A New NMR Method of Fluid Characterization in Reservoir Rocks: Experimental Confirmation and Simulation Results

Abstract

Abstract A new magnetic resonance fluid (MRF) characterization method for fluids in porous rocks has been developed. The method uses suites of spin-echo measurements that can be acquired in the laboratory or by a nuclear magnetic resonance (NMR) logging tool. In general, the data suites consist of spinecho measurements with different echo spacings, polarization times, applied magnetic field gradients, and numbers of echoes. These measurements are sensitive to the viscosities and molecular diffusion coefficients of the fluids and therefore provide the information needed for fluid characterization. The MRF method is based on the inversion of a general multifluid relaxational model that describes the decay of the transverse magnetization in porous rocks containing reservoir fluids. In its most general form, the relaxational model consists of separate contributions to the measured spin-echo signals from all of the fluids that can be present in reservoir rocks; i.e., brine, oil, gas, and oil-base mud filtrate (OBMF), including mixtures of gas dissolved in oil or OBMF. A key ingredient in the multifluid relaxational model is a new phenomenological microscopic constituent viscosity model (CVM) for hydrocarbon mixtures that links diffusion-free relaxation and molecular diffusion in crude oils. The CVM significantly improves the robustness of the inversion so that accurate fluid characterization is possible even when the brine and crude oil T1 and T2 distributions overlap one another. We present experimental results on live and dead hydrocarbon mixtures and crude oils that confirm the validity of the CVM. The results of a Monte Carlo simulation for a model carbonate formation containing brine, crude oil, gas, and OBMF demonstrate the robustness and accuracy of the inversion. Monte Carlo simulations conducted for different types of rocks containing different amounts and types of fluids demonstrate that the inversion provides quantitative estimates of total porosity, fluid saturations, fluid volumes, bulk volume irreducible water, crude oil viscosity, hydrocarbon-corrected permeability, crude oil T1 and T2 relaxation time distributions, crude oil diffusion coefficient distributions, brine T2 distributions, and apparent brine T1/T2 ratios. The MRF method was also tested in the laboratory on partially saturated Berea 100 and Indiana limestone rocks containing brine and oil. The water saturations estimated from the NMR data by inversion of the multifluid relaxational model are shown to agree well with the water saturations that were independently estimated from differential weight measurements. The oil viscosity estimates are also shown to be in good agreement with the known viscosity of the oil.