Toward a Unified Geomechanical Understanding: Core-Based Rock Mechanics and Evolving Earth Models in Heterogeneous Kuwait Reservoirs

Abstract This paper presents a standardized, core-based approach to mechanical property testing aimed at improving the accuracy of mechanical earth models in Kuwait's oil fields. Emphasis is placed on formations with high operational importance and limited prior mechanical data. The goal is to support drilling, stimulation, and injection decisions with a unified, continuously updated geomechanical framework that reflects the diversity in lithology, pressure regimes, and reservoir conditions across Kuwait. The core testing program was tailored to each formation's specific operational challenges and geomechanical characteristics. In the heterogeneous carbonates of Ratawi and Minagish, testing aimed to capture representative stress–strain responses to support stimulation, injection planning, and fracture risk assessment. In Wara, where wells intersect mechanically weak shales, the focus was on evaluating shear strength, anisotropy, and wellbore stability. Core selection followed a standardized methodology based on rock type, scratch test, log signatures, core quality, and spatial distribution. Laboratory data were rigorously quality-controlled and correlated with logs to build calibrated, evolving mechanical earth models. Rock strength varied significantly—from 3 to 30 ksi—even within the same formation, as confirmed by scratch, UCS, and multistage triaxial tests. When combined with borehole image logs and drilling event analysis, an unexpected pattern emerged: in some areas, stronger rocks failed during drilling, while weaker ones remained stable. This suggests local stress conditions, rather than strength alone, may drive failure. Multistage triaxial tests revealed plastic yielding and overcompaction in certain weak samples, indicating potential pore collapse or compaction-dominated failure. These observations point to the need for alternative failure criteria beyond conventional shear-based models. Biot coefficients varied significantly across formations, emphasizing the importance of direct measurement over assumed values. These observations provide new insights into differences in rock behavior during drilling, closure pressures, and stimulation outcomes. Core-based data enabled more accurate log correlations and significantly enhanced the predictive quality of geomechanical models. Historical models built on analogs often failed to capture evolving reservoir behavior. Consistent core selection and testing workflows led to high-confidence property-log correlations, even across geologically diverse regions. The development of a living geomechanical model—continuously refined with lab data and field observations—supports proactive decision-making and marks a shift from static interpretations to adaptive, data-driven reservoir management. This paper presents a standardized, data-driven approach for bridging laboratory core measurements with evolving earth models in reservoirs with diverse lithologies and pressure regimes. It introduces the concept of a "living" mechanical model, dynamically updated with new data. The approach enhances prediction accuracy and operational decision-making, particularly in complex settings with overpressure, depletion, or heterogeneous rock fabric.