ºìÐÓÊÓƵ

Topic

Understanding and accelerating the model-based design of multi-source energy systems

Department of Civil Engineering

Date & location

• Tuesday, January 14, 2024

• 10:00 A.M.

• Engineering Computer Science Building

• Room 467

Reviewers

Supervisory Committee

• Dr. Ralph Evins, Department of Civil Engineering, University of Victoria (Co-Supervisor)

• Dr. Curran Crawford, Mechanical Engineering, Uvic (Co-Supervisor)

• Dr. Madeleine McPherson, Department of Civil Engineering, UVic (Member) 

External Examiner

• Dr. Kristina Orehounig, Institute of Architectural Sciences, Tu Wien, Austria 

Chair of Oral Examination

• Dr. Gary Kuchar, Department of English, UVic

   

Abstract

The transition towards sustainable energy system requires innovative tools and methodologies to efficiently address the design of resilient and cost-effective multi energy systems (MES). The energy hub concept emerges as a powerful approach, capable to address optimal design and operation of energy systems while best exploiting potential synergies between technologies. Its scalability, limited by significant computational demands, is necessary to help capturing the multi-scale nature of energy systems and supporting a harmonized transition across all levels. This thesis investigate two key strategies to leverage this limitation. The first is the use of representative days to reduce the temporal complexity of the core application. The second is the development of surrogate models to rapidly estimate optimal designs. These strategies are examined across diverse contexts and system design problems through four independent but complementary studies. Each study led to a publication in international scientific journal or conference proceedings. 

The first study evaluates the robustness of selection methods for the use of representative days. The study highlights their sensitivity to context and the need for validation across multiple case studies. The second study examines the impact of shorter time horizons on the MES design. It emphasizes the strong influence on the system design, particularly on the design of storage. It also offers new insights into the design decision mechanisms. The third study compares district heating and cooling technologies. It demonstrates the superiority of the 5th generation network through diverse scenarios, emphasizing sensitivity to the topology and spatial demand distribution. The fourth study develops a novel surrogate modeling framework leveraging the use of machine-learning to address system design problems. The robustness of the framework is validated across diverse scenarios.  

Overall, this work emphasizes the importance of context-specific validations and sensitivity analyses. These aim to improve the understanding of design decision processes of energy hub applications, necessary to develop robust and effective complexity reduction methodologies. These contributions provide a foundation to improve the scalability and accessibility of MES design methods and, eventually, support the energy transition.