Wellbore Stability Analysis Based on Sensitivity and Uncertainty Analysis

Abstract A methodology based on sensitivity and uncertainty analysis is presented for performing a wellbore stability analysis in order to quantify model uncertainty and improve the predictions made. This methodology is applied to the safe mud window design for a study case. A wellbore stability model is set up, as of constitutive equation and failure criteria selection. Based on available data, a stochastic approach is used to assess the uncertainty associated to input parameters, or otherwise other alternatives like expert opinion are utilized. Secondly, a sensitivity analysis is carried out by means of OAT and RSM methods displaying results on tornado and spider diagrams. Finally, an uncertainty analysis is performed using Monte Carlo simulations for quantifying uncertainty in model outputs and obtaining the required mud weight to drill successfully as a probability function. Four approaches for assessing uncertainty in input data of wellbore stability models are presented. Sensitivity analyses showed maximum horizontal stress and cohesion as the most critical parameters in collapse pressure determination while minimum and maximum horizontal stresses were the major variables in fracture pressure evaluation. On other hand, vertical stress and Poisson ratio were considered as constants during the uncertainty evaluation due to their subtle impact in model results. Ultimately, an uncertainty analysis is presented via probabilistic graphs, which enables a better decision making since uncertainty in input data is considered and safe mud window is obtained as a probability function of drilling successfully. The workflow proposed makes up a complete uncertainty study useful for developing wellbore stability analyses which optimizes the decision making in drilling process design. Aspects like evaluation of input data uncertainty and incorporation of sensitivity analyses are novel components for complementing the simulation Montecarlo as uncertainty analysis method and risk analysis tool.