Gestion multi-agent et optimisation des systèmes énergétiques hybrides pour la cogénération industrielle de chaleur et d'électricité

Industry in the European Union is pressured by technical and political access to resources, increasing fuel and emission prices and more stringent environmental regulations. Many industrial complexes need superheated saturated steam to operate, which comes mostly from fossil fuel combustion. Both construction of new steam generating capacity and pricing of steam require analysis of present and possible future scenarios of financing, policies, resource prices and demand. Current analysis method uses levelized cost of energy, which represents average lifetime costs. The method does not capture well the variability and uncertainty of parameters nor does it provide useful information for short-term decision making. This work presents MATHESY - a tool for more comprehensive technoeconomic analysis which accounts real-world variability, capable of considering technology solutions alternative to fossil-fired boilers fairly. This is achieved through dynamic steam prices which account for transient resource prices and operating parameters of steam consumers and producers. MATHESY uses multi-agent approach to simulate interactions between steam consumers and producers, a proven approach in microgrid management, electricity trading and negotiation simulations. Discovery of steam prices uses two Dutch-style auctions and long-term contracts executed in 10-minute intervals. Agents use price-predicting models to negotiate. Steam producers track profits from trades and calculate costs using physics models of boilers and a High-Temperature Gas-cooled Reactor (HTGR).Results of scenarios identical to ones in levelized cost analysis done for the European Union project GEMINI+ are within 5%, which is the selected price increment between auction stages. Scenarios with variable and historical prices and demand curves show the system finds optimal results and demonstrates a good market opportunity for HTGRs at current high fossil fuel and emission prices. Utilization of long-term contracts decreased prices on average and qualitatively improved results of HTGRs. The cost of sub-optimal technology choices and market timing is visualized with our approach in the sensitivity analysis of external factors, variable demand and installed power. In scenarios with sufficient steam generating capacity, between 99,92% and 100% of steam demand was met.