Soutenance de thèse de Hemant SHARMA (CPP) - CEA vendredi 26 mars 2021 à 14h00 en visio

Compte-tenu des circonstances sanitaires, veuillez trouver le lien pour y assister à distance : Lien vers le site
ID de réunion : 944 5032 5323
Code secret : Soutenance

En cas de problème, l’audio sera également accessible par téléphone au +331 7037 2246,
ID de réunion : 944 5032 5323
Code secret : 7660970061

Les membres du jury :
 

• Monsieur Frédéric WURTZ, Directeur de Recherche au CNRS, Présiden
• Madame Catherine AZZARO-PANTEL, Professeure, INP-ENSIACET, Rapporteur
• Monsieur Nicolas PERRY, Professeur I2M, ENSAM, Rapporteur
• Madame Peggy ZWOLINSKI Professeure, Grenoble INP, Directrice de thèse
• Monsieur Guillaume MANDIL, Maître de Conférences, UGA, Co-Encadrant de thèse
• Madame Élise MONNIER, Ingénieure, CEA-Liten, Co-Encadrante de thèse
• Madame Emmanuelle COR, Ingénieure de Recherche, CEA-Liten, Co-Encadrante de thèse
   

Summary :

Emerging energy technologies are often assessed at a pilot scale, which allows the investigation of their performance in real world conditions. These are often called hybrid energy systems and are characterised by variable energy supply/demand, multiple energy vectors, etc. Existing software to analyse these systems only allow simplistic environmental analysis which might be insufficient to assess the new energy systems. This thesis thus aims to provide a general methodology to include environmental indicators in the design, planning and assessment hybrid energy systems. First, to facilitate integrated techno-economic-environmental modelling, a general framework to develop parameterised life cycle assessment (LCA) models of hybrid energy systems is proposed. This framework is applied to a case study involving hydrogen production from PEM water electrolysis. The results show that only improving influential aspects of the model might be sufficient to obtain accurate results. Then, the LCA indicators along with the techno-economic ones are included with a genetic algorithm, NSGA-III to compute the pareto front of the multi objective problem. Here again, a dimensioning case study involving hydrogen production is used for demonstration of the concept. Since several studies in the literature have shown that the LCA indicators are correlated, performance of the algorithm is tested with reduced set of objectives while avoiding impact transfer. For reduction of objectives, a new technique using a visual survey is proposed which can be applied to simple cases. Other objective reduction techniques like principal component analysis (PCA) can also be used for this purpose. The results show that using reduced set of objectives make NSGA-III more efficient in the search of optimal solutions. Similar quality solutions can be obtained with less number of evaluations. Furthermore, results show that inclusion of LCA indicators in dimensioning enables search of solutions with lower environmental impact. Not only does it showcase the trade-offs between environmental indicators and the conventional techno-economic indicators, it also uncovers the trade-offs between the LCA indicators themselves. This work will thus facilitate the development of energy systems with higher environmental efficiency.