Low-Carbon Energy System Design: Methods, Software and Applications

Global warming and climate change induced by anthropogenic greenhouse gas emissions have recently been recognised as major threats to human societies. One of the cornerstones of most plans aiming to reduce anthropogenic emissions consists in massively harnessing renewable resources for power generation and electrifying a variety of end-uses, both directly and indirectly. However, the integration of renewable resources presents considerable challenges that notably stem from their inherent variability and stochasticity. Several solutions have been advocated, including tightly integrating energy systems and carriers in order to provide flexibility to the power system or produce synthetic fuels and feedstocks from renewable electricity, as well as taking advantage of the heterogeneous distribution of renewable resources in space and time to site renewable power plants strategically and reduce the variability of their output.

The first part of this thesis focusses on the planning of multi-carrier energy systems. First, a modelling abstraction for structured mixed-integer linear programs that is particularly well-suited for tackling such problems is presented. Building upon this framework, a modelling language and open-source tool for mathematical programming are then introduced. The language is designed to simplify problem encoding, enable model re-use, speed up model generation, and interface with both off-the-shelf and specialised structure-exploiting solvers.

Two case studies leveraging these tools are then proposed. The first case study analyses the extent to which power-to-gas and carbon capture technologies can help to achieve deep cross-sector decarbonisation targets and increase renewable energy penetration in Belgium. Results suggest that power-to-gas technologies can only play a minor supporting role in cross-sector decarbonisation strategies, while carbon capture technologies form a key plank of low-carbon energy systems designed to serve high methane and hydrogen demand levels, although they also exacerbate fossil fuel dependence. The second case study investigates the economics of carbon-neutral synthetic methane production from solar and wind energy in North Africa. Results show that the cost of synthetic methane delivery to Northwestern European consumers would be around $150$ \euro/MWh (HHV) by 2030 and suggest that using time-resolved, integrated energy supply chain models is essential to produce accurate cost estimates.

The second part of this thesis analyses the spatiotemporal complementarity that renewable power plants exhibit and explores the benefits that strategically siting them may bring about. First, a framework for evaluating renewable resource complementarity is presented. In essence, this framework measures the empirical probability of observing simultaneous low power generation events across most plants considered. The framework is illustrated by a case study investigating the complementarity between wind regimes within and across continental France and Southern Greenland. Results reveal that a reduction in the occurrence of system-wide low RES generation events can be achieved when deployment patterns include locations from both areas simultaneously.

Then, the problem of siting renewable power plants while taking their spatiotemporal complementarity into account is cast as a combinatorial optimisation problem and closely analysed. Multiple heuristics proposed to solve it are tested in a case study using large amounts of high resolution climatological data to site onshore wind power plants in Europe. Several heuristics are found to consistently provide solutions that are at least as good or strictly better than the ones returned by a state-of-the-art mixed-integer programming solver at a fraction of the computational cost. Results also suggest that the wind does not always blow hard somewhere in Europe, and it therefore seems unlikely that onshore wind power plants will be capable of supplying a constant share of the electricity demand at all times, even when strategically siting them on a continental scale.

Finally, the siting framework is extended and combined with a joint generation, transmission and storage capacity expansion planning model to assess the impact of various renewable power plant siting strategies on power system design and economics. Results suggest that simply selecting onshore wind locations in the most productive European regions could suffice to obtain sets of locations that display a relatively high degree of complementarity and yield the cheapest system designs.