| Wellbore instability problems in the petroleum industry result in annual losses exceeding US;To minimize the risk of wellbore failure during well drilling or production an approach is taken to quantify the effects of uncertainties such as in the in-situ stresses, rock characteristics, etc. Probabilistic and possibilistic techniques are used in this work. The probabilistic approach utilizes a Monte Carlo simulation, while the possibilistic approach uses fuzzy sets.;A wellbore stability model considering mechanical and chemical aspects of wellbore instability has been selected for this study. The focus is on proper mud weight design to avoid wellbore instabilities. In addition, a sensitivity analysis is performed to identify the parameters that are most significant in wellbore instability problems. The impact of various failure criteria also is investigated.;It is concluded that both the probabilistic approach and possibilistic approach are viable tools in the assessment of uncertainty effects of input data on wellbore stability model predictions. However, the probabilistic approach appears to be more suitable for defining the safe range of mud weights. The results of the sensitivity analysis indicate that uncertainties in the in-situ stresses, rock characterization parameters, and type of failure criterion used have very significant effects on wellbore stability model predictions. The main contribution of this work is an alternative well design approach involving uncertainty analysis. |
