Scenario modelling and optimisation of renewable energy integration for the energy transition

A large number of countries have engaged themselves in an energy transition towards more re- newable energy in their energy systems. Motivations stem mainly from the need to reduce CO2 emissions, and from a desire of their population to phase out technologies such as nuclear. Most of these countries promote biomass, wind and solar energy sources, among other possibilities. How- ever the current rate of deployment of renewable energy systems globally is not sufficient to reach the CO2 emissions reduction that would allow to maintain the global average temperature increase below the 2°C threshold. The main barriers to a wider integration of renewable energy systems are i) their limited realisable potential, ii) their still limited competitiveness, iii) their intermittence; iv) public acceptance often related to poor level of energy literacy amongst citizens. Citizens are key decision-makers. They must decide on energy policies and on the energy technologies they use, hence they have the power to foster or halt the energy transition. This thesis presents two different strategies for addressing the problem of the integration of re- newable energy sources for energy transitions. The first one (Chapter 1) consists in developing an energy modelling tool to help decision-makers understand the energy system and find their own answers. The modelling approach also includes a new methodology for the calculation of the total cost of a national energy system. A model of the Swiss energy system has been created following this approach, which serves as basis to develop the Swiss-Energyscope online calculator. This calculator and its model present an optimal trade-off between scientific rigour and user-friendliness, which allows the reproduction of the energy transition scenarios conceived by the Swiss Government, and consequently its use for energy policy making. The second strategy (chapters 2 and 3) profits from the possibilities offered by mathematical mod- elling and optimisation to analyse national energy systems, and derive insights for policy and decision-makers. First, a methodology using a mix-integer linear programming (MILP) model analyses biomass usage pathways to determine its optimal use in Switzeland in 2035. Second, in order to study the role of biomass, non-linear optimisation is applied to create future scenarios. (Chapter 3) focuses on the solutions to deal with the variability of renewable electricity. To this end, a MILP model with hourly time resolution is conceived to study the use of flexible electricity supply and demand options for the integration of renewable electricity. The optimisation methodologies are validated on case studies for the Swiss energy system. Regarding biomass, the results reveal that woody biomass chemical conversion technologies can allow for an overall better performance in terms of CO2 avoided emissions compared to direct combustion, as long as the produced biofuels are used in efficient technologies. Results also show that the combination of the gasification-methanation process of woody biomass with the production of H2 produced from excess electricity would allow to reduce the Swiss natural gas imports to zero by 2050. Concerning the integration of variable renewable electricity, the cost difference between using flexible electricity supply- and demand-options or electricity imports to deal with variable renewable electricity is below 2.5% of the total cost of the energy system.