FEASIBILITY OF DIGITAL ROCK PHYSICS FOR PETROPHYSICAL CHARACTERIZATION OF COMPLEX RESERVOIRS

Abstract

Hydrocarbons (practically oil & natural gas) are one of the major sources of energy in the entire world today and are expected to continue their domination as a vital resource in the near future (Barclays, 2019). These are found trapped in the subsurface and are formed due to deposition and subsequent decay of organic matter, happened over geological ages (millions of years). The hydrocarbon reserves are differentiated as conventional and unconventional types, depending upon whether the source and host rock is same or different, respectively. Carbonate reservoirs are those reservoirs which though are conventional type but behave as unconventional one’s due to the diagenetic processes occurring in them that results into heterogeneous distribution of grains and pores. This thesis discusses the applications of Digital Rock Physics (DRP) technique in determining the petrophysical properties of a complex and heterogeneous reservoir, like Carbonate reservoir. Initially, laboratory measurements are performed to determine the petrophysical properties that characterize a reservoir. However, due to limited availability of cores and non-repeatability of measurement values which adds non-negligible cost to any E&P program, these measurements are rarely performed. Thus, scanned volumes are preferred to characterize a reservoir in terms of its petrophysical properties. I have used two sets of Carbonate samples in the present work. The first set consists of eight (08) samples from Middle Eastern Reservoir and the second set consists of eight (08) samples from Western Offshore Basin, India. DRP workflow has three (03) main steps, beginning with image acquisition or scanning of the samples. The carbonate samples were scanned using Micro X-ray Computed Tomography (CT) instrument facility available at Colorado School of Mines, Golden, USA. This is performed at multiple resolution, in order to map the pores of all sizes. The first set of samples was scanned at ii a resolution of 25m and second set of samples was initially scanned at a resolution of 12m and then three (03) samples were selected for multi-resolution scanning at a resolution of 25m and 3m. Conditioning of acquired images helps in improving the appearance of multiscale features by enhancing signal-to-noise ratio. The conditioned image is then processed for segregation of pore and grain space by implementing segmentation algorithms (Gonzalez and Woods, 2008). In this thesis, the conditioning and segregation is performed by running a sample through a set of workflows defined by a combination of different conditioning and segmentation algorithms. This is performed in order to achieve a robust segmentation algorithm for the set of samples used in this work. Finally, DRP provides a platform to simulate digitized rock volumes with the physical mechanisms to understand the dominant flow regimes and simulate ultimate recovery. So, I present a quantitative estimate of total porosity and its size fractionation using digital rock workflows. The porosity fractions were categorized into three sizes: micro, meso and macro pore types based on the modification of the classification system in carbonates from Lonoy (2006) to a classification based on pore area.. This porosity partitioning allows us to understand dominant flow regimes in complex reservoirs and determine fluid flow regimes (continuum to statistical (free molecule)) based on the calculated Knudsen number (KN) for gas reservoirs Thus, the present work discusses the effect of resolution on the static and dynamic petrophysical property measured from segmentation and simulation workflow. This work can further be extended to study the effect of other parameters and determine different petrophysical properties from simulation of respective physical processes.