Calculating Variation of Fluid Density in Near-Wellbore Environment and Its Application to Density/Porosity Calculations

Abstract

Abstract Knowledge of fluid density within the pore volume investigated by gamma-gamma density logging tools is necessary to accurately calculate porosity. As porosity increases the effect of fluid density on density porosity estimates increases. The algorithm presented here calculates "true fluid density" as a function of fluid saturations, hydrocarbon composition, temperature and pressure. The algorithm can handle mixtures of oil, gas, and water at different salinities, temperatures, and pressures. More accurate fluid densities are used to improve accuracy of porosities derived from gamma-gamma density measurements. Because the gamma-gamma density tool responds to some function of "electron density" and not the "true density", the concept of "apparent density" was developed during the 1950's for quantitative interpretation of gamma-gamma density logs. "True density" is related to "electron density" by the ratio (Z/A) of the atomic number (Z) to the atomic weight (A). The ratio (Z/A) varies with the fluid composition, hence the accuracy of calculated densities is influenced by fluid composition. The algorithm developed in this paper follows two paths. In one path, only stock-tank oil density is known and in the other path, the composition of the oil and gas are known. In both paths, water composition (salinity or total dissolved solids) is required. The paper includes a sensitivity study for the calculation and a field example in which this method is applied to a high porosity oil field where accurate density porosity calculation is hindered by fluid density variation related to water injection, depletion, large temperature anomalies, and drilling mud type. Results from density porosity calculations using "apparent fluid density" are compared with core porosity measurements and with other density log porosity calculations. Density porosity based on "apparent fluid density" matches core results more closely than other calculations.