This paper focuses on the case study of the geomechanical evaluation of a tight gas reservoir in Oman. Rock strength was characterized using data inputs/measurements with the objective to reduce uncertainty on predictions of wellbore stability in four deep gas wells. As a primary input for many standard geomechanical models, rock strength parameter is routinely measured on rock samples using triaxial/uniaxial tests. This parameter is traditionally named as Uniaxial Compressive Strength (UCS) measured during the axial loading and crushing of cylindrical "core plugs" extracted from cores. Although considered as a standard for rock strength evaluation, this method has some limitations such as (i) sample destruction, (ii) natural bias in weak formations (iii) natural dispersion, (iv) discretization of the measurement along sampled intervals and (v) sample preparation challenges.
The scratch test has been developed as a rapid and cost effective rock strength testing method addressing these limitations. The relative advantages of the scratch test as an alternative to standard rock strength testing are discussed and its added value demonstrated in the context of practical applications for a tight gas reservoir. Geomechanical data was acquired from four different wells with the following measurements: (i) Acoustic wireline log, (ii) Uniaxial compressive strength resulting from laboratory tests on plug samples, (iii) a continuous high-resolution strength profile interpreted from scratch tests performed on whole cores. Then, the data from all these sources has been integrated following a dedicated workflow designed to reduce the uncertainty in the output of strength models from wireline logs, through careful handling of data resolution differences and heterogeneity mapping.
Encouraging correlations between core-based measurements and scratch test suggests that the scratch test has emerged as a valid alternative to standard rock mechanical tests in suitable situations. Robust statistics are provided for strength and explanations are proposed for outlier values from tests on plug samples. Furthermore, the continuity and the high resolution of the strength profile enables a much better calibration of strength proxies from acoustic wireline logs. Finally, the scratch test yield values from shale intervals that were traditionally overlooked by plug site selections for rock mechanical testing.
The strength assessment protocol from the scratch test handles rock heterogeneity with a much larger data set than conventional rock strength testing methods. This enables more robust core property-log calibration at different length scales. Such continuous high resolution profiles of rock properties leads to a significant reduction in uncertainty in petrophysical and geomechanics models, and better decision making in well design and field management.