Developing flexible management strategies in infrastructure: The sequential expansion problem for infrastructure analysis (SEPIA). (August 2020)

• Record Type: Journal Article
• Authors: Lozano, Jorge-Mario; Zuluaga, Santiago; Sánchez-Silva, Mauricio
• Is Part Of: Reliability engineering & system safety. Volume 200(2020)
• Publication Date: 2020-08
• Subjects: Infrastructure expansion -- Sequential decision-making -- Flexibility -- Value of information -- Infrastructure systems; Reliability (Engineering) -- Periodicals; System safety -- Periodicals; Industrial safety -- Periodicals; Fiabilité -- Périodiques; Sécurité des systèmes -- Périodiques; Sécurité du travail -- Périodiques; 620.00452
• DOI: 10.1016/j.ress.2020.106951
• Languages: English
• ISSNs: 0951-8320

Abstract

Highlights: Present a model to find optimal expansion decisions for infrastructure networks. The model incorporates flexibility in decision-making for infrastructure. Address the problem of sequential decision-making under uncertainty. Compare the model against traditional methodologies, which are static and myopic. Show that flexibility/adaptability lead to efficient decisions in infrastructure. Abstract: Infrastructure systems are usually designed to operate over long time periods. This leads to complex decisions in which the incorporation of risk management strategies is crucial to guarantee a successful operation. Traditional engineering practices focus on finding a fixed set of parameters that fulfill operational objectives throughout the project. This approach is mostly prescriptive and therefore cannot adapt to unexpected changes of external conditions. This work considers an infrastructure network designed to attend a stochastic demand expected to increase over time. The paper presents a model that is capable of integrating different aspects of the decision-making process in a cost-effective manner and incorporates the idea of flexibility. The proposed approach is referred to as Sequential Expansion Problem for Infrastructure Analysis (SEPIA). This approach is based on real performance data acquisition, optimization of decisions given possible future scenarios, and a scheduling strategy that allows operating interconnected components in an infrastructure network. The paper also presents a computational implementation that can effectively find solutions for large decision spaces. Through two examples, it is shown that incorporating flexibility in the decision strategies allows simulating more realistic decision scenarios, as well as improving the long-term performance of the system compared to traditional approaches.