The use of (ppm level) niobium impurities in reactor pressure vessel (RPV) steel for neutron flux estimations based on the reaction 93Nb (n,n′) 93mNb has been reported previously [1]. The method has now been further investigated and refined. Small niobium fractions in RPV steel (≈ ppm) and plating (≈1 %) materials have been separated by ion exchange chromatography in one to three steps. The measured Nb fractions in samples from some four pressure vessel (RPV) base materials were 1 to 3 ppm. The purification of tens of milligrams of RPV material provides sufficient amounts of niobium for mass determination with a highly sensitive (10−5 ppm) Inductively Coupled Plasma Mass Spectrometer (ICP-MS). The 93mNb and small remaining 54Mn activities were measured with a calibrated Liquid Scintillation Counter (LSC) based on dual label technique and almost 100 per cent efficiency to 93mNb. One purification is needed for plating materials (≈1 % Nb) and two purifications of about one gram of steel with Nb impurities in order to resolve the needed activities (≈10 Bq 93mNb/ug Nb). Other neutron induced reactions in Mo produce interfering 93mNb, 92mNb and 95Nb. With about 0.5 per cent Mo in the RPV steel, several tens of ppm of Nb (Nb/Mo > 0.01) are typically needed in order to obtain a clearly dominant 93mNb activity from the 93Nb (n,n′) 93mNb reaction of interest. RPV plating materials with 0.5...1 % of alloyed Nb have insignificant degradation from Mo on the measured 93mNb activities. Thus, a sufficient chemical purification of Nb was achieved and the specific 93mNb activities could be measured for all types of available RPV and plating samples. However, other competing and 93mNb producing reactions such as 92Mo(n,γ)93Mo (3500 y), disintegrating to 93mNb, strongly degraded the results in samples with low Nb/Mo ratios. Samples with sufficiently high Nb/Mo ratios may be eventually selected for dosimetric purposes since the Nb and Mo mass fractions vary with position and type of material. The achieved accuracy of the measured specific 93mNb activities was about ±3 % (lσ) in irradiated RPV plating materials and about ±4 % for Nb ppm impurities.

No abstract

A comprehensive re-evaluation of all neutron dosimetry results for the Loviisa reactors (WER-440, 2 units) started in 1992. Revised and better validated neutron spectra, a refined semiem pirical approach utilizing the computer program PREVIEW [4], unified and up-dated nuclear data and refined and extended calibrations of the applied Liquid Scintillation Counter (LSC) for 93mNb activity measurements are applied. The kernel-based computer program PREVIEW makes it possible to take into account the detailed local irradiation history e.g. the influence of the calculated time-dependent axial flux changes during a cycle. The main features of this re-evaluation are reported together with updated results for two IAEA-CRP surveillance chains (SC) denoted “4K” and “5K”. The fluence (>1 MeV) estimates increased by +16 to +20% and the dpa estimates by +24 to +27% as compared to previous results. The largest changes, about +20% in the >1 MeV fluences, arose from the revised and better validated neutron spectra. The refined treatment in PREVIEW of time-dependent variations fin the axial flux distribution accounted for 0 to −5% in the fluence estimates based on 54Fe (0 to −16% for 58Co). Nb, Fe, Ni, Ti and Cu dosimeters were used for validations of the calculated neutron spectra built into the PREVIEW kernel library. The quality of the neutron fluence estimates could be signifi cantly improved by applying revised and better validated neutron spectrum shapes and by applying a refined semiempirical approach taking into account calculated axial flux changes during an operating cycle. The remaining significant uncertainties in the neutron spec trum shapes and cross sections emphasise the development and applica tion of additional and more reliable dosimeters for the lower fast neutron energies.