Exploring flexible strategies in engineering systems using screening models : applications to offshore petroleum projects

Abstract

Engineering Systems, such as offshore petroleum exploration and production systems, generally require a significant amount of capital investment under various technical and market uncertainties. Choosing appropriate designs and field development strategies is a very challenging task for decision makers because they need to integrate information from multiple disciplines to make decisions while the various uncertainties are still evolving. Traditional engineering practice often focuses on finding "the optimal" solution under deterministic assumptions very early in the conceptual study phase, which leaves a large amount of opportunity unexploited, particularly the value of flexible strategies. This thesis proposes a new approach to tackle this issue - exploring flexible strategies using midfidelity screening models. The screening models interconnect and model physical systems, project development, and economics quantitatively at the mid-fidelity level, which allows decision-makers to explore different strategies with significantly less computational effort compared to high fidelity models. The screening models are at a level of detail that gives reliable rank orders of different strategies under realistic assumptions. Flexibilities are identified and classified at strategic, tactical, and operational levels over a system's lifecycle. Intelligent decision rules will then exercise flexible strategies as uncertainties unfold. This approach can be applied as a "front-end" strategic tool to conduct virtual experiments. This helps identify good strategies from a large number of possibilities and then discipline-based tools can be used for detailed engineering design and economics evaluation.

(cont.) The present study implemented the use of such screening models for petroleum exploration and production projects. Through two simulation case studies, this thesis illustrates that flexible strategies can significantly improve a project's Expected Net Present Value (ENPV), mitigate downside risks, and capture upside opportunities. As shown in the flexible tieback oilfield development case study, the simulations predicted a 82% improvement of ENPV by enabling architectural and operational flexibility. The distributions of outcomes for different strategies are shown in terms of Value-at-Risk-Gain curves. This thesis develops and demonstrates a generic four-step process and a simulation framework for screening flexible strategies with multi-domain uncertainty for capital-intensive engineering systems.