Navigation path

Evaluation of Reactor Pressure Vessel Embrittlement of South Ukraine NPP including embrittlement aspects

Evaluation of Reactor Pressure Vessel Embrittlement of South Ukraine NPP including embrittlement aspects


The VVER 1000 reactors (19 reactors in operation) represent the most modern East- European reactor type, where the irradiation conditions of the reactor pressure vessel (RPV) are close to those of the western reactors, that is, the fluence rate at the RPV wall is relatively low due to the large water gap between the core and the RPV, and the irradiation temperature is high. The contents of impurities like P and Cu, which normally account for a great deal of the irradiation embrittlement of the beltline materials, have also proved to be very low both in the base and weld metals. These materials also exhibit low transition temperatures, both Charpy and To (transition temperature corresponding to 100 MPa√m) in the initial condition, partly owing to the nickel used as an alloying element in these steels.

South Ukraine NPP (SU) was interested in performing integrity assessment of the RPV of the 3 operating units, taking into account irradiation embrittlement of the core region material. The material of the VVER 1000 RPV differs from the VVER 440 type of RPV by the addition of Ni. Ni is known to enhance embrittlement due to neutron irradiation. Therefore it was of interest of the utility to clarify this problem as SU NPP is concerned. In order to do this the surveillance and other experimental results had to be reviewed, the irradiation embrittlement properties had to be evaluated, Pressurized Thermal Shock (PTS) and Pressure-Temperature (P-T) limits were to be analysed for the RPVs.

The objectives of the project were to:

  • Collect, review, evaluate and document the existing SU RPV material property data;
  • Check the origin and quality of the surveillance specimens;
  • Analyse the surveillance test results and compare to predictions;
  • Perform neutron fluence calculations;
  • Carry out PTS and P–T limit studies for the RPV in order to ensure integrity;
  • Evaluate the embrittlement rate of the RPV core region material and the safe life time of the RPV;
  • Propose measures to mitigate irradiation embrittlement and reduce the risk of vessel failure.

3)Project results–achievements

Base line material data (Materials Certificate), Task 3
The initial material properties of the RPVs of the SU NPP measured by the RPV manufacturers were reviewed at the NPP. The available data include the chemical composition of the three core shells and the two circumferential core welds of the RPV. Tensile test results were available from only one core ring of units 1 and 2 and two core rings in Unit 3. Only tensile test results for both core welds of all three units were available. Only little information on toughness properties (Charpy V and Fracture Toughness (FT)) were found because only a few specimens are needed for confirming, that the Tko is below specified value. The reported Tko values were lower or equal to the specified ones. Because no impact test transition curves or fracture toughness results were available, it was not possible to properly evaluate the exact Tko values or to compare the FT behaviour of different parts with each other based on the baseline certificate data.

The measured chemical compositions were in accordance with specified requirements The material specification allows relatively large ranges of variation in the Ni contents, i.e. 1.0...1.5% for the base metals arid 1.0...1.9% for the weld metals. In the base metals the Ni contents seem to be below 1.3 in all units. In Unit I the Ni content of weld 4 was only 1.2%, whereas in the other welds the Ni contents were higher, i.e. 1.7-1.8%.

All tensile test results of the core shells, which were chosen for surveillance programmes (lower belt rings I and also support ring III of Unit 3), satisfy the specification requirements. In the case of core weld metals and welded joints all test results fulfil the requirements of the specifications.

Evaluation of irradiation embrittlement (Surveilance program), Tasks 2, 4, 5, 6, 7 and 9
In this tasks available surveillance data of South-Ukraine I and 3 were collected and the origin of the materials checked from the available documents. Most of the data (excluding fatigue test results) were tabulated and analysed. No surveillance data was available for Unit 2. The data consisted of the tensile, Charpy and fracture mechanics test results which were provided for this project by the South-Ukraine NPP. The transition curves and the transition temperatures (Tk) were calculated for the materials. The fracture mechanics test results were analysed using the statistical cleavage fracture model of VTT. The surveillance programmes of the South-Ukraine 1 and 3 had been carried out according to the Russian Guide.

The tensile test results showed some irradiation embrittlement response but this data were not analysed in detail because accurate neutron fluences were not available. In general, the Charpy test data showed Tk values lower than specified, but the scatter was large, mainly due to inaccuracies in the given neutron fluences. The specimens had mostly been irrdiated in the standard surveillance positions of VVER 1000 reactors, i.e. in the upper part of the reactor, which leads to high variability in the specimen fluences and low fluence rates. Since for most data series the specimen fluences had not been measured, the capsule fluence was first estimated from the fluences of the neutron dosimeters, given by the South-Ukraine NPP, and this mean fluence was then used for all specimens of the capsule. This method is, however, not accurate enough especially when the fluences are low and there exits large variation in the neutron flux over the capsule.

Due to the above uncertainties, it was difficult to give estimations for the embrittlement rates of the materials studied in the project. As no distinct violations of the specified trend curves or, chemistry factors (Af) were observed on any of the investigated materials, applying the mean values of Af there is no evidence on the basis of the present surveillance data of South-Ukraine land 3 that the RPV beltline materials would not behave in the way specified in the Russian Guide. Due to the uncertainties one cannot, on the other hand, conclude from this surveillance data that the present trend curves would be conservative for these materials. No estimations for the expectable irradiation rates or predictions for end-of -life Tk are therefore given here. Such estimations could, however, be made also from the maximum values of Af, but due to the reasons discussed above this procedure would probably lead to unnecessary conservative estimates. The values of Af obtained in this project for the South-Ukraine 1 and 3 RPV materials should anyway be regarded as very tentative.

The following conclusions can be made from the material test results produced in the surveillance programmes of the South-Ukraine 1 and 3 included in this study:

  • The initial mechanical properties and Tko values of the surveillance materials differ somewhat from those of the materials tested previously by Izhora and Atommash (above Task 3). The differences are in part caused by the different determination procedures used.
  • The surveillance test results show large scatter in the 47 J and 100 MPa√m transition temperature shift values. This scatter is most likely caused by the uncertainties associated with the given specimen fluences, resulting from the high fluence rate gradients during the irradiation and the subsequent variability in the specimen fluences (mostly only mean fluences were available).
  • Due to the low specimen fluences, the dependence of transition temperature Tk on the fluence cannot be reliably estimated for either of the RPVs (for which surveillance data were available) from the present test results. According to the fracture mechanics test results of the weld and base metal specimens irradiated in the lower position of the capsule channel, the embrittlement rate can be relatively high in both the base and weld metals, i.e. approximately that specified by the Russian Guide for these materials.
  • The irradiation temperatures of the surveillance specimens are not known accurately enough. Temperature differences of 0-30C are possible, and the effect on the embrittlement rate is not well known. The uncertainty should therefore be taken into account by an extra margin. According to the surveillance test results the true initial transition temperatures (Tko) of the RPV base and weld metals are likely well lower (i.e. better) than the limits specified for the beltline materials.
  • The transition temperatures of the RPV base and weld metals at the end of the planned operating time cannot be reliably estimated from only the surveillance test results presented. They consisted mostly of low-fluence data. Similarly, the conservativity of the specified trend curves, i.e. the fluence dependence of Tk, cannot be judged.
  • The shift in To (100 MPa√m) due to irradiation can be lower or higher than the shift in Tk (47 J). The shift of To due to thermal ageing (at 290 C) is higher than the shift of Tk.
  • The effect of thermal ageing on the embrittlement should be taken into account, e.g. with an extra margin, when long-term effects are evaluated.

The uncertainties discussed above mainly result from generic shortcomings in the surveillance programmes of VVER 1000 NPPs.

Neutron fluence calculations, Task 10
Neutron fluence was calculated for the Unit 3 RPV of the South Ukraine NPP considering fuel loading cycles 1 to 6. The DETA code was used for the calculation. The code is designed for calculations of fast neutron field characteristics in core and near vessel space of the VVER 1000 type reactor. The calculations are performed by Monte-Carlo method in multi-group approximation of neutron transport theory. In the calculation a symmetry sector of 60 degree at the middle level of the core was considered. For definition of displacements per atom (dpa) the neutron spectra in some points of the RPV were calculated. The results of calculations were compared with experimental data during the 6th fuel loading cycle.

Comparison of experimental and numerical results showed a rather good correspondence. The calculated total fluence above energy E > 0.5 MeV varied on the RPV inside surface depending on location from 2.48 to 6.27*1022 n/m2 for the first six fuel loading cycles of Unit 3. The 1st fuel loading was performed 21 September 1989, the 6th loading 21 December 1994 and it was in use up to 2 October 1995.

Pressure-Temperature limits, Task 10
The pressure-temperature limits were determined according to the Russian Guide and the calculation of the stress intensity factors according to ASME XI, Appendix A. The initial toughness for the welds and base metals given in the Russian Guide were used in the analysis. The analysis was conducted for heat-up and cool-down transition modes. The neutron fluences used in the analysis corresponded to the end of life condition (40 years of operation; The neutron fluence used in the analysis corresponded to the end of life condition (40 years of operation, that is 5.7*1023 n/m2 (E > 0.5 MeV) for the lower beltline shell and weld 3, and 4.5*1023 n/m2 (E > 0.5 MeV) for weld 4. The analysis demonstrated that the pressure margin between the operation values and limiting values is sufficient in all above mentioned modes.

Pressurized Thermal shock, Task 10
As no information concerning pressurised thermal shock (PTS) analyses were available at SU NPP, an additional PTS analysis was performed at VTT. The computation was performed for one severe transient, which causes strip-like cooling conditions with a coolant level simultaneously in the reactor pressure vessel. The relevancy of the transient depends on the thermal hydraulic details of the examined nuclear power plant unit, which were not examined in the work. The material properties Tko and Af for the selected areas of the RPV were taken from the Russian Guide, since real values were not available for this unit, as mentioned above.

The thermal and mechanical computations were made using the Abaqus 5.5 finite element code. Stress intensity factor was calculated using a weight function method. The computations were performed for two crack depths, corresponding to 11 % and 25 % of the wall thickness (cladding excluded). The crack aspect (a/c) ratio was 2/3.

For each of the three locations examined, the most critical case (crack depth/orientation) was plotted against the fracture toughness, emergency condition curve of the Russian Guide. The allowable value for the transition temperature (Tka) was determined so, that the computed crack loading curve was tangent to the fracture toughness curve.

The results are shown in Table 1. It compares the computed allowable values of transition temperatures (Tka) values with the fracture toughness (Tk) values. An extra margin of 10 C due to thermal ageing has been added to the Tk. The neutron fluence used in the analysis corresponded to the end of life condition as above for P–T limit calculations (40 years of operation, that is 5.7*10 (23) n/m2 (E>0.5 MeV) for the lower beltline shell and weld 3, and 4.5*10 (23) n/m2 (E>0.5 MeV) for weld 4). According to the Table the allowable fracture toughness is exceeded in both core welds 3 and 4, which means, that there is no margin against brittle fracture at the end of life condition.

If the value of fracture toughness (Tk) is limited to the allowable values, 73 C for weld 3 and 54 C for weld 4. The corresponding allowable fluences of 3,4*10 (23) and 1,1*10 (23) n/m2 (E>0.5 MeV), respectively. The latter is less than 1/4th of the end of life fluence which means 10 years.

Table 1. The fracture toughness (Tke) values corresponding to 40 years of operation (end of life condition) according to Russian Guide at the depth of 20.6 mm. The calculated allowable (Tka) values according to the performed PTS analysis.

PartTke (20.6mm) (C)Tka (C)
Lower beltline shell 60 < 84
Weld 374> 73
Weld 468> 54

If applying the real measured baseline material properties instead of the specified ones (0 C) as in above study, the results will be fully acceptable. The Tko for both welds was - 65 C and the specified value is 0 C. Accordingly the Tke value for the welds will be + 9 C? for weld 3 and respectively + 3 C for weld 4, which is much less than the above limit values 73 C respective 54 C. Nevertheless, the results of this study showed that when using specified material properties and irradiation embrittlement factors according to Guidelines, the result would be unacceptable in this transient as shown in above the table 1.

The following mitigation measures to reduce the risk of brittle failure of the RPV in PTS were proposed:

  • Reduce neutron fluence by core loading modifications (low leakage core etc.);
  • Annealing of the core region of the RPV to recover the toughness (as in VVER 440 NPPs);
  • Reducing stresses in a possible PTS by warming the ECCS water.

This project was extremely difficult since just little information and test results from the surveillance programs of the selected NPPs were available. The neutron fluence of the 1st sets of surveillance specimens is so small, that practically no shift (within scatter tolerance) in transition temperature can be observed. Furthermore due to the large variation in neutron flux in one set of surveillance specimens a proper transition curve with sufficient amount of test specimens (> 6 specimens) with equal neutron fluence cannot be provided. This is a typical and generic shortcoming of the surveillance program of the VVER 1000 NPP.

Since the completion of this Tacis project new findings regarding the surveillance program of the VVER 1000 RPV have been reported. In Tacis 2.06/96 “Surveillance program of VVER 1000 NPP” it was confirmed, that the irradiation temperature of the surveillance specimens is < 10 C above the operation temperature. Accordingly, the surveillance test results can be considered reliable and a better confidence can be given to received results.

Two new Tacis regional projects: Tareg 2.01/00 “Validation of neutron embrittlement for VVER 1000 & 440/213 RPVs, with emphasis on integrity assessment” and 2.01/03 “Embrittlement assessment and validation of embrittlement and re-embrittlement models for VVER reactor pressure vessels” have been launched in 2003 and 2005. The aim of the Tacis regional projects are to improve surveillance test results in Russian and Ukrainian VVER NPPs by optimizing reconstitution of tested surveillance specimen halves and conduct re-testing in selected test temperatures. The aim is at increasing the number of test results for selected neutron fluence levels in order to have the requested amount of test results for elaborating toughness curves according to Russian and Ukrainian Guidelines. Based on the improved test results, upgraded neutron fluence calculations and databases new PTS studies will be carried out in order to confirm the integrity of the RPV of VVER NPPs.

Further Information
Further information on the project results could be sought from the EC archive in IE-JRC in Petten or the beneficiary organizations in Ukraine.

General Information

Evaluation of Reactor Pressure Vessel Embrittlement of South Ukraine NPP including embrittlement aspects
€ 140.493,24
Budget year: 
Types of activities: 
Design safety
DS - Design safety (incl. Off-Site Emergency Preparedness)
South Ukraine
Duration (months): 
Contracting authority: 
European Commission
VTT (Technical Research Centre of Finland)
CRIS number: 
Old reference: 
Project reference: 
U1.02/92 A2
Decision number: 
Method of procurement: 
Informal Consultation
Signature date: 
Effective contract date: 
Contract end date: 
Closure date: