Reservoir transport and poroelastic properties from oscillating pore pressure experiments

Abstract

Hydraulic transport properties of reservoir rocks, permeability and storage capacity are traditionally defined as rock properties, responsible for the passage of fluids through the porous rock sample, as well as their storage. The evaluation of both is an important part of any reservoir characterization workflow. Moreover, permeability and storage capacity are main inputs into any reservoir simulation study, routinely performed by reservoir engineers on almost any major oil and gas field in the world. An accurate reservoir simulation is essential for production forecast and economic analysis, hence the transport properties directly control the profitability of the petroleum reservoir and their estimation is vital for oil and gas industry. This thesis is devoted to an integrated study of reservoir rocks' hydraulic, streaming potential and poroelastic properties as measured with the oscillating pore pressure experiment. The oscillating pore pressure method is traditionally used to measure hydraulic transport properties. We modified the method and built an experimental setup, capable of measuring all aforementioned rock properties simultaneously. The measurements were carried out for four conventional reservoir-rock quality samples at a range of oscillation frequencies and effective stresses. An apparent frequency dependence of permeability and streaming potential coupling coefficient was observed. Measured frequency dispersion of drained poroelastic properties indicates an intrinsically inelastic nature of the porous mineral rock frame. Standard Linear Model demonstrated the best fit to the experimental dispersion data. Pore collapse and grain crushing effects took place during hydrostatic loading of the dolomitic sample and were observed in permeability, coupling coefficient and poroelastic measurements simultaneously. I established that hydraulically-measured storage capacities are overestimated by almost one order of magnitude when compared to elastically-derived ones. The fact that the values of storage capacities as estimated from the hydraulic component of the oscillating pore pressure experiment are unreliable was also demonstrated by comparing poroelastic Biot and Skempton coefficients. These coefficients were estimated both from hydraulic and strain measurements and the comparison of two datasets points out ambiguity of hydraulic measurements. I also introduce a novel method, which allowed us to estimate the permeability from the full range of acquired frequency data by utilizing a nonlinear least-squares regression. I additionally performed numerical simulation of oscillatory fluid flow. The simulated frequency-dependent results displayed an excellent agreement with both analytical solution and experimental data. This agreement proves that numerical simulation is a powerful tool in predicting frequency response of a porous rock sample to harmonic pore pressure excitations.