Mud-Filtrate Invasion in Laminated and Spatially Heterogeneous Rocks: High-Resolution In-Situ Visualization and Analysis Using Time-Lapse X-Ray Microcomputed Tomography (Micro-CT)

Abstract

High-resolution X-ray microcomputed tomography (micro-CT) was used to investigate the process of mud-filtrate invasion in laminated and spatially heterogeneous rocks. Laboratory experiments involved the injection of pressurized water-based drilling mud into a borehole located at the center of cylindrical rock core samples that exhibited a range of petrophysical properties, including a cross-bedded sandstone, a finely laminated anisotropic sandstone, an oolitic limestone with significant pore-scale heterogeneity, and a spatially heterogeneous vuggy dolomite. As drilling mud was injected into the initially dry (air-saturated) core samples, high-resolution X-ray micro-CT scanning was continuously performed to visualize the flow of mud filtrate through the core samples during mud-filtrate invasion. Because the properties of the drilling mud were consistent for all of the experimental cases, differences observed in the invasion behavior could be attributed to the unique properties of the core samples. Despite exhibiting a wide range of petrophysical properties, including permeabilities ranging from 8.9 to over 1,800 md, the mud-filtrate invasion rate was similar for all of the experimental cases, aside from a brief initial period of spurt loss that lasted only a few seconds, indicating that properties of the deposited mudcake were responsible for controlling the rate of mud-filtrate invasion rather than the unique properties of the core samples. Conversely, displacement of the invasion front and the spatial distribution of mud filtrate in the core samples differed significantly between cases, both spatially and as a function of time. Initially, mud-filtrate flow in the core samples was dominated by viscous forces, and air was forcibly displaced from around the borehole. Shortly thereafter, we observed a rapid transition to capillary-dominated flow as the deposition of mudcake on the borehole wall quickly reduced the rate of mud-filtrate invasion. As a result of the displacement process being controlled by capillary forces, we observed that mud filtrate within the invaded zone became more uniformly distributed over time, with mud filtrate from around the borehole being drawn further into the core samples. The average mud-filtrate saturation within the invaded zone initially decreased before stabilizing between 43 and 51%, indicating that a significant volume of trapped gas (air) remained behind the invasion front. Experimental results indicate that borehole measurements acquired at early times during the mud-filtrate invasion process, such as those acquired while drilling, may be more sensitive to small-scale heterogeneity than wireline measurements acquired later once capillary forces have caused the mud-filtrate saturation within the invaded zone to become more uniformly distributed. Furthermore, because capillary forces were primarily responsible for controlling the mud-filtrate saturation in the invaded zone, fluids from around the borehole may not be displaced to their true residual saturation, potentially causing nuclear and shallow-sensing resistivity measurements to underestimate the actual mobile hydrocarbon saturation of the formation.