Fluid identification based on geophysical well logs in ultralow-porosity tight sandstone reservoirs

Abstract

Deep tight sandstone reservoirs typically exhibit ultralow porosity and permeability owing to their small pore size, narrow pore throats, and poor pore connectivity. In heterogeneous tight formations, well-log interpretation is uncertain, and it is difficult to identify fluid types using the classic picket-plot (resistivity versus porosity crossplot) method. To better identify and characterize gas-bearing zones in heterogeneous tight sandstone reservoirs, effective sensitivity factors for fluids need to be constructed. From the conventional logging principle, the acoustic neutron porosity difference, density-neutron porosity difference, and triple-porosity ratio are all sensitive parameters for gas-bearing layers. In addition, the density and nuclear magnetic resonance (NMR) porosity difference and [Formula: see text] geometric mean of the movable fluid are sensitive to gas saturation. Based on these parameters, a series of fluid typing crossplots was constructed, and their effectiveness was compared. In contrast, the interpretation results from NMR logging are better, and the dual-porosity difference and triple-porosity ratio methods from conventional logs also are effective when NMR logging is not available. In practice, these methods are used to provide their results, and then they are combined using a voting strategy according to the performance of each crossplot. In two case studies of the Kuqa Depression of the Tarim Foreland Basin, China, the new fluid typing method proved effective, and it provides an alternative method for fluid identification using nonelectrical logs (when electrical logs are not available or applicable).