Incorporating Uncertainties in Renewable Energy System Design: A Novel Stochastic Optimization Technique for Improved Design Strategies

Details

Supervisors

Contino, Francesco

Degree label

Mobilité entrante en sciences de l'ingénieur

Abstract

Designing future energy systems requires information on regulatory and economic performance as well as information about the future climate and energy demand. Despite these parameters are likely to become uncertainties due to its variations during the system lifetime, it is still more an exception rather than the norm considering them in the design procedure. Also, usually the data resources on these variations are limited. This omission of the uncertainty can result in a large difference between the simulated model and its actual performance, leading to a kill-by-randomness of the system. A previous collaboration between Universitat Politècnica de Catalunya (UPC) and Université Catholique de Louvain, published in Energies [1], extensively examined these economic and regulatory uncertainties. This MSc thesis builds upon the groundwork laid by the aforementioned collaboration, guided by the expertise of Prof. Maria Elena Martin Cañadas (UPC), Prof. Francesco Contino, and Dr. Diederik Coppitters (UCLouvain). The primary focus of this thesis is to develop and implement a novel stochastic optimization method, that will aim to optimize renewable energy systems, considering both economic and regulatory uncertainties, focusing on expected performance, and robustness [2]. The work begins with an extensive literature review on existing stochastic optimization methods [3]–[6]. Afterwards, the student will implement a new renewable equipment in an established Python package on optimization under uncertainty, developed at iMMC in UCLouvain [7]. The student’s goal is to develop and integrate novel approaches, relevant to renewable energy system design under uncertainty, and analyse its performance and possible scenarios in the location of Brussels. Also, a decision tree will be done in order to identify the key attributes and splits in the system’s dataset. The successful implementation of the proposed stochastic optimization method is expected to bring valuable information and improvements in renewable energy system design in an uncertain operating environment. Potential results of this project involve sophisticated design strategies for hybrid renewable energy systems optimized not only for expected performance but also proving increased robustness in the search of unforeseen uncertainties.