APPLICATION OF CDZNTE DETECTORS FOR CONTROL OF INITIAL ENRICHMENT OF FRESH NUCLEAR FUEL AT REFUELING

The task of determining the initial enrichment of nuclear fuel (NF) at nuclear power plants in real time is relevant for both economic indicators of operation of the power unit and for all components of nuclear safety (nuclear safety, security and safeguards). The technology used by nuclear fuel cycle companies is not suitable for use at nuclear power plants. It is proposed to use a measuring system based on CdZnTe semiconductor detectors designed to control the burn-up depth of nuclear fuel for estimating the initial enrichment of fresh nuclear fuel in fuel assemblies (FA). The initial enrichment is estimated with help of the gamma radiation of uranium isotopes of fresh nuclear fuel. A distinctive feature of the system that has been developed is the use of passive tomography technology to take into account the self-absorption of radiation in the FA. The promise of the proposed method for estimating initial enrichment is confirmed by experimental measurements and simulation results. It is shown that the system developed makes it possible to determine the enrichment of nuclear fuel, taking into account its profiling by the volume of FA. Also, the results of measurements confirmed the possibility of using the proposed technical solutions for CdZnTe detectors to create portable monitoring systems for special nuclear material in the field. The solution of this task is of great importance for national security.

Full control over the state of nuclear fuel, at all stages of its life cycle, at nuclear power plants necessarily requires an instrumental assessment of the initial enrichment of fresh fuel before it is loaded into the reactor core. Nuclear fuel (NF) is a mixture of 235 U isotopes with a range of content of 1.6-4.4 %, therefore the initial isotope composition of nuclear materials is often the object of measurements. However, there are no such systems for determining the initial enrichment of nuclear fuel at nuclear power plants. Therefore, at present, it is necessary to rely solely on the data declared by the manufacturer. Knowledge of the actual deviation of the initial enrichment is of great importance in order to optimize the fuel permutations, especially when operating at the final stages of the campaign.
The determination of the initial enrichment of NF (i.e., the content of the fissile isotope 235 U) in real time can be carried out through passive radiation measurements of the sample's self-radiation. Like all passive measurements, this method of analysis is non-destructive [1]. The determination of uranium enrichment level in samples is a key measurement in technological processes and in the production control on the uranium enrichment and nuclear fuel companies, it acts an important part in the international nuclear safeguards inspections to grantee that uranium fuel is used for peaceful purposes. The principles of enrichment measurement can be used to determine the content of any isotope if their radiation characteristics are known and some special measurement conditions [2,..,5] are met.
The main radiation used in passive non-destructive analysis of uranium samples is gamma radiation. When measuring the enrichment of 235 U, the gamma radiation line with a gamma-ray energy of 186 keV is most often used. It is the most distinguished single gamma-emission line for any uranium samples enriched by 235 U above the natural level.
The method for determining the enrichment of uranium by gamma radiation was for the first time used to control the gas cylinders with UF6 [1,4,6]. The method of the measurement included recording the radiation of the uranium sample through the channel of the collimator by the detector of the gamma radiation. The enrichment was determined by the intensity of gamma radiation of the 235 U isotope with 186 keV energy. If the sample is thick enough, the detector detects gamma rays with an energy of 186 keV only from an insignificant part of the total sample, because of the strong absorption of low-energy quanta in the fuel assembly materials. In this sense, in the terminology of [1], fuel element is a "thick sample". The "visible volume" of the sample is characterized by the configuration of the collimator, the detector geometry and by the length of the free path of radiation with an energy of 186 keV in the sample material [1,3,4].
We propose a measuring system based on CdZnTe detectors [7, 8,9], created for the burn-up analysis of spent nuclear fuel, can be used to estimate the initial enrichment of an overloaded fresh fuel. In this case, the main problem may be the low detection efficiency of the detectors used. To test the possibility of using CdZnTe detectors, fuel pellets with an initial enrichment of 4.4 % were measured. One of the obtained spectra is shown in Fig. 1 and Fig. 2, using different scales of the axes for displaying the count results (i.e. measured intensity) in the channels of the multichannel analyzer. It is done to provide greater visibility of the peak of the total absorption with gamma radiation with energy of 235 U 205.3 keV. In the spectrum, well-identifiable lines of 186 keV and 205 keV of selfradiation of 235 U are observed. The lines of sample self-radiation of 238 U with energies of 742 keV, 766 keV, 786 keV and 1001 keV are visible as well. The 1001 keV line is usually used for analytical purposes. For comparison, Fig. 3 shows the spectrum of the self-radiation of a metal sample of depleted uranium. The mass of uranium in both series of measurements was approximately the same. It can be seen that the analytical lines of the 235 U isotope are absent, and the radiation intensity is much lower. In both cases, the value of the parameter of the thickness of the sample is the "infinite thickness".
The determination of the initial enrichment using the 235 U self-radiation intensity measurements with 186 keV energy is carried out according to the following formula [1,11]: , (1) where is the net counting rate at the peak of full absorption; and are the total counting rate in the peak of the total absorption and the counting rate of the background and the continuous component of the Compton distribution, respectively; are the experimentally determined coefficients for taking into account self-absorption and scattering of radiation in the fuel matrix material and fuel cladding walls.

Fig.3 -Spectrum of the self-radiation of a metal sample of depleted uranium
Such a method is fairly simple when carrying out measurements of typical homogeneous samples, for example, in the measurement of barrels containing uranium dioxide. However, FA is an object with a complex geometry and in this case it cannot be considered as a "thick" bung. As a rule, it can be assumed that the main contribution to the counting rate of the full-absorption peak with 186 keV energy is given by one, or at most two rows of external fuel rods. To increase the recording efficiency and reduce the measurement time, the detector should be positioned as close as possible to the fuel assemblies, but in this case, there are observation points where so-called "firing" from fuel rods located in the inner rows are recorded, while the gamma radiation intensity is very uneven depending on the viewing angle [10]. An example of such a phenomenon, obtained by the modeling described below, is shown in Fig. 4. This leads to a strong dependence of the obtained results on the location of the detector relative to the measured FA. This effect is even more pronounced when measuring fuel assemblies containing fuel elements with different enrichment (for example, fuel assemblies marked "ED" contain an external row of fuel rods with an enrichment of 3.6 %, and internal fuel rods have an enrichment of 4.4 %).
To reduce the measurement error, we took into account the effect of self-absorption and scattering of radiation in the material of the fuel matrix and fuel cladding walls, enrichment can be determined on the basis of data on the ratio of the counting rate at the peak of full-energy absorption of 186 keV of the 235 U self-radiation and the counting rate at the peak of full absorption of 1001 keV of self-radiation of 238 U. Fig 5 shows the results of measurements of gamma-radiation intensity of one (gray bars), two (black bars), and three (white bars) layers of VVER-1000 fuel pellets. It is clearly seen that an increase in the thickness of the measured layer leads to a significant increase in the emission intensity of the 1001 keV line of self-radiation of 238 U and secondary X-ray radiation due to the interaction of the primary radiation with the fuel matrix.  The dependence of enrichment on the ratio of the self-radiation intensities of 235 U and 238 U for FA will obviously be nonlinear due to the different absorption of gamma quanta with energies of 186 keV and 1001 keV. For the empirical determination of this dependence, it is necessary to carry out a sufficient number of measurements for a fuel assembly with a different enrichment in a statistical sense, which in principle is not a problem under NPP conditions [10,12]: To confirm the possibility of estimating the initial enrichment of nuclear fuel, a twodimensional plane model of the field of the gamma radiation of uranium isotopes contained in the fuel assemblage was constructed, according to which the location of the detector at the n-th observation point at a distance Rn from the FA axis the gamma-radiation of j-th isotope of uranium with energy Eγ at the location of the detector is: (2) where is the activity of the -th uranium isotope for -th fuel element taking into account the actual value of enrichment; is the quantum yield of the -th gamma line for the -th isotope; is the coefficient of the contribution of the -th fuel element to the intensity of the gamma radiation of the -th isotope with energy , taking into account the effects of attenuation when the gamma-ray beam propagates from the -th fuel element to the -th point of observation, .
Taking into account the registration efficiency, the measured intensity (counting rate) for an isotope with energy the measured gamma radiation intensity of the -th isotope with energy is expressed as . (3) In the future, we omit the indices jk , since we are talking about the intensity of gamma radiation at the point of registration of the -th isotope with energy , i.e. in the simulation it is assumed that the peaks of the gamma spectrum are isolated and processed.
The attenuation in the propagation of gamma radiation is mainly connected with two physical mechanisms [13]: geometric attenuation associated with the spatial divergence of the beam , where is the active area of the detector; is the distance from the center of the -th fuel rod to the -th location of the detector; -physical attenuation of gamma radiation during passage through various media (in particular, when passing through UO2 in other fuel elements located on the path of gamma-ray run to the detector, in their zirconium shells, and also in the water environmentboth inside the FA and outside of it, in the overload pool); the physical attenuation of radiation from the -th fuel element is expressed as , (5) where is the coefficients of linear attenuation of gamma quanta with energy for the corresponding media; is the length of run of gamma radiation in a medium of type (uranium dioxide, zirconium, or water).
Thus, the contribution coefficient of the -th fuel element for the -th position of the detector, associated with attenuation, is . (6) According to the formulated model, the fields of gamma self-radiation with an energy of 185 keV of the 235 U isotope and with an energy of 1001 keV for 238 U at different locations of the detector were calculated. All the results presented below are calculated from the relations (2), (3), (6) for the "FA axis -detector" distance of 22.5 cm. This is because of the fact that the developed equipment for monitoring the fuel burn-up depth was specially designed for placement on the working bar of the reloading machine (RM), and the specified size is determined by the diameter of the working rod of the RM. The angular diagram of the intensity distribution of gamma radiation with an energy of 185 keV 235 U is shown in Fig. 6. It can be seen that the intensity distribution practically repeats the shape of the fuel assembly. At separate viewing angles, there are irregularities and "sharp increases", the magnitude of which is small and does not exceed 7-8 % of the mean. There are smooth maxima at the angles corresponding to FA edges. The coefficients of the contributions of individual fuel elements are investigated by the relation (6). For the energy 185 keV of 235 U isotope in the Fig. 7, the inverse values of the contributions are given in a logarithmic scale. The type of the contribution pattern is typical for lowenergy isotopes, in particular, at the observation angles corresponding to the FA edges, there is an acute "ridge", and a pronounced "multi-ridge structure" is observed on the midpoints of the faces.  Fig. 9. Their appearance is also quite typical for high-energy gamma quanta: in particular, the "ridge structure" is much less pronounced than in Fig. 7. To research the practical possibility of determining the initial burn-up of fresh nuclear fuel, the ratio of measured intensities of 238 U with energy of 1001 keV and 235 U with an energy of 185 keV is modelled, taking into account the registration efficiency [12] depending on the angle of the detector location (Fig. 10). It can be seen from the figure that the ratio of the measured intensities repeats to a considerable extent the picture in fig. 6: there are irregularities and small "sharp increases" observed in the form of peaks, and besides the peaks being concentrated at the viewing angles corresponding to the midpoints of edges. However, the absolute value of the peaks is small, which allows us to assume that the detectors are arranged exactly opposite to the edges. Quantitative analysis of the contribution coefficients for both analytical lines of 235 U, 238 U isotopes made it possible to determine the following: with an allowable unevenness of contribution coefficient of 40%, placing the detector in opposite to the midpoint of the fuel assembly allows coverage of the entire external row of fuel rods and 4 central fuel elements of the second row for 185 keV 235 U energy, and almost completely two fuel rods close to the detector for energy 1001 keV 238 U. Considering the homogeneity of fresh nuclear fuel inside the fuel assembly (except for the first row), this is quite sufficient to estimate the initial enrichment. The number of detectors should be determined experimentally, depending on the required accuracy of measurements.
The above experimental spectra and simulation results show that the developed technical facilities to control the burn-out of spent nuclear fuel can be successfully used to determine the initial enrichment of fresh nuclear fuel. The developed equipment makes it possible to determine the enrichment of nuclear fuel when it is transferred by a reloading machine to the reactor core without additional time. The introduction of such a system provides a qualitatively new level of monitoring of the state of nuclear fuel at all stages of its life cycle at nuclear power plants. In particular, there will be an automatic analysis of the initial enrichment of nuclear fuel, a burn-out analysis for each campaign based on the results of measurements during the overload, a control of the fuel cladding tightness during the overload, and an analysis of burn-out during shipment to dry storage. In addition, close integration with the control system of the overload machine ensures automatic maintenance of a database on the history of nuclear fuel, i.e. fulfillment of the IAEA requirements for nuclear safeguards. Continuous accounting of the burn-out of each fuel assembly is also the main source of information to refine the methods of nuclearphysical calculations of the reactor core.