Methodology of uncertainty analysis prediction based on multi-well data fusion

Abstract

During drilling activities, geological parameters of a well to be drilled (target well) can be predicted within a limited interval based on multi-well data fusion which aims at ensuring a drilling safety, enhancement of drilling efficiency, reduction of drilling cost as well as acquiring accurate measurements in respect to Oil and Gas protection layers. This work presents a method of uncertainty analysis prediction of pressures using fusion data (formation pressures) from adjacent multi-well. The Eaton method, effective stress theory, and mathematical confidence interval were the various methods used to establish the formation pressure matrix of the target well. The results revealed that due to the complexity and variability of the formations, data interpretation errors of the geological parameters were inevitable. Therefore, the probability density distribution function was established through stratigraphy, probability statistics, and information diffusion. Moreover, the real value of the wells’ formation pressure (target well) was within the distribution interval of multi-well data fusion. Hence, the developed method cannot only effectively reduce the interval of geological parameter of the target well but also enhance the accuracy of parameters prediction.