In the context of the controlled designing areas for accelerators of less than 10 MV, specific attenuation curves and patient dispersion coefficients with uncertainty values of less than 1% were produced by using Geant4. A tool that allowed us to evaluate different geometric configurations, having as study parameters the spectra of the beam used, the compositions and densities of the materials used and their dimensions. The results found are divided into two lines. On the one hand, from the analysis of the beam attenuation in concrete, attenuation curves were obtained for accelerators with energies of 4 and 6 MV, among which there were spectra of accelerators such as Varian Clinac 21EX, Siemens, among others. The trend shown by the data found was verified and compared with results currently used that refer to fixed TVL values, or that differentiate a first TVL and an equilibrium TVL, finding that, as expected, the values currently used are conservative and therefore can be improved. An additional result clearly shows the dependence between the attenuation values and the density of the attenuating material, which corresponds to a linear relationship. This result is considered of great value as it provides us with a valuable generalizing tool when producing the TVL value for different densities. Finally, three different concrete compositions were evaluated, including Portland concrete, where the relationship between the attenuation capacity and the effective atomic number of a material became evident. A second line of analysis was performed through the scattering by the patient coefficients. Where it was possible to validate the simulations made in this work, through the reproduction of the scattering coefficients currently used (the ones published in the IAEA report 47 and in the NCRP report 151) with the difference that in this work we managed to report the uncertainties associated with each value. Secondly, a simpler calculation geometry was proposed for these coefficients, which we consider adequately addresses the study problem, thus implying much smaller uncertainties (less than 1%) without the results found differing considerably with those used today. Finally, it was possible to report dispersion coefficients in the patient for a higher number of angles than are currently available (this is due to the fact that their uncertainty was reduced) and from them it was determined that they adjust adequately to an exponential type curve that provides us with a powerful generalization tool in case of requiring values for unreported angles in practice.