Maréchal, FrançoisKantor, Ivan DanielSantecchia, Alessio2022-04-072022-04-072022-04-07202210.5075/epfl-thesis-8424https://infoscience.epfl.ch/handle/20.500.14299/186889Europe is currently transitioning from fossil energy sources to renewable generation of electric power. Although fundamental to reach net-zero targets, intermittent renewables are disrupting conventional methods used in operational planning and design of processes and the electrical grid. This work focuses on optimal operating strategies in the field of process industries and design of large-scale power systems, with the aim of facilitating the path towards a renewable, robust and intelligent energy system. The first step requires deep analysis of environmental impact from the electrical grid; therefore, the first development is creation of a novel dynamic life-cycle assessment-based (LCA) tool to construct series of impact data from the electrical grid. The tool connects to public databases to quantify the real-time price and environmental impact of electricity consumption. Historical grid impacts and weather data are used to train random forest regressors, which are able to forecast week-ahead carbon emissions in each country with hourly granularity. The forecasts are further embedded into a model predictive controller (MPC) that optimizes short-term scheduling of an industrial batch process. The method follows a rolling scheduling approach that allows for coordination between production scheduling and procurement of electric power targeting minimum environmental impact. The comparison between avoided emissions (5-30\% reduction compared to BAU) and resultant operating cost enables calculation of the minimum carbon tax that would favour adoption of carbon abatement strategies in industry. The same use case is used to introduce the second novel concept of representing flexible processes as equivalent batteries, which store electricity from low-cost periods as intermediate products and consume the embedded energy during high-cost periods. Cost related to providing flexibility combined with the profits from optimized process scheduling contribute toward monetization of flexibility as ancillary services for the grid. Balancing between these services and the cost of implementing demand-side response (DSR) solutions creates a seminal pricing strategy for grid flexibility, quantification of which is unprecedented. Lastly, the work is expanded to focus on design of the future European power system with 100% renewable generation and deep electrification of demand. Based on hourly capacity factors of generation, the indispensability of long- and short-term electricity storage is demonstrated to increase renewable penetration (+64pp from current) and avoid massive investments in generation overcapacity. Different combinations of storage technologies and generation shares are explored using a Monte Carlo approach with pseudo-random drawing from a Sobol sequence. The comparison between results obtained for independent countries with isolated grids and a single European interconnected system shows that electrical synergy can significantly decrease energy cost and total greenhouse gas emissions by 18% and 24%, respectively. Moreover, the same comparison introduces a novel approach to estimate the price that countries should expect to pay for security of supply, or their compensation for providing inexpensive renewable energy. Transitioning to renewable-based generation and storage reduces greenhouse gas emissions associated with electricity consumption by 90% and prospects a promising 85% reduction in carbon intensity of the European economy.enreal-timeelectrical griddemand-side responsemodel predictive controlequivalent batterytransmission expansionrenewable energycarbon emissionscarbon taxgrid digitalisationthesis::doctoral thesis