A Comparison of Energy-Window and Spectral-Fitting Methods for the Estimation of Carbonate Content in Rocks Using Neutron-Induced Gamma Rays

When a nuclear explosive is tested underground, rock close to the explosion is heated to very high temperatures. If any of that rock contains carbonates, these are dissociated and the carbon may combine with oxygen to form carbon dioxide. When the rock cools, the carbon dioxide remains gaseous, possible causing subsurface pressures greater than atmospheric, which could increase the hazard of a delayed leak. To reduce that possibility, formations that might contain significant amounts of carbonates are analyzed for those materials. We have compared measurements of carbonate content from two commercial neutron-induced gamma-ray spectroscopy logging system in emplacement boreholes (2 to 4 m in diameter) at the Nevada Test Site (NTS). Both systems use high-energy neutrons to excite carbon and oxygen nuclei inelastically, and a spectrometer to detect the resulting gamma rays. One system uses broad energy windows around the main carbon and oxygen gamma-ray peaks and relates the carbonate content to the ratio of the two window count rates. This ratio requires correction for the gap between the sonde and the borehole wall, which is measured mechanically, and for the formation water content, which is obtained from a neutron log. The other system determines carbonate content by fitting the entire energy spectrum to a set of standard spectra reflecting the system response to pure elements in the formation. Water corrections are obtained directly from the spectroscopic measurements; no gap correction is needed. We tested both systems in a calibration facility to determine their responses to different values of carbonate, gap, and water content. They were then run in two boreholes at the NTS. Cuttings samples from both holes were analyzed for carbonate content for comparison with log results.