**3. Simulation of gamma-ray spectra using Geant 3.21 Monte Carlo code for sources distributed in 220l volume**

Application area of the radiation transport modeling through Monte Carlo method is ex‐ tremely large, from the nuclear reactor design to parameters calculation of complex detection systems, from the simulation and the interpretation of various experiments to the calculation of the dose coefficients. Nowadays, this area is expanding, both by tackling new problems in modeling and by the inclusion of some details, previously neglected, of the respective phenomena [17]. What characterizes the Monte Carlo method is the remarkable fidelity with which it can describe physical phenomena, without approximations [18, 19]. Theoretically, the accuracy of results is limited only by the accuracy of the nuclear data (the effective cross section of interaction) used, and no other method is competitive with the Monte Carlo method in this regard. Accordingly, the Monte Carlo method is often used as a reference method; the simplified calculation procedures, faster, based on some approximations are tested in relation to this method, and even nuclear data can be validated by interpretation of the Monte Carlo complex experiments [18, 19].

In the Monte Carlo method, a problem of radiation transport is solved by simulating the evolution of a large number of radiations and the analysis of their fate. The evolution (the history) of each radiation starts with its emission continues with the undisturbed propagation through the environment between the successive interactions and by changing its parameters (possibly even its disappearance) in the points where the interactions takes place. In the case of photon transport at keV–MeV energy range, complex effects are important, such as Compton photoelectric and electron–positron pairs generation effects. Following the Compton effect, the incident photon energy is transferred to the recoil electron and to the scattered photon. In the photoelectric effect, the photon is absorbed, an electron is emitted, and the atom remains in an excited state. In the pair production effect, the photon energy is consumed to produce a pair of electron–positron particles. Thus, except for the Compton effect, the history of each photon ends at the interaction point. However, in terms of energy dissipation in substance, the primary interaction of the photon is only the starting point. The resulted electrons give their energy causing ionization and excitation, emitting new photons (secondary radiation), and the atoms will relax from excited states, emitting new photons and possibly Auger electrons; the positron will annihilate producing annihilation photons that interact further. Obviously, if more details of the secondary radiation will be included, the computing time will increase, and the computation program becomes more complex with more branched. At every interaction, the evolution of the resulting products and then of the successive generations of secondary radiations should be evaluated. Which details are relevant and which are not depends on the problem to be solved for achieving an optimal compromise between modeling finesse, the required accuracy and the programming effort and the necessary computing time. For incident photons with energy up to a few MeV, characteristic X-rays, the bremsstrahlung radiation (emitted by electrons resulting from interactions) have energies much lower than the primary photon energy. Consequently, in many problems, X-rays can be considered locally absorbed.

The development of the Monte Carlo methods and the improvement of computational technologies have led to the development of several Monte Carlo simulation programs for simulating radiation transport. Simulation codes used are GEANT 3 [20], GEANT 4 [21], MCNP [22], GESPECOR [6], FLUKA [23], ETRAN [24], EGS [25–27], PENELOPE [28], etc.

In this section, the application of Monte Carlo simulation to the study and examination of the response function characterization of two gamma-ray spectrometry systems used for meas‐ uring large sources was reported. For this purpose, GEANT 3.21 code was applied for the spectra simulation expected to be obtained for 50–2000 keV energy range for volume sources measured with both systems. Although the prevalent application of Monte Carlo simulation for efficiency calibration of HPGe detectors for the measurements of small volume samples up to several dm3 , extensive realistic computations by Monte Carlo methods have not been carried out until now for the measurement of big volume samples like 220l waste drums.

#### **3.1. Experimental configuration**

source-detector distance, the detector-absorber distance, ETNA software was applied for the computation of the efficiency for the other measurement geometries for the HPGe detector. The experimental efficiency curves were compared with the prediction of the ETNA software. Excepting the case of the smallest distance from the source to the detector, the discrepancies between ETNA and the experimental results were generally below 3%. In the case of the measurement at 2 cm distance from the detector, the discrepancies were higher being sensitive to the detector geometrical data. This is because the detector specifications established by the manufacturer of the detector were used without any optimization. Furthermore, the uncer‐ tainty of the distance between the crystal and the end cap (the manufacturer value) has a contribution in the uncertainty of the transfer factor, because of the change in the solid angle

The HPGe detector efficiency transfer method has also been used for the efficiencies evaluation for the specific cylindrical sources. For this purpose, the matrix was considered water equiv‐

The default attenuation coefficients foreseen by ETNA code were used for the matrices

In the case of the soil matrix, containing 137Cs, the ratio between the ETNA software values and the experimental values of the efficiency was 1.038 for *h* = 0 cm and 0.966 for *h* = 2 cm. The higher discrepancies in the case of 134Cs results (gel matrix) in comparison with the results for 137Cs (soil matrix) can be attributed to the uncertainty of coincidence summing effects and of

**3. Simulation of gamma-ray spectra using Geant 3.21 Monte Carlo code for**

Application area of the radiation transport modeling through Monte Carlo method is ex‐ tremely large, from the nuclear reactor design to parameters calculation of complex detection systems, from the simulation and the interpretation of various experiments to the calculation of the dose coefficients. Nowadays, this area is expanding, both by tackling new problems in modeling and by the inclusion of some details, previously neglected, of the respective phenomena [17]. What characterizes the Monte Carlo method is the remarkable fidelity with which it can describe physical phenomena, without approximations [18, 19]. Theoretically, the accuracy of results is limited only by the accuracy of the nuclear data (the effective cross section of interaction) used, and no other method is competitive with the Monte Carlo method in this regard. Accordingly, the Monte Carlo method is often used as a reference method; the simplified calculation procedures, faster, based on some approximations are tested in relation to this method, and even nuclear data can be validated by interpretation of the Monte Carlo

In the Monte Carlo method, a problem of radiation transport is solved by simulating the evolution of a large number of radiations and the analysis of their fate. The evolution (the

and soil composition for the matrix with *ρ* = 1.4 g/cm3

.

[10].

alent for the matrix with *ρ* = 1.0 g/cm3

**sources distributed in 220l volume**

complex experiments [18, 19].

involved in the study.

126 Nuclear Material Performance

the matrix effects.

The first system used is an ISOCART from Ortec (Geom1) and has a p-type detector with a relative efficiency of 25%. The second system used is a WS1100 Segmented Gamma Scanner from Canberra (Geom2) and has a p-type detector with a relative efficiency of 44.4%. The characteristics and dimensions of the detectors are presented in **Figure 7**.

**Figure 7.** ISOCART (left) and Segmented Gamma Scanner WS1100 (right) gamma-ray spectrometry systems.

The volume source considered in simulation was a 2201 radioactive waste drum typically used for conditioning of radioactive waste in Romania. Several studies were reported using this kind of sample [29–32]. The source matrix was considered to be concrete with standard composition, and the axis of the detector was perpendicular on the axis of the cylinder. The distance from the center of the coordinate system associated to the detector to the center of the cylinder was 50 cm for both geometries.
