Šárka’s project is associated with MoZEES RA4: Policy & techno-economic analysis (Task 3.4), and can be viewed at NTNU Open. The work has been carried out at the department of industrial economics and technology management at the Norwegian University of Science and Technology (NTNU) and supervised by Professor Peter Schütz and Professor Asgier Tomasgaard.

Summary

According to the Paris Agreement on climate change, CO2 emissions must be decreased by 40% towards 2030. Decarbonization of the transportation sector is a crucial step to meet the emission reduction targets in the Paris Agreement. Hydrogen is considered a promising low-emission fuel alternative in the transportation sector. However, low experience with hydrogen and limited availability are the main challenges when introducing hydrogen as a low-emission fuel in the transportation sector. Hence, designing the hydrogen production infrastructure and providing a reliable hydrogen supply is essential to encourage potential customers to switch from fossil fuels to hydrogen.

This thesis formulates optimization models and develops efficient solution methods to solve the problem of designing hydrogen production infrastructure in Norway. The papers in this thesis contribute to the state-of-the-art on deterministic and stochastic multi-period facility location problems with capacity expansion. To accurately represent the costs of hydrogen production, we consider specific aspects of hydrogen production technologies in our mathematical models. These aspects include economies of scale in investment, capacity-dependent short-term production costs, and considerations of minimum production requirements. Further, we present solution methods for deterministic as well as stochastic formulations based on Lagrangian relaxation. Furthermore, we provide managerial insight into the cost analysis of hydrogen production in Norway.

This thesis consists of four research papers. In Paper I, we formulate and solve the problem of locating hydrogen production facilities in Norway considering deterministic demand. In Paper II, we present a solution method based on Lagrangian relaxation for the problem introduced in Paper I. In the last two papers, future demand is considered an uncertain parameter. In Paper III, we use sample average approximation to solve the problem, while in Paper IV, a solution method based on Lagrangian relaxation is developed.