Estimation of thermal neutron absorption cross-section from K, U and Th concentrations for Miocene rocks from the Carpathian Piedmont in Poland using artificial neural networks.

The radiometric K, U and Th concentrations and neutron absorption cross-section sigma a of rock samples obtained from coring are analysed. The cores are from wellbores located in the Sucha-Jordanów region (Carpathian Mountains) and from gas producing Miocene formations in the Carpathian foothills. Correlation coefficients between the neutron absorption cross-section (sigma a) and K, U and Th concentrations are presented. Neural network representation of the function sigma a = f(K, U, Th) obtained for a region can later be used for sigma a estimation from spectrometric probe results in uncored wells.