PRESSURE VESSEL INVESTIGATIONS OF THE FORMER GREIFSWALD NPP: FLUENCE CALCULATIONS AND NIOBIUM BASED FLUENCE MEASUREMENTS

摘要

Pressure vessel integrity assessment after long-term service irradiation is commonly based on surveillance program results. Nevertheless, only the investigation of RPV material from decommissioned NPPs enables the evaluation of the real toughness response. Such a chance is given now through the investigation of material from the former Greifswald NPP (VVER-440/230) to evaluate the material state of a standard RPV design and to assess the quality of prediction rules and assessment tools. The operation of the four Greifswald units was finished in 1991 after 12-15 years of operation. In autumn 2005 the first trepans (diameter 120 mm) were gained from the unit 1 of this NPP. Some details of the trepanning procedure will be given. The paper mainly deals with the retrospective dosimetry based on Niobium, which is a trace element of the RPV material. The reaction ~(93)Nb(n,n')~(93m)Nb with an energy dependence highly correlated to radiation damage and a half life of the reaction product of 16.13 years is well suited for retrospective fast neutron dosimetry. Fluence calculations using the code TRAMO were based on pin-wise time dependent neutron sources and an updated nuclear data base (ENDF/B-VI release 8). The neutron spectra were determined at the trepan positions. The different loading schemes of unit 1 (standard and with 4 or 6 dummy assemblies) were taken into account. The calculated specific ~(93m)Nb activities for February, 2006 at the sample positions were determined to 16.3 Bq/μg Nb for sample 1, (0.1cm distance from inner wall), and 4.0 Bq/μg Nb for sample 2 (11.5 cm distance from inner wall). Unfortunately, a second neutron reaction besides ~(93)Nb(n,n') leading to ~(93m)Nb-activity is the reaction ~(92)Mo(n,γ)~(93)Mo. ~(93)Mo decays by electron capture to ~(93m)Nb with a half life of 4000 years and a branching ratio b_r=0.88. As (n,γ)-reactions are produced mainly by low energy neutrons, being less important for material damage, the ~(93m)Nb-activity generated through the Mo-path should be determined separately and subtracted from the measured activity. For the . sample 1 in the maximum flux azimuthal position of weld SN4 with a Nb-content of 8 ppm and an Mo-content of 4000 ppm for February 3, 2006 was obtained a Mo-induced ~(93m)Nb-activity of 80 Bq/g steel, amounting to 37.7% of the total ~(93m)Nb-activity. It turns out that the ~(93m)Nb generation on the second path is nearly of the same order as the fast neutron induced generation from Niobium. For the experimental determination of the ~(93m)Nb-activity the Nb-content was determined by ICP-MS (inductive coupled , plasma mass spectrometry) after dissolution of the material sample. The radiochemical isolation of Nb was done by anion exchange separation. The radiochemical separation was accompanied by determination of the chemical yield of Nb using again the ICP-MS method. The measurement of the ~(93m)Nb activity was realized by Liquid Scintillation Spectrometry (LSC). The first comparison between the calculated and the measured ~(93m)Nb activities resulted in deviations between 15 and 50%. Possible reasons for the observed differences are discussed.