New Analytical Method To Calculate Matrix- and Fluid-Corrected Total Porosity in Organic Shale

Abstract

Estimation of total porosity from neutron and density porosity logs in organic shale (source rock) is challenging because these logs are substantially affected by fluid and matrix-composition effects. Conventional interpretation of neutron and density porosity logs often includes corrections for shale concentration in which the main objective is to improve the calculation of nonshale porosity in hydrocarbon-bearing zones. These corrections are not desirable in unconventional rock formations because shale pores can be hydrocarbon-saturated. Neutron and density porosity readings across shale zones are sometimes averaged by use of the root-mean-square (RMS) method.We introduce a new and simple analytical expression for total porosity that effectively separates both matrix and fluid effects on neutron and density porosity logs. The expression stems from a new nonlinear mixing law for neutron migration length that is coupled with the linear-density mixing law to calculate total porosity and fluid density. The method is applied in two sequential steps: First, separate corrections for only matrix effects are implemented to enhance the neutron-density crossover for qualitative interpretation of fluid type; second, the coupled equation is used to estimate fluid density and actual porosity devoid of matrix and fluid effects. Calculated porosity and fluid density can be used further to calculate water saturation from density logs. One remarkable feature of this method is the ease with which it can be applied to obtain accurate and reliable results. Application of the method only requires knowledge of single-component nuclear properties and mineral volumetric concentrations. One can obtain nuclear properties from a set of charts for multiple fluid types and minerals provided in this paper, whereas one can calculate mineral concentrations on the basis of available triple combo logs or gamma ray spectroscopy logs.Two synthetic and four field examples (two conventional and two shale-gas reservoirs) are used to test the method. First, we describe an application in a conventional siliciclastic sedimentary sequence in which only shale concentration calculated from gamma ray logs is required to improve the estimation of porosity in shaly sections. Second, we document several applications in which gamma ray spectroscopy logs are used together with a reliable hypothesis for clay type to define mineral properties. Results compare well with nuclear-magnetic-resonance (NMR) and core measurements, whereas the new method outperforms the conventional RMS procedure, especially in the cases of gas-bearing, low-porosity organic shale. The new analytical method can be readily implemented on an Excel spreadsheet and requires minimal adjustments for its operation.

CO2-Containing Reservoir Evaluation Based on the Joint Inversion of Nuclear Logs and Resistivity Logs: A Case Study of Ultrahigh-Temperature and High-Pressure Gas Reservoirs in the Yinggehai Basin, China

“…The rapid simulation of neutron logs is mainly achieved by calculating the migration length. 13 , 24 According to the model shown in Figure 2 , the migration length can be expressed as where ξ is the migration length, ξ w is the migration length for pure water, ξ CO2 is the migration length for CO 2 , ξ CH4 is the migration length for CH 4 , and ξ m is the migration length for matrix minerals. η is a nonlinear coefficient relevant to shale content: …”

“…The resistivity correlation value in the Jacobian matrix is obtained using the finite difference method. The neutron porosity correlation value (migration length) in the Jacobian matrix is obtained using the method proposed by Ortega and Torres-Verdı́n (2015) . The stabilization parameter α in eq is automatically selected by the L-curve method …”

“…The neutron porosity correlation value (migration length) in the Jacobian matrix is obtained using the method proposed by Ortega and Torres-Verdı ́n (2015). 24 The stabilization parameter α in eq 1 is automatically selected by the L-curve method. 25 3.2.…”

“…28,29 The rapid simulation of neutron logs is mainly achieved by calculating the migration length. 13,24 According to the model shown in Figure 2, the migration length can be expressed as…”