A new methodology to reduce uncertainties in reservoir simulation models using observed data and sampling techniques

This paper presents a new method to reduce uncertainties in reservoir simulation models using observed data and sampling techniques. The proposed methodology is able to deal with problems with a high number of reservoir uncertain attributes and includes the development of a probability redistribution algorithm using observed data. The use of Latin Hypercube technique in the construction of uncertainty curves that is a quantitative representation of the overall uncertainty of the problem studied was also proposed. Based on new probability distributions, selective samples are carried out through the Latin Hypercube technique. The methodology was evaluated using two case studies. The first one, used for validation purposes, is a simple reservoir with 8 attributes; the second one is a more complex case with 16 attributes. The results presented in the paper showed that the proposed methodology can be efficiently used in the integrated study of history matching and uncertainty analysis, providing a practical way to increase the reliability of prediction through reservoir simulation models reducing the uncertainty through observed data.