Abstract
Each service company that offers pulsed neutron logs includesmeasurements based on the gamma rays from inelasticscattering. The gamma rays that originate from the high-energy neutrons that undergo inelastic scattering necessarilyoriginate near the source. The number of these gamma raysthat reach detectors relatively far from the source reflects thegamma ray scattering by and thus the density of the materialbetween the scattered neutrons and the detector. Because thereare many factors other than density that influence the numberof gamma rays detected, service companies offermeasurements based on inelastic scattering as empirical tracesthat are used in combination with measurements based onneutron scattering. Such combinations are useful primarily torecognize gas and distinguish gas from low porosity. Onecompany offers a trace that represents density from the gammarays due to inelastic scattering, but the process of deriving thedensity involves the very empirical step of subtractingdifferent portions of the gamma rays attributed to factors otherthan inelastic scattering until a satisfactory approximation todensity is achieved.
This paper presents a method for predicting the gammarays from thermal neutrons to be subtracted from the totalgamma rays measured during the neutron pulse and acomparison of the density-like trace so derived with open-holedensity logs recorded in Cabinda, Angola, Africa.
The method presented for predicting gamma rays due tocapture is based on representing the decay of gamma rays afterthe neutron burst by two exponential decays. The fraction ofgamma rays due to thermal neutrons is different at differenttimes during the pulse, but the overall effect is predicted fromthe amplitudes and decay times that fit the data used torepresent sigma. While this removes variations in the fractionof capture gamma rays subtracted to create a density-like trace,the relationship between the amplitude of the trace and densityis still determined empirically.
The comparison with open-hole logs makes use of the factthat many of the logs run were run to distinguish brine from oilthrough casing. This enables comparison of the densitiesderived from cased-hole logs to those recorded from open-holelogs under conditions where the open-hole logs still representthe reservoir. Some of the log data studied were from intervalswhere gas saturations changed between the open- and cased-hole logs, and these afford a chance to compare changes indensity with changes in neutron porosities. Examplecomparison are presented and limitations discussed.
