Real Options Models for Maintenance Decision Making for Offshore Jacket Platforms Available to Purchase

Real options models are currently available as one of the best tools for the assessment of investment projects. This is so mainly due to the capability of the real options models to: (1) account for uncertainties in financial variables that are crucial to the investment project; and (2) quantify the value of the possibility to make a decision on whether to defer, abandon, expand or reduce the project at one or several points along time. Recently, some researchers have proposed the use of real options models for the assessment of infrastructure projects for hydrocarbon exploitation from an economics point of view. The objective of this work is to develop real options models for decision making regarding inspection, maintenance and decommissioning of offshore facilities taking into account the financial and technical aspects of the project. In all cases it is considered that at some point in the future, within the service lifetime of the structure, the decision maker will have an option to carry or not an inspection, and take or not a maintenance or decommissioning action, which will determine the structural and financial performance of the project for its remaining lifetime. The in-service times with no structural failure and the rehabilitation times are modeled as random variables. The cash flows are modeled as stochastic processes considering interruption of operation due to repairs after failure. Analytical expressions are derived for the computation of structural reliability and availability depending upon maintenance actions. An example is given for a jacket platform subjected to fatigue deterioration and damage. Simple and compound options of maintenance and decommissioning options are analyzed. The value of the project is computed by means of an approach similar to that of Black and Scholes for financial options [2]. The results are compared to those obtained under the traditional Net Present Value approach.