1.14bi respectively. It can be seen that the smooth stress-strain curve in the unirradiated stainless steel (Fig. 1.14a) developed a distinct yield point subsequent to radiation exposure accompanied by a decrease in strain hardening and a decrease in the uniform and total elongations. The fact that the yield point and the Luders strain observed in unirradiated BCC mild steel (Fig. 1.14b) increases initially with increase in neutron fluence and eventually disappears after the highest value (1019 n/cm2) clearly demonstrates the decrease in source hardening with increased neutron radiation dose. On the contrary, in FCC metals the yield point appeared following radiation exposure indicating an increased source hardening in the irradiated material.

Friction hardening (o) arises mainly from the long range elastic interac­tions of moving dislocations with other (forest) dislocations as well as short range interactions with faulted dislocation loops, precipitates, etc., so that

°i = olr + osr = aGbiJp+рОЬ4ш, [1.21]

where the subscripts LR and SR represent the long-range and short-range stresses, G is shear modulus, b is the Burgers vector, p is dislocation density, N and d are the number density and diameter of the clusters (faulted loops, precipitates, etc.) and a and в are constants representing the strengths of long range and short range forces. In general the defect densities (p and N) are proportional to the fluence (or dpa) and thus

1 = a’Gbj^, [1.22]

Elongation (%)

1.14 ( a) Effect of neutron irradiation on stress vs strain curves for stainless steel (FCC) depicting the occurrence of yield points following radiation exposure.21 (b) Effect of neutron irradiation on stress vs strain curves for mild steel (BCC) depicting the absence of yield points following high neutron radiation exposure.22

where Ф is fluence (ф). The stress-strain curves shown in Fig. 1.14b on mild steel at room temperature as a function of fluence reveal that the yield strength varied as cube-root of fluence (Fig. 1.15) and not the square-root. This seems to stem from the fact that friction hardening (a) indeed varied

as square-root of fluence as noted in Fig. 1.16.23 Since os decreased with increase in fluence, the yield stress, being the sum of these two factors, should be a function of fluence raised to a power slightly less than 0.5. In FCC met­als such as stainless steel, source hardening is very small before irradiation

while it increases with fluence thereby resulting in small yield points at high fluences (Fig. 1.14a). It is believed that at high fluences saturation of radia­tion hardening occurs resulting in deviations from the square-root depen­dence of the hardening on the fluence.

Since the Luders strain in steels varies linearly with yield stress, it increased as cube-root of fluence (Fig. 1.15) where we note that the datum point at the highest neutron dose of 1.4 x 1019 n/cm2 is an extrapolation from high tem­peratures to ambient. Thus, at room temperature the highly irradiated mate­rial exhibited severe localized deformation and failed during Luders band propagation itself before reaching the strain-hardening regime. The increase in Luders strain and the decrease in source hardening, subsequent to irra­diation, imply that the work hardening should decrease as neutron dose increases. Indeed, the work-hardening exponent decreased from ~0.34 for the unirradiated mild steel to ~0.19 at a neutron fluence of 2 x 1018 n/cm2.23

The fact that the source hardening in BCC metals such as steels decreases on exposure to neutron irradiation implies that the concentration of inter­stitial C and N in solution decreases with increased neutron fluence. Murty24 examined the effect of incremental neutron dose on static strain ageing kinetics and demonstrated that the ageing kinetics are slowed and that flu — ences greater than 1018 n/cm2 rendered the steel non-ageing. In a correla­tion between the effects of neutron irradiation and dry hydrogen treatment, Murty and Charit25 demonstrated that the concentration of nitrogen in solu­tion decreases with neutron fluence, reaching a value very close to zero at

1018 n/cm2 (Fig. 1.17). These results imply that interstitial impurities com­bine with radiation-induced point defects such as vacancies and interstitials, either with individual defects or loops, to form complexes. These complexes are probably responsible for part of the increase in friction hardening and the corresponding decrease in solution hardening. McLennan and Hall26 found from internal friction experiments that the concentration of C in solution decreased by a factor of four in steels after irradiation to about

1019 n/cm2.

This is also the reason for the decrease in the intensity of dynamic strain ageing (DSA) in annealed mild steel, as depicted in Fig. 1.18a-1.18e, where the load drops in the stress-strain curves decreased with increase in radi­ation fluence, finally rendering the steel non-ageing after irradiation at 1019 n/cm2.27 It must be noted here that though radiation exposure results in reduced concentration of interstitial C and N in solution leading to reduced blue brittleness, radiation hardening and embrittlement can still occur. Thus the competing and synergistic effects of DSA and neutron irradiation could lead to increased ductility along with increased strength at appropriate temperature and strain-rates. Comparison of stress-strain for unirradiated material (~100°C) with those irradiated to different doses clearly reveals (Fig. 1.19)28 the typical embrittlement due to DSA in the unirradiated

material, whereas increased strength and ductility are noted following irra­diation to around 2 x 1018 n/cm2; but, after irradiating at the highest fluence level of 1.4 x 1019 n/cm2 the ductility decreased to around 2% with possi­ble fracture during Luders band propagation. These results are in contrast to those at room temperature where no DSA or blue brittleness is noted (Fig. 1.14b). At 100°C in mild steel where jerky flow started, the ductility decreased to 11% while it increased to ~20% following neutron irradia­tion to 1018 n/cm2. The fact that strength increased along with an increase in ductility implies that toughness (as defined by the area under the stress — strain curve) increases at temperatures where DSA is suppressed following radiation exposure. This is clearly shown in Fig. 1.20 which compares the toughness (J) for mild steel before and after neutron irradiation to 2 x 1018 n/cm2.28 Normal radiation embrittlement is noted at ambient temperature while an increase in toughness is observed at elevated temperatures follow­ing radiation exposure. The measured toughness is sensitive to the strain — rate of testing and a minimum toughness value is obtained when tested over a strain-rate range. This minimum in the unirradiated material occurs at higher temperatures for increased strain-rates and follows an Arrhenius relation (e = Ae~QIRTc where Tc is the temperature at which minimum tough­ness occurs) with the activation energy (Q) identifiable with that for diffu­sion of C and N in steel. Thus these synergistic effects of neutron irradiation

1.18 ( a) Stress-strain curves for mild steel at varied temperatures before irradiation27; (b) stress-strain curves for mild steel at varied temperatures following irradiation (3.9 x 1020 n/m2)27; (c) stress-strain curves for mild steel at varied temperatures following irradiation (2.8 x 1021 n/m2)27; (d) stress-strain curves for mild steel at varied temperatures following irradiation (2.0 x 1022 n/m2)27; (e) stress-strain curves for mild steel at varied temperatures following irradiation (1.4 x 1023 n/m2).27

1.19 Effect of neutron irradiation on stress-strain curves for mild steel at 373 K.28

and DSA could lead to beneficial effects on strength and ductility in certain temperature and strain-rate regimes.29 While these descriptions are limited to mild steels, such DSA and neutron radiation effects are also observed in steels used for nuclear reactor pressure boundary applications; we will discuss these in the next section under the radiation effects on radiation embrittlement of nuclear RPVs and support structures.

Kass and Murty30 have successfully used the Hall-Petch relation and fric — tion/source hardening concepts to explain the influence of fast and thermal neutrons, in the total neutron spectrum, on the grain size effects in pure iron and low alloy steels. They evaluated the effects of total and fast neutron spectra by irradiating samples with and without Cd-wrapping thereby elim­inating low energy (<~0.5eV) neutrons in the Cd-wrapped samples. The bar chart in Fig. 1.21 shows the effect of fast and total neutron radiation on the yield stress for Armco-iron and steels (1020, A516 and A588).30 All steels exhibited increased hardening due to the total neutron radiation exposure compared to only fast neutrons whereas in pure iron we note that as grain size decreases from 190 to 50 pm, exposure to the total neutron spectrum resulted in lower radiation hardening than only fast flux. This is explained on the basis of the Hall-Petch plot where the y-axis intercept represent­ing friction hardening increases with neutron radiation exposure while the slope decreases following fast neutrons. Exposure to the total neutron spec­trum with additional low energy neutrons would result in a further slight increase in friction hardening (or y-axis intercept) accompanied by a slight further decrease in the slope (i. e. decrease in the source hardening). This implies that the two lines (Fig. 1.22) representing the effects of fast and total (fast + low energy) fluences will cross over at a critical grain size below

1.21 Bar chart of yield stress for Armco-iron and steels (1020, A516 and A588) following fast and total neutron radiation exposures of 2.8 x 1018 and 3.4 x 1019 n/cm2, respectively.30

1.22 Effects of fast and total neutron fluences on Hall-Petch plots for pure iron.30

which (means larger values of the abscissa) fast neutron fluence would have a slightly larger effect on hardening than exposure to total flux.

Although ferritic steels such as those commonly used for pressure bound­ary and reactor support applications have very small grain sizes (<50 jam),

radiation hardening in these steels (1020, A516 and A588) is opposite to that noted in Armco-Fe of small grain sizes. A plausible explanation for this observation lies in the fact that the source hardening and the slope of the Hall-Petch plot are very small in these steels. Once these materials are exposed to fast neutrons, the hardening will essentially be due to fric­tion hardening with negligible contribution from source hardening thereby resulting in grain size independent yield strength (i. e. horizontal line in Hall-Petch plot). Exposure to the additional low energy neutrons along with fast neutrons is expected to result in a slight increase in friction harden­ing, thereby resulting in a Hall-Petch line parallel to that noted for fast flu — ence exposure alone. Hence, exposure to the total neutron energy spectrum in these steels will lead to an additional grain size independent hardening as observed in the bar chart in Fig. 1.21. Thus, identifying the yield stress as a sum of friction and source terms lends explanation and support to the experimental observations in both pure iron and steels.

Mechanisms

The first crack found in a steam generator tube was in a hot leg side tube of a steam generator in the Obrigheim plant in 1971. This was the first reported incidence of PWSCC at this site (Shah, 1992). As of 1994, at least 61 power plants had experienced PWSCC in the tubes; 32 plants experi­enced U-bend PWSCC and 5 plants experienced PWSCC in the denting zone. In most cases, it occurred in hot leg side tubes, but in some it occurred in cold leg side tubes. The cracks were found in 1-10 EFPY (effective full power year) for Alloy 600 LTMA (low temperature mill annealing), but they were detected only after 10 EFPY in Alloy 600 HTMA (high tem­perature mill annealing). However, some Korean power plants that have installed Alloy 600 HTMA in an explosive expansion method experienced circumferential SCC in three to seven years. PWSCC of the U-bend zone usually causes axial cracks. If there is mechanical expansion, axial cracks are primarily found, but in a French power plant, a circumferential crack was found in kiss roll expanded tubes in the sludge accumulation area.

Denting of Alloy 600 was first reported in the tube support plate (TSP) region in 1975 immediately after the secondary system water treatment changed from phosphate to all volatile treatment (AVT). Denting refers to the state where the cross-section of a tube does not maintain its original form due to the growth of oxides around the pipe. This damage mechanism was a major cause of tube plugging from 1976 to 1980, although more recently it has not been particularly serious. If chloride intrudes into the secondary sys­tem by steam condenser leakage and is concentrated between the tube and the support structure, it creates an acidic environment. When the oxygen content is high, corrosion of carbon steel tube support plates increases and a porous magnetite is formed which has double the volume of the base metal. The degree of superheat, presence of chloride and concentration of oxy­gen in the secondary system influences the corrosion rate of carbon steel. Copper oxides or copper ions can also be factors that supply oxygen to the water. Sulphate, like chlorides, can accelerate corrosion of carbon steel. The most serious cases occurred when seawater was used as the coolant of the steam condenser, such as in the RSG steam generators of Westinghouse and CE (Combustion Engineering in the United States). For a power plant that uses phosphate treatment, the environment remains alkaline so denting hardly occurs. Denting occurs rapidly, not occurring or spreading gradually like PWSCC or outer diameter stress corrosion cracking (ODSCC). When exposed to seawater containing chloride, denting occurs in a very short time (20 ppb acid chloride: 2.5 years), though when exposed to water containing neutral salt, it takes a more considerable length of time (20 ppb neutral chloride: 50 years).

The secondary IGSCC and IGA of Alloy 600 have been considered a serious corrosion problem since they were first reported in the early 1970s. These problems have been found in many freshwater cooling power plants, whereas fewer problems have been found in the power plants that use sea­water coolant. The causes of IGSCC corrosion are impurities concentrated by surface boiling due to coolant flow not being smooth in a certain area, as well as stress, materials, temperature, etc. IGA is different because it occurs when there is no stress; however, sometimes stress creates IGA.

Erosion-corrosion describes pieces of corroded metal peeling off the metal surface by the action of solid particles repeatedly colliding with the

, Stream outlet

Secondary side

Tube

supporl

plate

by S. S. Hwang of KAERI

7.2 Various types of corrosion found inside and outside the steam generator tubes.

metal surface when there is stable protection film on the metal surface. Depending on the size, shape and hardness of the particles and the cor­rosion environment, mechanical damage can be accelerated. If there was no protective film on the surface, only erosion of the metal would occur, but once the protective film is removed, both corrosion and erosion will be accelerated. Such corrosion occurs in OTSGs.

Figure 7.2 shows the diverse types of corrosion found inside and outside steam generator tubes.

Taking into account the specific features of the VVER-440/213 design of civil structures and also the lack of/missing analyses performed by the designer, eight analysis tasks were identified as necessary for the justification of LTO of the Paks NPP. The need to perform stress calculations for the valida­tion of the rather sparse information available for containment and other safety-classified structures was also recognized. Considering their content, these calculations are not typical TLAAs, however, without sufficient infor­mation on the design of civil structures the newly performed TLAAs would not have the design basis. The scope of TLAAs for Paks NPP includes the generic tasks, like: [19]

There are also several design-specific TLAAs in the case of the main reactor building at Paks NPP, for example:

• Fatigue analysis of the containment for increased pressure level during integral leak-tightness tests.

• Analysis of main reactor building settlement.

The allowable leakage value of VVER-440/213 containment is 14.7% per day at the design pressure of 2.5 bars. Each of the containments was tested at this design pressure in the start-up phase. The pressure of the yearly leak­age tests is 1.2 bar and tests at a pressure of 1.7 bars are also carried out during the outages. The leakage value for the nominal pressure of 2.5 bars is calculated via extrapolation from the leak rate results of tests. This practice has been criticized regarding correctness of the leak rate extrapolated from the measured ones, and investigations of enhancement of the test pressure level has been proposed. Nevertheless, the recent reduced pressure test pro­cedure has obvious advantages compared to the tests at enhanced pressure level: the time needed for the low-pressure test is short and the load on containment structures is moderate. According to the results of leak tests the correct leakage values at the nominal pressure of 2.5 bars can be deter­mined from the results of tests carried out at considerably lower pressure values. This statement is based on analyses of numerous tests, including the results of tests carried out at the design pressure of 2.5 bars at Paks NPP Unit 2.

Regarding Paks NPP, the analysis of settlement of the main building com­plex has been identified as a TLAA requirement since an excessive incli­nation of the main building complex due to differential settlement may result in non-allowed tilting of the RPV vertical axis, which may in turn cause problems with the control rods. Additionally, excessive inclinations can also cause extreme local loading resulting in degradations of the build­ing. It has to be mentioned that the VVER-440/213 type units at the Paks NPP have twin-unit-design, that is two main reactor buildings separated by a dilatation gap are built upon a common base mat. Detailed settlement control was started during the construction period at Paks. The measured results are to be evaluated and reported annually. A consolidation process, prolonged in time, was observed in case of the main reactor buildings, the settlement of which is still continuing. The phenomenon is related to the seasonal variation of the water level of the river Danube, which may reach a value of 9 meters. This variation of river water level influences the ground­water level. According to the data measured in the wells at and around the plant site, the groundwater level follows the variation of the water level in the Danube with a certain time delay. The water-table fluctuations influ­ence the stress-deformation conditions in the subsoil. This can explain the successive settlement of the raft foundation measured during past years. The settlement at Unit 4 is somewhat larger than at Units 1-3, which is due to the slight in homogeneity of the subsoil and the highest alteration of the level of the water-table, occurring in the vicinity of Unit 4.

Detailed analyses have been performed for the subsidence and differen­tial settlements of the main reactor buildings for the end-of-life situation taking into account the static loading (immediate settlement), groundwa­ter fluctuation, seismic settlement, dynamic settlement due to machinery and tectonic subsidence. The calculation model and procedure has been calibrated to the measured time-history of subsidence. An appropriate con­stitutive model has to be defined for the soil, which includes the develop­ment of a non-linear hardening model and proper definition of the decay curve for cyclic loading due to groundwater fluctuation based on soil tests results.

In regard to LTO, the analyses show that a value of differential settlement that may cause non-allowed tilting of the RPV axis due to the inclination of the building should not be expected. The structural integrity of the founda­tion and the containment part of the main building structures is not affected by the settlement and is not expected as a result of further subsidence.

In Zircaloy-4, the creep rate was reported to depend on the condition of the material — whether in CWSR or annealed condition; CWSR alloy showed a significant strengthening upon addition of hydrogen. The reason for this behaviour is attributed to hydrogen influencing strain-hardening rate and static recovery of the material. Biaxial tests in Zircaloy-4 showed that the presence of hydrides in the cladding will help to prevent the cold work microstructure from being annealed out of dislocations and thereby maintain lower creep rates in the spent fuel cladding.87 The same alloy in annealed condition showed an increase in creep rate when the hydro­gen is in solution and a decrease when part of hydrogen is precipitated as hydrides. This behaviour is attributed to the reduction in the stacking fault energy due to diffusion of hydrogen to the core of the screw disloca­tions and their increased mobility. On the other hand, when the hydrogen is present in the form of hydrides, it increases the matrix strength and reduces the creep rate. Other researchers noted that the increase in creep due to hydrogen content of around 200 wt ppm may be due to the reduc­tion in the modulus value when hydrogen was added.88 In a Zr-2.5 wt%Nb alloy, the creep rate at 723 K was reported to increase by 2-2.5 times for a hydrogen content of 160 wt ppm and the stress exponent reduced from 2.41 to 1.59, indicating a change in the creep mechanism.89

Since dry storage of spent fuel is gaining importance, it is necessary to assure clad integrity during interim storage. The high burnup rods are likely to con­tain large amounts of hydrogen (1000 ppm) and with a hoop stress of 100-120 MPa the clad should not creep to failure. In order to reduce the hydride prob­lem the initial level of hydrogen (and other impurities) is kept low and pickup during service is controlled by using new corrosion resistant alloys.

In the VVER-1000 models, all primary circuit surfaces are either made from, or are clad in, stainless steel. The 08X18H10T type stainless steel (08Cr18Ni10Ti, AISI 321) is used for the core structures, main circulating pumps and steam generator tubing, whilst the main loop pipework and steam generator collectors are manufactured from 10GN2MFA type carbon steel and the cladding is made from 08Cr18Ni10T stainless steel. The pres — surizer is also made from 10GN2MFA carbon steel, covered by cladding, with an inner layer of Sv-07Cr25Ni13 (similar to AISI 309) stainless steel and two layers of Sv-08Cr19Ni10Mn2Nb niobium stabilized stainless steel (similar to AISI 347). The reactor pressure vessel and head is made from the low alloy steel 15Cr2MNFA. The cladding of the reactor head has an inner layer of Sv-07Cr25Ni13 stainless steel and two layers of the niobium stabilized stainless steel Sv-04Cr20Ni10Mn2Nb (again similar to AISI 347). The phosphorus and copper contents in the welds of VVER-1000 RPVs are 0.005-0.014% and 0.03-0.08%, respectively.

It has been recognized that the standard surveillance programmes for VVER-1000/320 reactor pressure vessels have some deficiencies related to the design of the surveillance assemblies, for example the non-uniformity of neutron field within individual specimen sets, large gradient in neutron flux between specimens and containers, lack of neutron monitors in most of containers and no suitable temperature monitors (Brumovsky and Zdarek, 2005). The location of surveillance specimens does not assure similar condi­tions as the beltline region of reactor pressure vessels. A modified surveil­lance programme for VVER-1000/V-320C type reactors was designed and implemented at the Temelin NPP in the Czech Republic. The technical fea­tures of the surveillance test assemblies provide opportunities for implemen­tation of an integrated surveillance programme, using samples from several VVER-1000 units: Temelin 1 and 2 (Czech Republic); Belene (Bulgaria); Rivne 3 and 4, Khmelnitsky 2 and Zaporozhie 6 (Ukraine); and Kalinin 3 (Russia). Irradiation of these archive materials together with the a refer­ence steel JRQ (of ASTM A 533-B type) and reference steel VVER-1000 allowed a comparison of the irradiation embrittlement of these materials, and an opportunity to obtain more reliable and objective results, as no reli­able predictive formulae exist up to now because of a higher nickel content in the welds. Irradiation of specimens from the cladding region will help in the evaluation of resistance of pressure vessels against PTS regimes.

Several mitigation measures have been identified for the VVER-1000 RPVs. Based on the fracture mechanics analysis, heating up the hydro-accumulator water to 55°C was recommended to prevent injection of ECCS water with temperatures below 20°C for all the plants. The use of low neutron leak­age core loading patterns in VVER-1000 reactors would reduce RPV wall fluences by approximately 30%. For reducing the neutron flux on the reac­tor vessel, low leakage core design was introduced at some plants (i. e. fuel assemblies with high burn-up to be placed at the core periphery). In addition, the quality of manufacturing and alloy composition ensure the possibility of LTO for VVER-1000 reactors (Vasiliev & Kopiev, 2007).

The steam generators for VVER-1000 have been designed on the same principles as the VVER-440 plants, however the SGs at VVER-1000 plants are replaceable. At some units, throughout the design service life of the SG, there were problems resulting in necessary SG replacement. At the same time, the SGs at some plants could be operated beyond design service life. As operating experience has shown, it is the water chemistry of the second­ary circuit that is the main factor influencing operability of the SG tubing, as in the case of VVER-440 plants. Tube integrity is inspected by the eddy current method; the results of the testing can be used to determine the plug­ging criterion for defected tubes. Proper definition of the plugging criterion is an important challenge.

The ageing problems of the SGs at VVER-1000 plants are as follows (Trunov et al, 2006a): [13]

• corrosion of the heat-exchange tubes

• formation of deposit

• difficulties in measuring and regulating the SG water level.

A study performed by the International Atomic Energy Agency summa­rizes the status of knowledge on steam generator ageing: TECDOC-1577 (IAEA, 2007b).