Shadaab Syed Maghrabi

Modeling of Shale-erosion Behavior in Aqueous Drilling Fluids

Abstract Numerous shale-stability issues can occur while drilling with water-based muds (WBMs), including shale sloughing and cutting disintegration. These issues can be detrimental to the formation and pose difficulties with respect to rheology control, possibly reducing the rate of penetration (ROP). A “shale-erosion test” is a well-known laboratory test used to characterize the erosion of cuttings in WBMs. This paper documents a mathematical modeling tool known as an artificial neural network (ANN) used to model the erosion behavior of shale cutting in WBM. The ANN model establishes complex relationships between a set of inputs and an output based on computational modeling. For ANN modeling of shale-erosion behavior, the shale mineralogy and fluid composition constitute a set of inputs, while experimentally obtained “% erosion or % recovery” of the cuttings from the shale-erosion test represent the output. Experimental data for building the ANN model was obtained by performing approximately 150 standard shale-erosion tests using five different shales with varying mineralogy and WBMs with varying salt concentrations/types, shale stabilizers, and mud weights. For every test conducted, the input data (shale and fluid characteristics) and the output data (% recovery) was incorporated into the ANN model. The ANN model was then run to establish relationships between inputs and the output, which exhibited excellent correlation with R2 ≈ 0.85–0.90. The ANN model was successfully validated for an independent set of shale-fluid interactions. With the novel ANN model in place, erosion behavior of cuttings could be predicted in advance, thereby reducing the number of trials necessary in technical service labs. Mud engineers can use this model on a real-time basis as the shale chemistry varies with the depth of the formation drilling. The model could provide convenient measurement of fluid performance, enabling fluid optimization necessary to obtain desired shale behavior in advance, thereby minimizing drilling risks and costs associated with these oftentimes unpredictable shales.