METHOD FOR NON-DESTRUCTIVE MEASUREMENT OF URANIUM ENRICHE

Description

TECHNICAL FIELD

[0001] The technical field of the invention relates to a technique for non-destructive measurement of the enrichment of uranium in its isotope 235 PREVIOUS ART

[0002] Currently, the reference method for non-destructively assessing the enrichment of uranium in its isotope 235 is gamma spectrometry.

[0003] Gamma spectrometry is one of the most commonly used passive non-destructive nuclear measurement techniques to obtain qualitative and quantitative information on gamma-emitting radionuclides.

[0004] Uranium-235 and some uranium-238 descendants can be measured by gamma spectrometry. However, gamma spectrometry is quite complex to implement and requires certain assumptions about the depth and gradient of contamination, particularly in porous objects such as concrete walls. These assumptions are not necessarily valid.

[0005] Furthermore, when the enrichment in 235< U is high, and / or for small quantities of uranium, the acquisition time can be long, due to the low intensity of the gamma emissions of the 238< U descendant radioelements. Another technique for measuring uranium is known from ASHRAFI SALEH ET AL: "Measurement of natural radioactivity of Iranian granite samples using beta-gamma coincidence spectrometer and maximum likelihood method", ENVIRONMENTAL EARTH SCIENCES, vol. 78, no. 15. This article discloses a method for estimating the uranium activity of an object, using a detector, the detector being configured to form pulses when exposed to β particles, the detector being connected to a spectrometric measurement circuit, configured to establish a spectrum, the spectrum corresponding to a histogram of the energy of the pulses formed during a measurement period.

[0006] The inventor proposes an alternative method for estimating uranium enrichment. The method can be deployed on civil engineering structures, equipment, or collected samples. It is a method based on a device that is inexpensive and easy to implement. STATEMENT OF THE INVENTION

[0007] A first subject of the invention is, according to claim 1, a method for estimating an enrichment in 235< U of an object, using a detector, the detector being configured to form pulses when it is exposed to β particles, the detector being connected to a spectrometric measurement circuit, configured to establish a spectrum, the spectrum corresponding to a histogram of the energy of the pulses formed during a measurement period, the method comprising the following steps: a) arranging the detector facing the object and acquiring a measurement spectrum, the measurement spectrum being representative of an energy distribution of β particles emitted by descendants of 238< U and by descendants of 235< U; b) from the measurement spectrum, possibly forming a corrected spectrum; c) from the measurement spectrum or the corrected spectrum, determining a first spectral value in a first energy band extending from a first minimum energy, the first minimum energy being greater than or equal to 300 keV; d) from the measurement spectrum or the corrected spectrum, determining a second spectral value in a second energy band extending below a second maximum energy, the second maximum energy being less than or equal to 300 keV; e) calculating a ratio between the first spectral value and the second spectral value;f) application of a calibration function to the ratio determined during step e), so as to estimate the enrichment in 231< U.;

[0008] The process may include the following sub-steps: b1) interposition of a screen between the detector and the object and acquisition of a background spectrum, the screen being configured to absorb the β particles emitted by the object; b2) correction of the measurement spectrum, using the background spectrum, to obtain a corrected spectrum

[0009] The screen used in sub-step b1) may be an aluminum screen with a thickness greater than or equal to 3 mm or 4 mm.

[0010] According to one possibility, in step c), the first energy band extends up to a first maximum energy, the maximum energy being greater than or equal to 1000 keV or 1500 keV or 2000 keV.

[0011] According to one possibility, in step a), the distance between the detector and the object is: greater than 1 mm or 5 mm; and / or less than 20 mm or 10 mm.

[0012] When the object is a naturally radioactive object, the process may include: (i) taking into account a natural activity of the object; (ii) modeling a β spectrum of the natural activity of the object (SpRN) detected by the detector; step b) includes a correction of the measured spectrum using the β spectrum of the natural activity of the object, to obtain the corrected spectrum.

[0013] According to one possibility, step (i) results from a gamma spectrometry measurement carried out on the object.

[0014] The detector may comprise an organic scintillator material. The detector may comprise a semiconductor material.

[0015] A second object of the invention is, according to claim 10, a device for estimating an enrichment in 235< U of an object, the device comprising: A detector (10) comprising a detector material (11) configured to form pulses under the effect of exposure to β particles; a spectrometry circuit (13), connected to the detector, and configured to form a histogram of the amplitude of the pulses detected by the detector during an acquisition period; a processing unit (20), connected to the spectrometry circuit, and configured to implement steps b) to f) of a method according to the first subject of the invention

[0016] The detector may comprise an organic scintillator material, the thickness of the organic scintillator material preferably being between 3 mm and 10 mm. The detector may comprise a semiconductor material.

[0017] The detector may be of an envelope configured to absorb α particles of energy greater than or equal to 4.5 MeV or 5 MeV.

[0018] The device may comprise a movable screen, configured to move from a closed configuration, in which the screen is interposed between the detector material and the object, to an open configuration in which the screen leaves a space free between the detector material and the object. The invention will be better understood upon reading the description of the exemplary embodiments presented in the remainder of the description, in conjunction with the figures listed below. FIGURES

[0019] There figure 1 schematizes a measuring device. The figure 2 shows an evolution of the energy deposited in the detector material of the device by β particles as a function of the thickness of the latter. The figure 3 shows an evolution of the energy deposited in the detector material by β particles as a function of the thickness of an aluminum screen interposed between the detector material and the object. Figure 4Ashows a spectrum of the energy of β particles detected by the detector material as a function of different distances between the detector material and the object, with uranium contamination assumed to be surface-based. Figure 4B shows a spectrum of the energy of β particles detected by the detector material as a function of different distances between the detector material and the object, the uranium contamination being assumed to be volumetric, according to a thickness of 2 mm. The Figure 5 shows spectra of the energy of β particles detected by the detector material in the presence of an aluminum screen, without the aluminum screen. Also shown is a corrected spectrum resulting from the subtraction of the modeled spectra with and without the screen. The figure 6shows a spectrum of the energy of β particles detected by the detector material due to the natural radioactivity of the object, as well as spectra of the energy of β particles detected by the detector material with and without subtraction of the spectrum of the natural activity. figure 7 shows simulations of spectra of the energy of β particles detected by the detector material in different configurations. The Figure 8A shows spectral simulations of the energy of β particles detected by the detector material resulting from different radioelements, for a low enrichment of 235< U. The Figure 8B shows simulations of spectra of the energy of β particles detected by the detector material resulting from different radioelements, for a strong enrichment of 235< U. The figure 9 shows a spectrum of the energy of α particles detected by the detector material. The figure 10shows a spectrum of the energy of β particles detected by the detector material. The figure 11 shows the evolution of a transfer function, allowing a spectral value to be converted into an activity level, and this for different contamination depths. Figures 12A, 12B And 12C show experimental maps carried out with a device according to the invention. The Figure 13A shows the evolution of a ratio between a first spectral value and a second spectral value as a function of the enrichment in 235< U, taking into account surface contamination. The Figure 13B shows the evolution of a ratio between a first spectral value and a second spectral value as a function of the enrichment in 235< U, taking into account volume contamination. figure 14 shows the main steps of a method for determining U activity by implementing the invention. PRESENTATION OF SPECIAL EMBODIMENTS

[0020] There figure 1 represents a measuring device allowing measurement of the uranium activity of an object 2. The device comprises a scintillator detector 10, comprising a material scintillator 11 preferably organic, preferably based on Polyvinyltoluene (PVT). Under the effect of interactions between ionizing radiation and the scintillator material, light pulses are formed. These light pulses are converted into electrical pulses by one or more photodetectors 12. The electrical pulses are then processed by a spectrometry circuit. The spectrometry circuit 13 is configured to form an amplitude histogram of the pulses detected by the organic scintillator during an acquisition period.

[0021] By implementing an energy calibration function, resulting from an energy calibration, it is usual to perform a correspondence between the amplitude of the pulses and energy values. When an ionizing particle deposits all its energy in the scintillator, the amplitude of the pulse it generates corresponds to the energy of the particle before the interaction in the detector material.

[0022] The use of an organic scintillator detector is suitable for performing charged particle spectrometry, such as β-type. An organic scintillator is also sensitive to ionizing photons of the X or γ type. However, the materials forming an organic scintillator have a low atomic number, which makes them unsuitable for the formation of photoelectric interactions. Thus, an organic scintillator is considered unsuitable for X or γ spectrometry applications.

[0023] The scintillator detector is covered with an optically sealed envelope 14, of low thickness, for example an aluminized PET (Polyethylene Terephthalate) film, 18 µm thick to ensure sealing against ambient light. The low thickness makes it possible to minimize the attenuation of β radiation. In the rest of the text, the term β particle designates a β -< particle. The thickness of the envelope can reach 30 µm to stop alpha particles of energy 4.5 MeV or 5 MeV.

[0024] The thickness e of the scintillator detector is 4 mm. Details of the scintillator thickness will be given in connection with the figure 2. An organic scintillator material has the advantage of being relatively insensitive to gamma radiation, due to its low atomic number. In addition, this type of scintillator limits the backscattering phenomenon of β particles. Another advantage is a certain stability of the scintillator material's response to thermal variations. The response, in terms of light intensity produced, when exposed to the same radiation, is stable from 0° to 50°C, which is suitable for field use conditions.

[0025] The measuring device comprises a processing unit 20, configured to implement spectrum processing steps described below. The processing unit 20 is programmed to execute instructions coded in a memory, connected to the processing unit by wired or wireless connection. The processing unit 20 may in particular comprise a microprocessor.

[0026] Alternatively, the detector may comprise a semiconductor material suitable for β spectrometry. For example, it may be a silicon-type semiconductor, for example planar silicon.

[0027] The advantage of an organic scintillator material is the ability to be manufactured in different dimensions and shapes. When the object is a sample being taken, the shape can be adapted to the shape of the sample.

[0028] The detector 10 comprises a movable screen 15, acting as a shutter, configured to be arranged: in a closed position, between the detector 10 and the object to be measured 2; or in an open position, freeing the space between the detector 10 and the object 2, as shown in the figure 1 .

[0029] In this example, the device is primarily dedicated to the radiological monitoring of walls. In order to be able to be deployed along large surfaces, the surface area of ​​the detector material, in a plane perpendicular to its thickness, is 50 cm x 50 cm. This allows a large surface area to be addressed during each measurement, while being sufficiently compact and light to be easily handled.

[0030] In this example, the screen 15 is movable in translation in a plane parallel to the detector material 11. The screen is for example formed of aluminum, the thickness preferably being equal to 4 mm. Details relating to the thickness of the screen 15 are given in connection with the figure 3 The screen 15 can be connected to the detector material by a slide system, allowing translation of the screen relative to the detector material.

[0031] Object 2 is an object to be controlled, likely to have a mass or surface activity of Uranium, the isotopy being known or unknown. In the examples described below, the object is a concrete wall.

[0032] The invention is based on detection, by the detector 10, of β particles emitted by descendant isotopes of Uranium isotopes. This includes 234m< Pa, which is in radioactive equilibrium with 238< U, and emits β particles up to a maximum energy Eβmax equal to 2269 keV. It is also possible to exploit the detection of β particles emitted by 231< Th (descendant of 235< U), whose Eβmax = 308 keV or 234< Th (descendant of 238< U), with Eβmax = 198 keV.

[0033] There figure 2represents a quantity of energy (y-axis - unit MeV.Bq -1< .g)), deposited by β particles emitted by a homogeneous distribution of 234m< Pa in a concrete depth of 2 mm. The x-axis corresponds to the thickness (unit mm) of the scintillator material 10. The depth of 2 mm corresponds to the maximum path of β particles of energy 2269 keV (maximum emission energy of 234m< Pa) in the concrete. The figure 2 was obtained by simulation, considering a distance d=7.5 mm between the detector material and the concrete. It is observed that using a thickness e of 4 mm, the scintillator material can absorb 97% of the β emission energy of 234m< Pa. The simulation was carried out by the MCNP6 (Monte Carlo N Particles) calculation code, which is a reference code in the field of modeling interactions between ionizing radiation (β, γ, neutrons) and matter.

[0034] There figure 3represents a quantity of energy (y-axis - unit MeV.Bq -1< .g)), deposited by β particles emitted by a homogeneous distribution of 234m< Pa in a concrete depth of 2 mm. The x-axis corresponds to the thickness (unit mm) of the aluminum screen 15. The figure 3 was obtained by simulation, considering a distance d=7.5 mm between the detector material and the concrete. It is observed that beyond a thickness ε of 4 mm of aluminum, the gain in terms of absorption of β particles is negligible. A thickness of 4 mm is therefore sufficient to effectively seal the detector material against β particles emitted by 234m< Pa.

[0035] THE Figures 4A And 4Baddress the issue of the distance between the detector and the object to be controlled. Since the controlled object is potentially contaminating, it is preferable to avoid the detector 10 being in direct contact with the object. A slight recoil is useful, so as to allow the screen 15 to move from the open position to the closed position. Figure 4A shows simulations of spectra of the energy deposited by β particles in the scintillator material, as described in connection with the figure 1 . We considered a homogeneous surface distribution of 234m< Pa on a wall and we took into account several distances d from the detector in relation to the wall: d = 2 mm; d = 7 mm and d = 10 mm. On the Figure 4A, the abscissa axis corresponds to the energy (MeV) and the ordinate axis corresponds to the counting rate (s -1< .Bq 1< .cm 2< ). It is observed that between 2 mm and 10 mm, the spectrum is not significantly modified. The difference between the spectra acquired considering distances of 7 mm and 10 mm is negligible.

[0036] There Figure 4B shows simulations of the same type as those described in connection with the Figure 4A . On the Figure 4B , a homogeneous volume distribution of 234m< Pa was considered over a thickness of 2 mm of concrete. The same distances d from the detector to the wall were taken into account as on the Figure 4A The difference between the spectra acquired considering distances of 2.5 mm and 10 mm is negligible.

[0037] The inventor considers that the optimal distance between the object 2 and the detector 10 is between 5 mm and 10 mm. Beyond 10 mm, there is a risk of a loss of efficiency, i.e. a reduction in the number of β particles detected. Below 5 mm, handling the screen becomes complex, the latter being too close to the object, with risks of contact between the screen and the object. Subsequently, the distance between the detector and the object is equal to 7.5 mm.

[0038] There Figure 5 shows the effect of the detector shuttering by the screen 15 on the spectrum measured by the detector. The Figure 5 represents spectra measured experimentally in front of a uranium-contaminated wall. Y axis: count rate (s -1< ) ​​- X axis: energy (MeV). A spectrum without Spβγ screen was measured respectively (screen in the open position, as shown in the figure 1), a spectrum with Sp bdf screen (screen in the closed position). The spectra were measured using the same acquisition time of 120 seconds. The screen was a 4 mm thick aluminum plate. The spectrum with screen, noted Sp bdf, is considered to be representative of the γ background noise. The γ background noise is due to the γ-emitting radioelements present in the object (e.g., natural radioelements) or in the detector environment. Subtracting the spectrum with Sp bdf screen from the spectrum without screen gives a spectrum corrected for the γ background noise. The corrected Spβ spectrum, or raw β spectrum, is considered to be solely representative of the β particles emitted by the object.

[0039] Natural radioelements potentially present in the object include, for example, 40< K, as well as descendants of 232< Th and 238< U. Such radioelements are present, for example, in concrete objects. γ background noise can also result from artificial radioactivity present in the detector's environment.

[0040] When the object contains natural radioactivity, the raw β spectrum Spβ includes a natural component due to β particles emitted by natural radioelements. This natural component can, in a first approach, be neglected because it is generally quite weak. Table 1 shows the different β-emitting radionuclides of natural and artificial origin likely to be encountered in a concrete wall potentially contaminated with uranium. Table 1 Beta emitters Maximum energy (keV) Intensity (%) Origin 40K 1311.07 89.25 Natural 235U decay chain 231st 307.85 29 Natural & Artificial 227Ac 44.8 53 Natural 223Fr 1149.2 1 Natural 211Pb 1367 91.28 Natural 207Tl 1418 99.73 Natural Decay chain of 2380 234th 198 77.8 Nature! & Artificial 234mPa 2269 97.599 Natural & Artificial 214Pb 1019 9.2 Natural 2148i 3270 19.67 Natural 210Pb 63.5 19.8 Natural 210Bi 1162 100 Natural 232Th decay chain 228Ra 39.5 12 Natural 228Ac 2066 6 Natural 212Pb 569.9 13.3 Natural 2128i 2252.1 55.31 Natural 208Tl 1801.3 49.2 Natural

[0041] There figure 6 shows the Spβ-corrected β spectrum, as described in connection with the Figure 5 (raw β spectrum), as well as a β spectrum, also corrected for the γ background noise, measured in front of an uncontaminated concrete wall (natural background noise spectrum or RN background noise spectrum, noted SpRN). The uncontaminated concrete wall was located in the same installation as the contaminated wall, on which the corrected β spectrum was formed. Thus, the RN background noise β spectrum, noted SpRN, can be considered as representative of a natural β contribution in the corrected β spectrum. figure 6shows a net β spectrum, denoted Spβ', obtained by subtracting the natural background noise β spectrum SpRN from the raw β spectrum. The net β spectrum Spβ' is thus representative of the artificial β activity in the measured wall.

[0042] There figure 7 represents β spectra simulated with the MCNP6 calculation code, in a detector material as previously described (50 cm x 50 cm x 4 mm), taking into account mass activities of 232< Th, 238< U and 40< K measured by gamma spectrometry on a wall considered to be uncontaminated. The measurement by gamma spectrometry made it possible to estimate activity levels of 15 Bq / Kg for 232< Th, 22 Bq / Kg for 238< U, and 485 Bq / Kg for 40< K. These mass activities make it possible to simulate the natural component in the β spectrum measured by the detector. On the figure 7 , we have represented: the component due to 232< Th in the β spectrum; the component due to 238< U in the β spectrum; the component due to 40< K in the β spectrum; the component due to 235< U in the β spectrum, assuming natural enrichment of uranium: the mass of 235< U corresponds to 0.72% of the mass of uranium;

[0043] On the figure 7 , we also represented a simulation of β spectrum resulting from 234m< Pa taking into account a homogeneous surface activity of 0.4 Bq.cm -2< and an enrichment of 1% by mass in 235< U (UB spectrum), and taking into account an enrichment of 8.16% by mass in 235< U (UH spectrum).

[0044] In these simulations, the distance between the wall and the detector was considered equal to 7.5 mm, and the density of the concrete was considered equal to 2.3 g.cm -3< .

[0045] We observe that: the natural component of the β spectrum is dominated by 40< K, up to the energy 1311 keV, which corresponds to the maximum energy of the β particles emitted by 40< K; the measurement of the β spectrum resulting from 234m< Pa is largely superior to the natural contribution at low enrichment as well as at high enrichment, beyond 1 MeV. It is possible to model the natural component in a β spectrum measured by the detector, taking into account mass activities, established a priori or measured, of the main natural β-emitting radioelements.

[0046] Furthermore, simulations have shown that the γ component of the natural activity is not significantly absorbed by the aluminum screen. Also, subtracting the background noise spectrum, measured in the presence of the aluminum screen, allows us to eliminate the γ component of the natural activity.

[0047] THE Figures 8A and 8Bshow simulations of β spectra measured by a detector as described in the figure 1 , taking into account a homogeneous surface activity of uranium, on a wall element of 50 cm on each side, taking into account respectively an enrichment in 235< U of 1% and 82.55% by mass. By surface activity, we mean an activity distributed over a depth of 10 µm. In each of these figures, the contributions of 234m< Pa, 234< Th and 231< Th are represented. 234m< Pa and 234< Th are descendants of 238< U, while 231< Th is a descendant of 235< U. The distance between the detector and the wall was considered equal to 7.5 mm.

[0048] In each of these figures, the y-axis corresponds to a counting rate (s -1< .Bq -1< .cm 2< ) and the x-axis corresponds to an energy level (MeV).

[0049] THE Figures 8A and 8Bshow that whatever the level of enrichment, the spectrum is dominated by the contribution of 234m< Pa, the latter being weaker at high enrichment ( Figure 8B ) than at low enrichment ( Figure 8A ). We observe that by considering a first energy range ΔE1 delimited by a first minimum energy E1min, greater than or equal to 300 keV, and by summing the spectrum over this first energy range ΔE1, the spectral value obtained depends only on the activity of 234m< Pa. By spectral value, we mean a value measured from the spectrum. It can in particular be a counting rate or a count in the energy range. The spectral value obtained over the first energy range ΔE1 is a first spectral value noted N1.

[0050] Preferably, the first energy range ΔE1 extends up to a first maximum energy E1max greater than or equal to 2269 keV, the latter corresponding to the maximum emission energy of a β particle per 234m< Pa. For example, the first energy range ΔE1 is [300 keV; 2500 keV]. Such an energy range makes it possible to collect the entire useful spectral content of the β activity of 234m< Pa. The first maximum energy E1max may be equal to 2300 keV, or be less than 2000 keV or 1500 keV or 1000 keV. It is noted that the first maximum energy E1max is preferably less than or equal to 2500 keV, or even 2300 keV. Thus, although the first optimal energy range is [300 keV; 2500 keV], a narrower energy range within this interval may be suitable. An energy range extending from 300 keV, or from a first minimum value E1min greater than or equal to 300 keV, and less than or equal to 1000 keV, or 1500 keV may be suitable.It is noted that beyond 1000 keV, the value of the spectral channels decreases. It is therefore preferable that the first minimum energy E1min is between 300 keV and 1000 keV. It is possible not to take into account a first maximum energy: E1max = +∞: the range considered is then [E1min, +∞], with 300 keV ≤ E1min≤1000 keV.

[0051] The first spectral value N1, in the first energy range ΔE1 = [E1min, E1max] previously defined, is then intended to be processed by a transfer function, so as to be able to estimate an activity level in 238< U. The establishment of the transfer function is described subsequently, in connection with the figure 11 . Taking into account the enrichment η then makes it possible to convert the activity in 238< U into activity in U.

[0052] THE Figures 8A and 8Bshow another interesting aspect of the invention: we can define a second energy range ΔE2 , between 0 keV and 300 keV, in which the β spectrum is influenced by 231< Th, this radioelement descending from 235< U. Thus, we can define a second energy range ΔE2=[E2min, E2max], between a second minimum energy E2min≥0 and a second maximum energy E2max≤300 keV, in which the spectral value N2, called the second spectral value, depends on 235< U, in particular at high enrichment.

[0053] A ratio N1 / N2 of the first and second spectral values ​​depends on the enrichment η. Thus, the experimental measurement of this ratio can be used to estimate the enrichment η of uranium.

[0054] When the enrichment η of uranium in 235< U is unknown, the first spectral value N1 allows to estimate 238< U, using the transfer function FT.Indeed, the spectral content, in the first energy range, does not depend on the enrichment. The ratio of the first and second spectral values ​​allows the enrichment η to be estimated. The activity in 238< U can then be used to estimate the activity in U. The enrichment in 235< U can also be determined experimentally by gamma spectrometry.

[0055] Uranium contains α-emitting isotopes, primarily 234< U or 238< U. In order to estimate the contribution of α particles in the measured β spectrum, a 244< Cm source, with an activity of 2800 Bq, was placed 3 mm from the detector material. 244< Cm emits α particles with energies greater than 5700 keV, which is much higher than the energies of α particles emitted by U isotopes, for example 4775 keV for 234< U. The figure 9shows the spectrum measured over an acquisition period of 900s (y-axis: number of detected pulses - x-axis: energy (keV)). It can be seen that the contribution of α particles is limited to low energies, below 270 keV. Thus, the spectral value, in the first energy range, can be considered as unaffected by α particles emitted by uranium isotopes.

[0056] The α particles detected by the detector may be likely to influence the spectral value N2 in the second energy band ΔE2, or even the spectral value N1 in the first spectral band ΔE1. ​​This is particularly the case when the detector comprises a semiconductor material, for example Si. In order to avoid an influence of the α particles, the envelope 14 is advantageously sized to stop the α particles of energy 4.5 MeV or 5 MeV. When the envelope is made of aluminized PET (Polyethylene Terephthalate), a thickness of 30 µm makes it possible to stop the α particles, while absorbing, in a manner considered negligible, the β particles.

[0057] On the figure 9 , the few shots observed at an energy greater than 270 keV correspond to y photons after subtraction of the background noise, resulting from statistical fluctuations.

[0058] There figure 10shows a β spectrum measured experimentally on a concrete wall contaminated with 1% enriched uranium. On this spectrum, the first energy range was materialized, between 300 keV and 2500 keV. The detector was an EJ200 detector (Eljen Technology), with a sensitive surface area of ​​2430 cm 2< (493 mm on each side). The energy calibration of the spectrum, i.e. the correspondence between the value of the pulse amplitudes and the energy, was carried out using a 207< Bi source, which emits electrons, at discrete energy values, by internal conversion.

[0059] In order to convert the spectral value, in the first energy range, into an activity value of 238< U, it is necessary to take into account a transfer function. The transfer function can be obtained by modeling β spectra corresponding to known activities in 238< U, and therefore in 234m< Pa. Different transfer functions can be determined, corresponding respectively to different hypotheses of contamination distribution in the analyzed object. Uranium contamination can be considered as surface (for example on metallic objects), or having diffused in the object, for example in the case of liquid contamination of porous objects, such as a floor or a concrete wall.

[0060] There figure 11shows the result of numerical modeling of a detector as previously described, placed at a distance of 7.5 mm from a concrete wall, of density 2.3, whose uranium 238< U activity was distributed according to different depths. A unit activity of 238< U was considered (1 Bq / g or 1 Bq / cm 2< in the hypothesis of surface contamination). The detector model was validated in the laboratory by comparing β spectra respectively measured and modeled by exposing the detector to a standard source of 207< Bi.

[0061] In the case of surface contamination, activity distributed over the first 10 µm of the wall was modeled. The transfer function is 1050s -1< .Bq -1< .cm 2< ± 15%.

[0062] In the case of volumetric contamination, the maximum path, in the concrete, of the β particles emitted by 234m< Pa was taken into account, which could be estimated at 2 mm. Given this low thickness, the contamination gradient was considered to be homogeneous.

[0063] On the figure 11 , the y-axis corresponds to the value of the transfer function FT (unit s -1< .Bq -1< .cm 2< ) and the abscissa axis corresponds to the depth of contamination considered. On

[0064] there figure 11 , the transfer function corresponds to the first spectral value N1, in an energy range of [300 keV - 2500 keV], for a unit activity of 1Bq.g -1< .

[0065] If N1 corresponds to the counting rate in the first energy band ΔE1, the activity A [238 U ], in Bq.g -1< or in Bq.cm -2< , is obtained according to the expression: A 238 U = N 1 FT

[0066] The previously described detector was implemented on a gaseous diffusion uranium enrichment facility. The detection limit was such that: LD = k 1 − β 2 + 2 × SD FT × T S

[0067] With k 1 -β = 1.96: type risk level β for a 95% confidence interval; TS : acquisition time, in seconds; FT = transfer function, expressed in Bq.g -1< or in Bq.cm -2< SD is the detection threshold, such that SD = k 1 − α × 1 + 1 n × r bdf × T s k 1- α = 1.96: type risk level α for a 95% confidence interval; n: ratio of the respective acquisition times of the background noise and the measurement r bdf : count rate measured in the first energy band ΔE1, in a background noise measurement, on a wall considered to be uncontaminated. Unit counts per second.

[0068] Depending on the unit of the transfer function, the detection limit is expressed in Bq.g -1< or in Bq.cm -2<

[0069] Tables 2 and 3 represent detection limit values ​​expressed respectively in surface activity and in mass activity. The values ​​are calculated from expressions (2) and (3), considering a measurement duration equal to the duration of the background noise measurement. A contribution of natural activity in the spectrum, described in connection with the figure 7 The activities of the natural isotopes taken into account were 15 Bq / Kg for 232< Th, 22 Bq / Kg for 238< U, and 485 Bq / Kg for 40< K. Each measurement includes a measurement with the aluminum screen and a measurement without the aluminum screen, of the same durations. Tableau 2 LD en Bq.cm -2< Radionucléide ROI r BDF (s -1< ) FT (s -1< .Bq -1< .cm 2< ) 60 180 Ts (s) 300 600 1800 238< U ( 234m< Pa) [300 - 2500] keV 48 ± 8 % 1050 ± 15 % 4,8.10 -3< 3,4.10 -3< 2,2.10 -3< 1,5.10 -3< 8,7.10 -4< Tableau 3 LD en Bq.g -1< Radionucléide ROI r BDF (s -1< ) FT (s -1< .Bq -1< .g) 60 180 Ts (s) 300 600 1800 238< U ( 234m< Pa) [300 - 2500] keV 48 ± 8 % 149 ± 15 % 3,4.10 -2< 2,4.10 -2< 1,5.10 -2< 1,1.10 -2< 6,1.10 -3<

[0070] It is observed that the method allows low detection limits to be obtained, within reasonable acquisition times. The detection limits presented in Tables 2 and 3 concern 238< U via the detection of the β spectrum of 234m< Pa. As the enrichment increases, the amount of 238< U decreases relative to the total amount of uranium. Therefore, if we wish to maintain a low uranium detection limit, we must increase the acquisition time.

[0071] Maps were made on the walls of the gaseous diffusion uranium enrichment facility. figure 12A shows measured counting rates (cps = counts per second) on surface elements of dimensions 50 cm x 50 cm. The figure 12B shows the corresponding 238< U activity levels (Bq / m 2< ). Since the 235< U enrichment is known, U activity levels were estimated, and these are shown on the figure 12C (Bq / m 2< ). When the enrichment is not known, it can be determined by usual methods, for example by gamma spectrometry.

[0072] As previously described, in connection with the figures 8A et 8B , the enrichment η in 235< U can be evaluated from a ratio between the first spectral value N1 and the second spectral value N2. The second energy band ΔE2 is between 0 keV and 250 keV. For example, a second energy band ΔE2 can be taken into account, extending from E2min=150 keV to E2max=300 keV. The second minimum limit E2min of 150 keV was determined in order to limit the influence of gamma radiation: between 150 keV and 300 keV, it was found that the contribution of gamma radiation in the β spectrum is stable, and can be easily subtracted by taking into account the measurement with the aluminum screen. Considering a second minimum energy E2min lower than 150 keV, the contribution of gamma radiation in the β spectrum can fluctuate more, due to the higher sensitivity of the plastic scintillator material to low energy photons.

[0073] There figure 13A shows the evolution of a ratio between the first and second spectral values. The first energy band is [300 keV - 2500 keV] and the second energy band is [150 keV - 300 keV]. The contamination is assumed to be surface-based, i.e. considered to be distributed over a thickness of 10 µm.

[0074] There figure 13B shows the evolution of a ratio between the spectral values ​​(count rate) in the first energy band [300 keV - 2500 keV] and in the second energy band [150 keV - 300 keV] in the case of mass contamination, distributed over a thickness of 2 mm.

[0075] On the figures 13A et 13B , the y-axis corresponds to the N1 / N2 ratio and the x-axis corresponds to the enrichment in 235< U, expressed in %. The figures 13A et 13B were established on the basis of models with the MCNP6 calculation code.

[0076] A correlation is observed between enrichment and the N1 / N2 ratio. Thus, by applying an enrichment calibration function, as described in connection with the figures 13A ou 13B , we can estimate the enrichment from the N1 / N2 ratio. figures 13A et 13B show the establishment of enrichment calibration functions, respectively for surface activity and volume activity.

[0077] It is specified that the use of β spectrometry to estimate an enrichment rate η may constitute an independent aspect of the invention.

[0078] In practice, the use of a background noise spectrum, obtained according to the closed configuration, with the screen 15, is not necessary. Indeed, if one wishes to carry out a rough control, measurements carried out only according to the open configuration, i.e. without a screen, may prove sufficient to carry out a first-level control. The coupling between the open configuration and the closed configuration makes it possible to subtract the background noise, essentially due to the y photons, which makes it possible to obtain more precise measurements: more precise quantification of the 238< U activity or more exact enrichment value.

[0079] Furthermore, experience has shown that in the first energy band ΔE1 , extending from 300 keV to 2500 keV, the contribution of the gamma background noise due to natural radioactivity is stable, in the same installation. However, carrying out measurements with and without a screen increases the total measurement time. Given the low dispersion of the contribution, it is possible not to carry out a systematic measurement in the closed configuration.

[0080] It is possible to subtract, from a spectrum measured in open configuration, a stored spectrum, carried out in the closed configuration.

[0081] There figure 14 summarizes the main steps of a method according to the invention.

[0082] Step 100: Arrange a detector facing the object to be characterized

[0083] Step 110: acquisition of a spectrum, in the open configuration: this is the measured spectrum Sp.

[0084] Step 120: acquisition of a background noise spectrum Sp bdf , in the closed configuration, or taking into account a spectrum acquired in the closed configuration.

[0085] Step 130: Subtraction of the spectrum resulting from step 120 from the spectrum resulting from step 110. The spectrum resulting from this step is a raw β spectrum Spβ.

[0086] Step 140: Taking into account the natural β activity of the object. This step includes the following sub-steps: Sub-step 141: taking into account a natural activity level of the object: the natural activity level may have been estimated by a measurement on a comparable object, considered to be uncontaminated. Sub-step 142: estimation of a natural SpRN β spectrum (or RN β spectrum) resulting from step 141.

[0087] Alternatively, step 140 includes a sub-step 143 of measuring a raw β spectrum carried out on an object considered to be representative of the measured object, and not contaminated. This involves implementing steps 110 and 120 on the object considered to be representative, which makes it possible to obtain the natural β spectrum.

[0088] Step 140 is preferred, but optional.

[0089] Step 150: correction of the raw β spectrum, resulting from step 130, taking into account the natural β spectrum resulting from step 140. This step is optional. Step 150 makes it possible to obtain a net β spectrum (or corrected β spectrum) noted Spβ'

[0090] Step 160: determination of a first spectral value N1 in the first energy band ΔE1, and of a second spectral value N2, in the second energy band ΔE2 from the net β spectrum Spβ' resulting from step 150 or from the raw β spectrum resulting from step 130.

[0091] Step 170: Application of the enrichment calibration function to a ratio between the first spectral value and the second spectral value, so as to estimate the enrichment rate η in mass of 235< U.

[0092] Steps 130 to 170 may be implemented by the processing unit 20.

[0093] The invention may be implemented in fuel enrichment or manufacturing installations.

Claims

1. Method for estimating the 235U enrichment of an object using a detector, the detector being configured to form pulses when exposed to β particles, the detector being connected to a spectrometric measurement circuit configured to establish a spectrum, the spectrum corresponding to a histogram of the energy of the pulses formed during a measurement period, the method comprising the following steps: - a) positioning the detector opposite the object and acquiring a measurement spectrum (Sp), the measurement spectrum being representative of an energy distribution of β particles emitted by descendants of 238U and by descendants of 235U; - b) from the measurement spectrum, possibly forming a corrected spectrum (Spβ; Spβ'); - c) from the measurement spectrum or the corrected spectrum, determining a first spectral value (N1) in a first energy band (ΔE1) extending from a first minimum energy (E1min), the first minimum energy being greater than or equal to 300 keV; - d) from the measurement spectrum or the corrected spectrum, determining a second spectral value in a second energy band (Δ E2) extending below a second maximum energy (E2max), the second maximum energy being less than or equal to 300 keV; - e) calculating a ratio between the first spectral value and the second spectral value; - f) applying a calibration function to the ratio determined in step e) in order to estimate the 235U enrichment.

2. Method according to claim 1, wherein step b) comprises the sub-steps: - b1) interposing a screen (15) between the detector and the object and acquiring a background spectrum, the screen being configured to absorb the β particles emitted by the object; - b2) correcting the measurement spectrum using the background spectrum to obtain a corrected spectrum.

3. Method according to claim 2, wherein the screen (15) used in sub-step b1) is an aluminium screen with a thickness greater than or equal to 3 mm or 4 mm.

4. Method according to any one of the preceding claims, wherein during step c), the first energy band extends up to a first maximum energy (E1max), the maximum energy being greater than or equal to 1000 keV or 1500 keV or 2000 keV.

5. Method according to any of the preceding claims, wherein in step a), the distance between the detector and the object is: - greater than 1 mm or 5 mm; - and / or less than 20 mm or 10 mm.

6. Method according to any of the preceding claims, wherein the object is a naturally radioactive object, the method comprising: - (i) taking into account natural activity of the object; - (ii) modelling a β spectrum of the natural activity of the object (SpRN) detected by the detector; - Step (b) involves correcting the measured spectrum using the β spectrum of the natural activity of the object to obtain the corrected spectrum.

7. Method according to claim 6, wherein step (i) results from a gamma spectrometry measurement performed on the object.

8. Method according to any of the preceding claims, wherein the detector comprises an organic scintillator material.

9. Method according to any of claims 1 to 8, wherein the detector comprises a semiconductor material.

10. Device for estimating the 235U enrichment of an object, the device comprising: - a detector (10) comprising a detector material (11) configured to form pulses upon exposure to β particles; - a spectrometry circuit (13), connected to the detector, and configured to form a histogram of the amplitude of the pulses detected by the detector during an acquisition period; - a processing unit (20), connected to the spectrometry circuit, and configured to implement steps b) to f) of a method according to any of the preceding claims.

11. Device according to claim 10, wherein the detector comprises an organic scintillator material, the thickness of the organic scintillator material being between 3 mm and 10 mm.

12. Device according to claim 10, wherein the detector comprises a semiconductor material.

13. Device according to any of claims 10 to 12, wherein the detector is covered by an envelope (14) configured to absorb α particles with energy greater than or equal to 4.5 MeV or 5 MeV.

14. Device according to any of claims 10 to 13, wherein the device comprises a movable screen (15), configured to move from a closed configuration, in which the screen is interposed between the detector material (11) and the object (2), to an open configuration in which the screen leaves a space between the detector material and the object.