Methodology to estimate the value of flexibility under endogenous and exogenous uncertainties

Risk analysis is essential in conceiving petroleum production projects, but it is, many times, underutilized to mitigate risk and increase value. Brazilian offshore fields tend to have high reservoir uncertainties and are heavily dependent on oil prices. Current projects have low flexibility and, in many cases, are ill suited to handle uncertainties. However, flexibilities can be embedded in projects to better prepare for uncertain scenarios. To support the decision process, we must be able to estimate the value of flexibility. The lack of methodologies to do so using reservoir simulations is the main motivation for this work. This paper proposes a five-stage methodology to estimate the value of flexibility in petroleum production projects under endogenous and exogenous uncertainties. Each stage is split into steps to identify the details of problem modeling and strategy selection. The EVoF (expected value of the flexibility) is a probabilistic concept similar to EVoI (expected value of information) based on projections and calculated prior to starting a project. The use of reservoir simulation models considering flexibility to predict production improves estimates of EVoF. High-fidelity simulation models capture flow and geological characterization in porous media, and representative models evaluate uncertainties. Economic uncertainty is represented over time using geometric Brownian motion. The overall process integrates the complex representation of attributes from different sources with a dynamic approach to allow the implementation of flexibility over time. This new integrated approach focuses on identifying drivers that evidence the need for flexible projects and investigates flexibilities to face uncertainties while observing their impact on results. Therefore, decision makers are able to compare rigid and flexible projects based on reliable predictions. The study case highlights various aspects related to the conception of flexible projects, such as proper choice of design variables and identification of the potential of the flexibility, and it shows that the methodology is suitable for a decision-making process that integrates simulation modeling and uncertainty analysis. We estimate EVoF from the distribution of NPV differences and show the increase of EVoF as design variables and decision rules change and optimized strategies for representative models are applied.