SEQUENTIAL AND SYSTEMS APPROACHES FOR EVALUATING INVESTMENT DECISIONS—INFLUENCE OF FUNCTIONAL DEPENDENCIES AND INTERACTIONS

Abstract

Development decision-making practices in the oil and gas industry focus on detailed modelling of every decision parameter—such as reserves, production schedule, facilities design, costs and prices—individually, without due attention to the dependencies and interactions between these parameters. Dependencies and interactions are captured in modelling of uncertainties in some parts of the system (such as reserves and production schedules) but not in others.Modelling uncertainty without modelling dependencies fails to benefit from investing in the values of flexibility and information. Modelling of dependencies and interactions promotes the integration of all the relevant parameters of petroleum projects in a holistic manner.Separate and sequential modelling of individual components of investment decisions limits the ability to examine how changes in one component impact on other components of the system. On the other hand, the systems approach views a development decision as an integrated unit, including components such as reserves, production schedule, facilities design, costs and prices.This paper hypothesises that modelling dependencies and interactions should not be limited to estimating reserves, but should be extended to model the total petroleum system. We believe that there is potential for adding value to petroleum projects by modelling dependencies and interactions in a holistic systems-based stochastic environment.The objective of this paper is to demonstrate the impact of functional dependencies and interactions on the development decision of a hypothetical offshore oil field. Specifically, we show the difference between the treatment of functional dependencies and interactions together with their implications for the sequential and systems approaches using Monte Carlo Simulation (MCS) based stochastic modelling to capture uncertainties.The systems approach captures interactions and dependencies while the sequential approach ignores them. Ignoring interactions leads to under-estimating the mean Net Present Value (NPV) as well as the standard deviation (by 54% and 44% respectively in our example). Furthermore, in our example, the P10, P50 and P90 (NPV’s) are all under-estimated by 20%, 50% and 50% respectively. These results clearly show that proper systematic treatment of dependencies and interactions can have significant impact on petroleum project evaluation.