A New Tool To Evaluate the Feasibility of Petroleum Exploration Projects Using a Combination of Deterministic and Probabilistic Methods

Abstract

Abstract In this paper we have presented a new tool to evaluate the feasibility of an exploration and production project. An exploration and production project can be evaluated for feasibility using deterministic methods or probabilistic methods. These two methods vary chiefly in the way they manage uncertainty in the input parameters of the decision model. Deterministic methods neglect the changes in the input parameters with time, but focus on estimating the probabilities of the different outcomes possible at each stage in the decision model. A classical tool of the deterministic method is the decision tree. Probabilistic methods assume that a quantity to be estimated (e.g. reserves) can never be accurately determined. Rather it presents the possible number as a distribution. A typical tool in decsion making using probablitics methods is the Monte Carlo simulation. Each of these two methods, probabalistic and deterministic have thier own advantages and limitations. Our paper discusses combining the benefits of decision trees and Monte Carlo simulation. The new tool combines the simplicity of the deterministic approach and yet takes care of uncertainty in the manner of the probabilistic method. The result obtained by the deterministic method is fed into the probabilistic model as in input. The probability function for each input parameter in the probabilistic method is defined. Other input constraints and limitations are imposed and the model is simulated. The resulting output is analyzed and the associated risk or uncertainty for each output parameter is quantified. The output of the probabilistic model is fed back into the decision tree and the optimum decision is determined. The overall uncertainty in each decision outcome is determined by combining the individual uncertainties. The various options available in the project execution are compared using the new tool and the best project is selected on the basis of the highest return on investment or lowest risk. The advantages of the new tool as compared to using only deterministic or probabilistic methods are also demonstrated.