In-situ measurement of Cs distribution in the soil. Sabina Markelj

Transcription

1 In-situ measurement of Cs distribution in the soil Sabina Markelj Supervisor: Matej Lipoglavšek April 12, 2004 Abstract In-situ gamma spectrometry is a new method for measuring radionuclides in soil. It has appeared to be a very useful method for measuring radioisotopes if the shape of their depth distribution in soil is known. Caesium is the main representative of man-made world contamination of soil. For determination of Cs in soil we have to assume its depth distribution. A new function for the distribution of Cs has been proposed together with a new way of measuring the function parameters. They could not be determined with only a single measurement due to the fact, that Cs-137 emits gamma rays with only one energy, whereas other radionuclide decays usually have more gamma ray energies. Additional equations were obtained from measurements with a lead plate placed in front of the detector at various distances. A new model for migration of caesium into deeper layers of soil has been developed and its solution describes the depth distribution very well.

2 Kazalo 1 Introduction 3 2 Radiation Radioisotopes Measurementofdose Humanexposuretoradiation Biologicalconsequencesofradiation In-situ measurements 5 4 Depth distribution of Cs in soil 6 5 Analysis of in-situ gamma spectra 7 6 Results 11 7 Conclusion

3 1 Introduction People are exposed to radiation every day, so it is in our interest to evaluate the radiation, to which we are exposed. Sources of it are cosmic radiation, natural and man-made radioisotopes. In this seminar we will be mostly interested in radioisotopes which are in the soil. They contribute to the external radiation and are also the cause of internal radiation via the foodchain or inhalation of radionuclides. Natural radioactivity is composed of long-lived terrestrial radioisotopes that are found in all environments ( 40 K, 238 U series, 232 Th series ), but their contents and ratio of individual isotopes can be modified by both natural processes and certain human activities. Man-made radioisotopes, mostly 137 Cs and 134 Cs, are the consequence of environmental contamination due to some uncontrollable or uncontrolled event. Worldwide contamination was caused by atmospheric nuclear weapon tests in the late 50s and early 60s, and large scale contamination followed the Chernobyl accident in 1986 [1]. For determination of dose from γ-emitters in soil we have to know their activity and depth distribution. In the case of natural radioisotopes their distribution is usually rather uniform unless some human activity such as deposition of layers of different activity took place. In case of contamination with man-made isotopes from nuclear tests or nuclear accidents, when the most important contamination path is via the atmosphere, the depth distribution is very inhomogeneous. The most complete information about radionuclide contents in soil is given by measuring different soil samples from layers at different depths and by subsequent laboratory analysis of individual samples. This method is very time-consuming and representativeness of the sampling site may be questionable. Another way of determination of radionuclide depth distribution is by measuring the γ-ray spectrum on a certain location. This way of measuring is called in-situ γ spectrometry and it has proven to be a very powerful tool for determination of specific activity of soil [2]. For isotopes that emit γ-rays of different energies some parameters of depth distribution may be determined by analysis of a single measurement based on the different absorption of γ- rays in the soil. That is due to known energy dependence of attenuation coefficient in soil. Rays with different energy are absorbed differently. Combining information from different energies enables us to asses the depth distribution. One in-situ measurement takes about one hour, while laboratory analysis takes for each sample one day. A drawback of the described in-situ method is that, that it is not applicable for radionuclides, emitting γ-rays of single energy. Such radionuclide is 137 Cs (energy of γ ray is 662 kev) which is the most common man-made radioisotope in the environment. In this seminar a new method will be introduced by which the activity distribution, of radioisotopes that emit γ-rays of single energy, can be determined [3] and [2]. We will mostly discuss about the distribution of Cs isotopes. Natural radioisotopes are distributed uniformly, while in the case of Cs the distribution is not uniform. During the contamination phase most of activity of man-made radioisotopes is limited to the surface of the undisturbed land and later it gradually migrates into deeper layers. For doing any kind of calculations with measurements distribution of fallout has to be assumed. Usually an exponential distribution was assumed [6], which exhibits a maximum at the soil surface. In recent years the maximum activity has shifted to a finite depth so a different distribution has to be assumed. The purpose of this seminar is to describe the new method, which is faster and better in determination of the distribution of radionuclides in the soil when the distribution has the maximum at certain depth. And I will also briefly show a simple model, developed by Matej Lipoglavšek, of how to explain the penetration of Cs in soil [3] and [7]. 3

4 2 Radiation 2.1 Radioisotopes Every radioisotope is characterized by some typical basic parameters such as decay path, half-life and activity. Most often nuclei decay by α or β decays. In the first, the nucleus emits an α particle and in the other it emits an electron or positron and neutrino or antinevtrino. Such decays are often accompanied by emission of γ rays. Every radioactive isotope has a different lifetime. The basic information is commonly the half-life of nuclide, which is the time in which one half of the initial nuclei will decay. They can decay very quickly (10 12 s)orslowly(10 9 years). Activity of a radioactive substance is determined by the number of decays in a time interval. It is measured in becquerel [1 Bq = decay per second]. 2.2 Measurement of dose Effectiveness of radiation is measured by dose. Energy which was absorbed in a unit of mass is called the absorbed dose (unit is Gray): 1Gy =1J/kg But the biological effects are not only dependent on the absorbed dose but also on the type of radiation. That is why biological effects are declared by the equivalent dose [H]: H T = R w R D Where D is the absorbed dose and w R is the factor which takes into consideration that damage is not the same for all types of radiation (w R (γ,e) = 1, w R ( thermal neutrons)= 2, w R (fast neutrons)=20 and w R (protons) = 10, Q(α) = 20). Equivalent dose is measured in Silvert [Sv]. At this point we can also stress that different tissues are also differently sensitive to radiation. That is considered by tissue weighting factors(w T ) and the called dose is effective equivalent dose [H E ]: H E = w T H T 2.3 Human exposure to radiation Radionuclides are in our environment. There are two ways of being exposed to radiation: external radiation and internal radiation. Most of external radiation is from radionuclides in human environment, so we are exposed to it everywhere. Internal irradiation is the consequence of inhalation of air or eating contaminated food. Cosmic radiation comes from the Universe. The top layer of atmosphere is constantly bombarded by high energy particles such as protons, helium and heavily particles. They come from the galactic and extragalactic sources but the majority of them comes from the sun. In the atmosphere they hit molecules of air and produce secondary cosmic radiation (electrons, muons, neutrons,... ). There are also some other radioisotopes which are produced with absorption of cosmic radiation in the atmosphere, the most important are 3 Hand 14 C. Cosmic radiation mostly contributes to external radiation. Natural long-lived radioisotopes are in the environment since the formation of the solar system. The most important are 40 K(t 1/2 =1, years) and two decay series 238 U(t 1/2 = 4, years) and 232 Th (t 1/2 =1, years) which with numerous α, β, γ decays come to 4

5 stable nuclei. They are found in all soils and rocks. In series of U and Th there are important products of 222 Rn (radon) and 220 Rn (toron). Both are noble gases which partially come in our atmosphere. On average natural radioisotopes contribute to external radiation 410 µsv/year and to internal radiation 1600 µsv/year. Most of internal radiation is the consequence of inhalation of radon. The whole yearly dose is 2400 µsv/year, where 2/3 is from internal and 1/3 from external radiation. Less than 1/2 of the contribution to external radiation is the cosmical radiation and the other half natural radioisotopes. Man made radioisotopes are mainly the consequence of nuclear explosions of bombs and nuclear accidents. The most important are short-lived 131 I, 90 Sr and long-lived 137 Cs. In explosion of 3 Hand 14 C, which are also products of cosmic radiation. Due to its long half life, 137 Cs is the main representative of man-made world contamination of soil. 2.4 Biological consequences of radiation Since the very beginnings of working with radioisotopes, it has been known that radiation can cause damage to living organisms. It can cause serious health problems and serious consequences in any living creature, but the damage depends on the dose that the body or certain part of it receives. The effects of radiation can be divided in two types. First the deterministic effects which are the direct consequence of radiation, such as nausea, falling out of hair or even death if the dose is bigger than 4-5 Gy. On the contrary, the stochastic effects show up a few years after the irradiation. They are dose dependent, are coincidental, and can cause diseases such as cancer, inheritable effects and so on. Since the year 1977 all countries, including Slovenia, have agreed with the basic principles of standard radiological protection, proposed by ICRP report from year ICPR recommends a system of dose limitation, which stands on three principles: cause for exposure to radiation has to be justified, optimization of exposure and individual limitation to dose. For more information see [1] 3 In-situ measurements In-situ γ-ray spectroscopy measurements are performed with portable germanium detector which provides a practical way to characterize dispersed radionuclides in the soil. It is placed 1 m above the ground. When the measurement is started multichannel analyzer counts the rate of events. One measurement takes about one hour respectively, for the spectrum to give satisfying statistical results in present conditions in Slovenia. Measurements are performed on undisturbed soil when the migration of Cs is investigated. 5

6 Figure 1: Portable gamma detector [5] As mentioned earlier, there is another way, the traditional one, of measuring the depth distribution of radionuclides. With that method every cm of soil is sampled separately, for which you need a special shovel to take the samples out. Samples are then specially prepared and dried in the laboratory, for the measurement to be preformed. One measurement of a sample takes about one day to get satisfying results which are analysed afterwards. If we compare these two methods of measuring, it is not difficult to see that in-situ gamma spectroscopy is the easier one. Any discrete sample taken for laboratory analysis will only identify what was at that specific, very small, sample site. This means that for cases where the contamination is not uniform, some hot spot places could be missed. In situ gamma spectroscopy on the other hand, effectively detects all the radioactivity over as much as 10 4 m 2 of area, and for high energy gammas, even detects radioactivity buried below the surface of soil. With in-situ gamma spectroscopy there is much higher probability that nothing will be missed. 4 Depth distribution of Cs in soil The first time when caesium appeared in soil was after tests with nuclear bombs. The second major deposition came with fallout from Chernobyl accident. Since then the migration of it in undisturbed soil has been carefully monitored in Europe. Cs was deposited in a relatively nonsoluble form. So the transport of the isotope in soil is therefore due to washing of the fallout to deeper layers and attachment to specific soil components. It has been confirmed that this is rather slow process. Cesium from the nuclear bombs was deposited in a rather long period. That is why the maximum of the distribution was constantly on the top of the ground, despite the washing into deeper layers. So the distribution of Cs was satisfyingly described with an exponential function. On the other hand the Cs of the Chernobyl accident was deposited rather quickly so the maximum is more discrete and in some years it should be shifted into deeper layers. The latest measurements have really showed this difference in distribution of radioactive cesium in soil. It differs from the old one in the position of the maximum of the distribution. It has shifted to lower layers of soil. There have been a few suggestions on how to describe this new distribution of Cs such as with the difference of two exponential functions and so on, but none of them gave any satisfying results. Lipoglašek and colleagues suggested that an appropriate simple model to describe this obviously complex phenomenon is a diffusion process in a moving medium with given boundary conditions. The distribution of specific activity (a) of Cs in a moving medium is described by the following differential equation: D 2 a = a t + v a 6

7 Where D denotes the diffusion constant and v the average transport velocity of cesium. The term v a describes absorption of radionuclides in the soil and it is assumed to be responsible for the observed shift of maximum activity to deeper layers. Equation can be taken as one dimensional, because it is supposed that the soil is homogeneous and semiinfinite. Equation can be solved by Green s function method. In the case of Chernobyl accident, radionuclides were deposited in a time interval of a few days, so the appropriate boundary condition is that the flow of particles is zero on the soil surface. is: (av D a z ) z=0 =0 So the correct Green s function which solves the equation and obeys the boundary condition G(z,t) = e v vt (z 2D πdt where Φ is probability integral 2 ) [e z2 v 4Dt 2 πt D e v vt (z+ 2D Φ(x) = 2 x e t2 2 dt 2π 0 2 ) 1 (1 Φ( (z + vt)))] 2Dt This function is a solution of our problem in case if radionuclides fall on the surface in a short period of time, as happened at Chernobyl accident. The initial distribution of activity is therefore a δ-function peaked at the surface. 5 Analysis of in-situ gamma spectra The data in which we are interested is the specific activity and should be determined from the measured in-situ gamma spectra. It can be seen, from the spectra, how many photons with a certain energy the detector has detected. The number of them is a measurement for the isotope activity. The count rate n in a full energy peak in the spectrum is calculated from the following equation. n = tb η(r, z)a(r, z)e µ(z,e)x 4π x 2 dv where b denotes the branching ratio, factor e µ(z,e)x considers the absorption from the source to the position of detector, t the time of measurement and η denotes the detector efficiency. Now we substitute x with proper coordinates which are in our case cylindrical (r and z) and consider the expression for µ which denotes linear attenuation coefficient: µ(z) =(µ a h + µ s z)/(h + z) where µ a and µ s are linear attenuation coefficients in air and soil respectively. Attenuation coefficient in soil is calculated from known attenuation coefficients of soil components. Each of them is multiplied with proper factor, which tells how much of that component is in the soil. 7

8 Figure 2: Geometry [3] n = tb µah+µsz exp( h+z r 2 +(h + z) 2 ) a(z)η(r, z) r 2 +(h + z) 2 rzdz (1) h is the distance between the detector and earth surface. The detector efficiency can be parametrized as: h 2 η(r, z) =η 0 r 2 [A +(1 A)eB exp( B r 2 +(h + z) 2 )] +(h + z) 2 h + z where η 0 denotes the absolute efficiency for a point source at a distance h on the axis of the detector and parameters A(E) and B(E) describe the detector anisotropy. It is obtained from measurements with known γ emitted at different positions from detector. Now all we need to do is to give the expression for activity in the equation and integrate it. As one can see Green s function for the problem is rather complex that is why it was approximated by Gaussian depth distribution: a(z) =a 0 exp( (z z 0) 2 2σ 2 ) In this case three parameters are required: the depth of the maximum, z 0, the maximum activity, a 0, and the standard deviation σ 0, describing the spread of the distribution. The exact solution and the Gaussian distribution are compared in Figure 3. Figure 3: Comparison between the Green function and Gaussian function (vt/ 4Dt= 0,35 ) [2] 8

9 Both functions are drawn with parameters z 0 = vt and σ0 2 =2Dt and normalized to the same deposit, for two different ratios vt/ 4Dt, which correspond to the measurements. Deposit is the whole activity of radioisotopes per unit area, which for instance in the case of a nuclear accident fall on a certain surface. So the integral of specific activity over depth has to be made while in case of laboratory measurements the specific activities of samples are multiplied by thicknesses of layers to get the deposit. It is measured in Bq/m 2. The integration of equation over the radius gives the result: n = bta 0 η 0 2πh 2 exp( (z z 0) 0 2σ 2 ) [ae 1 (p + µ s z)+(1 A)e B E 1 (p + µ s z)]dz where p and p denote µ a h and p+b respectively and E is exponential integral. E n (x) =x n 1 e t x t n dt The integration over the axial coordinate has to be performed numerically. As mentioned earlier there have been some other suggestions of how to describe the distribution of caesium activity. One of them was the difference of two exponential functions. a(z) =C 1 e αz C 2 e βz In this case we get the following solution for the number of counts in the full energy peak: n = tbη 0 2πh 2 [ C 1 α [A(E 1(p) e αp µs E 1 (p(1 + α µ s )))+ (1 A)e B (E 1 (p )+e αp µs E 1 (p + α µ s ))] [ C 2 β [A(E 1 (p) e βp µs E 1 (p(1 + β µ s )))+ (1 A)e B (E 1 (p )+e βp µs E 1 (p + β µ s ))]] We have four unknown parameters in the equation, C 1,C 2,α and β, so we need at least four data from the measurements for their determination. On the other hand with the Gaussian function we have to determine only three parameters, σ 0,a 0 and z 0, and as it can be seen from the figures 4 and 5, it also gives better accordance with the measurements. Figure 4: Depth distribution of (a) 137 Cs and 134 Cs at some location. Solid and dashed lines are computed from in situ measurements using Gaussian and double exponential depth distribution. [2] 9

10 Figure 5: Depth distribution of 137 Cs at location. [2] For isotopes which emit γ-rays of different energies each peak in the spectrum yields an equation from which the unknown distribution parameters are calculated by the method of least squares. For isotopes which emit γ-rays at less than three energies, additional equations are necessary in order to determine the depth distribution. In that case in-situ measurements are performed with a lead-plate absorber placed coaxially to the detector at various distances. The distance between the plate and the detector defines the average depth from which the detected γ-rays originate. So the additional equations are obtained from the measurements with the absorber positioned at different distances. A circular lead plate of thickness 3 cm and with radius of 14,5 cm was used in these measurements. Figure 6: Position of lead plate. [2] In the calculation it is assumed that the plate absorbs the γ-rays emitted at radii smaller than (z + h) tan (ϑ). The altered lower integration limit effectively changes the quantities p, p and µ s in the integrand to p/ cos ϑ, p / cos ϑ and µ s / cos ϑ in equation 1. In that way we get more equations if the measurements are performed at different distances from the detector and so it enables us to get the parameters of the distribution even in the case of nuclei emitting γ-rays of a single energy. Due to a finite sensitivity volume of the detector the plate only partly prevents the registration of γ-rays emitted at interval of radii around the lower integration limit. To minimize this interval the plate should not be placed too close to the detector. On the other hand the plate should not be placed too far from the detector in order to conceal a substantial solid angle. It 10

11 was estimated that the optimum position of the plate is between 10 and 20 cm from the detector. It was found that beside the measurement without the plate only two measurements with the plate are sufficient to determine the depth distribution. 6 Results In-situ measurements and soil sampling were made by Ecological Laboratory at three different locations in Slovenia. The detector was placed 1 m above a flat meadow. In addition to measurements without the absorber, two measurements with plate at different distances were made. The distance between detector and plate varied between 10 and 20 cm. On the figures rom 7 to 11 a comparison between laboratory and in-situ measurements is shown. Statistical errors are labelled on each of the measurements. They were calculated from the spectrum: the statistical error of the peak area ( N), and the error of background. Gaussian model for distribution of specific activity was chosen for the modelling. Full line on the figures represents specific activity calculated from in-situ measurements. Errors for the isotope 134 Cs are bigger and some measurements from certain depths are absent due to the low activity of 134 Cs. Activities at the missing distances are too small, so the detection system can not detect them. Typical values for D in Slovenia are few cm 2 /year, for v are few mm/year and for z 0 are few cm. Figure 7: Depth distribution of specific activity of 137 Cs at location 1 (village Drnovo( Krško polje)). Laboratory measurements ar labelled with dots and solid line is computed from in situ measurements. [3] 11

12 Figure 8: Depth distribution of specific activity of 137 Cs at location 2 (village Mrtvice( Krško polje)). [3] Figure 9: Depth distribution of specific activity of 137 Cs at location 3 (village Brege( Krško polje)). [3] Figure 10: Depth distribution of specific activity of 134 Cs at location 1. [3] 12

13 Figure 11: Depth distribution of specific activity of 134 Cs at location 2. [3] Figure 12: Depth distribution of specific activity of 137 Cs at location 3. The solid line is computed from in-situ measurements using double exponential distribution. [3] Figure 13: Depth distribution of specific activity of 134 Cs at location 1. The solid line is computed from in-situ measurements using double exponential distribution. [3] 13

14 One can notice that measurements made deeper in the soil give higher values for specific activity than those calculated with the model of Gaussian distribution from the in-situ measurements. The reason for this discrepancy can be perhaps found in the caesium which was left from the nuclear weapon tests and was transported deeper in the soil than the Chernobyl caesium. The model does not take into account caesium from the bombs. Another possible source of disagreement can be also found in the inhomogeneity of soil density. Figures 12 and 13 also present the specific activity calculated with the model of two exponential functions. It can be seen that there is better agreement between in-situ and laboratory measurements when a Gaussian depth distribution is assumed instead of double exponential distribution. Calculation with the double exponential model was often impossible since it requires one parameter more than the Gaussian distribution. In table 1 you can see parameters, which were determined from measurements. Location Isotope a 0 [kbq/m 3 ] σ[cm] z 0 [cm] v [mm/y] D[cm 2 /y] Cs 125 8,4 4,4 6,4 5, Cs 8,3 4,7 3,7 5,7 1, Cs 130 4,7 3,7 5,7 1, Cs 9,2 3,3 2,2 3,4 0, Cs 139 5,6 3,9 6,0 2,4 7 Conclusion In this seminar an improved method for determination of depth distribution of radionuclides in soil has been presented. Gamma-ray spectra were measured in-situ by a germanium detector with an absorbing lead plate placed between detector and ground at different distances from the detector. With these measurements additional equations were obtained. These equations determine the depth distribution also for the isotopes that emit rays with single energy such as 137 Cs. The activity distribution, for the radioisotopes that are a direct consequence of Chernobyl nuclear accident, 137 Cs and 134 Cs, was described by Gaussian model which has a physical background. Gaussian distribution is especially appropriate for evaluation of in-situ measurements since it has fewer unknown parameters than the double exponential distribution that has been proposed earlier. Reasonable agreement with laboratory measurements of soil samples was achieved. 8 Literatura [1] R. Martinčič, B. Pucelj, Reference book of Ionizing radiation Ljublana (1991) [2] M. Korun et al., Nucl. Inst. Meth. Phy. Res. B 93 (1994) 485 [3] M. Lipoglavšek, Diploma 1993 [4] A. Likar et al., Enviro. Radioactivity 57 (2001) 191 [5] Canberra In-situ Spectroscopy system, [6] M. Korun et al., Nucl. Inst. Meth. Phy. Res. A 300(1991) 611 [7] A. Likaretal., J. Phys. D: Appl. Phys. 33 (2000)