Data Quality Control to Increase Geomechanical Modelling Accuracy

Abstract

Abstract

Most of the techniques used in geomechanics require correlations of rock properties and stress, and assume the use of high quality data without discussion of how the data was acquired or its validity. The logs used in building a model are affected by factors beyond variations in mechanical rock properties or stresses. These variations are assumed to be valid without quality control of the measurement itself, the resulting error can significantly impact the final result. Considering the origin of the data and the factors that influence the measurements, properly setting up the tool's recording parameters such as it's real time compressional slowness reading tool mode, can help the analyst to mitigate some types of model uncertainties. Furthermore, some model inputs, such as lithology columns, are the product other professional's interpretations and the result depends on the background of the analyst and the purpose of the interpretation.

Based on recent studies performed for different companies in Latin America, we defined a suite of techniques for quality control of the data needed to determine the in-situ stress magnitudes and orientations, and the rock mechanical properties. This initial workflow provides guidelines to identify quality control problems and confirm whether the data is valid before used in geomechanical calculations. These techniques are broadly used during petrophysical analysis where the interpreted data is provided to geomechanical analysts for further modeling.

The application of these techniques can help to avoid common drilling problems, such as kicks, inflows, lost circulation and wellbore instability, especially during real time monitoring, where the data uploaded generally can't be assumed to be properly constrained and processed in real-time. In this paper, we present real cases from Mexico, Argentina and Peru of how misinterpretation of the data could affect the modeling results. Examples from before and after performing the quality control are presented.

AAPG Datapages/Search and Discovery Article #90260 © 2016 AAPG/SEG International Conference & Exhibition, Cancun, Mexico, September 6-9, 2016