Re-Examination of Offset Field-Data Calibration for the Purpose of Abnormal Pressure Prediction

Abstract

Abstract Until now, the calibration portion of the industrially available techniques for predicting abnormal pressure heavily depends on statistical regression analysis, as an analytical tool. For convenience and timely use, such statistical regression analysis has been programmed into the Excel program of Personal Computers. Unfortunately, where there is high demand for accuracy and data reproducibility, this analytical tool is secretly less accurate. Therefore, this study introduces a new analytical tool that conveniently alleviates the potential inaccuracies of the statistical regression analysis. The implementation of the new analytical tool has produced striking improvements in the accuracy of prediction, and certainty of calibrated data reproduction when required as the predicted values. Introduction The unexpected encountering of abnormally high formation pressure has been posing serious drilling problems that have resulted in uncontrollable disastrous drilling operations all over the world. From the analysis of overpressured sediments in various geographical provinces, it has been found that no satisfactory, single universal correlation can be generated between geophysical or well log parameters and the overpressure values in all geographical provinces.1-35 However, adequate correlation has been attained for individual geographical province. Such geophysical and well log parameters include the earth effective stress, sonic interval transit time, seismic wave velocity, rock resistivity and conductivity, and "d" exponent. Therefore, the most reliable drilling techniques for detecting and predicting overpressures have been based on information from each geographical province. The approaches of overpressure detection and prediction by the available techniques can be classified into pre-drill, and during drilling. The pre-drill detection/prediction approach is achieved through the calibration of surface seismic survey data and the overpressure values from the offset well located in the same geographical province. During drilling, correlations relating drilling parameters and earth properties have been unsatisfactorily developed. The origin or source of formation overpressure has, initially been attributed to the undercompaction of shale sediments, which, consequently, results in significantly high porosity in such sediments. Under normal compaction of sediments or intervals, a plot of interval seismic velocities with depth should yield continuous increments of velocity with depth of burial. Any reversal from this normal trend is attributed to the presence of abnormal formation pressure. Based on this concept, many studies4-15,20,23 were carried out to improve the pre-drill pressure prediction that is highly needed for well programs, such as mud weight determination and casing setting depths. All of the initially available pre-drill prediction techniques5-10,31-34 were based on the comparison of calibrated data of surface seismic survey and measured formation pressures at the offset wells, with surface seismic data at the proposed drilling location. Although some degrees in accuracy of predicted pressure were obtained, well blowout during drilling operations never ceased due to the persistent pressure deviations between the predicted and actual pressure values. These earlier pre-drill pressure prediction approaches involve:plot of the shale derived properties with the formation pressure gradients5-10;plot of seismic velocities with formation pressure gradients9-12,19&20 anduse of overlays on the plot of depths versus shale resistivities or interval transit time12-14.