Детальная информация

Nuclear-assisted high-temperature steam electrolysis (HTSE) enables low-carbon hydrogen production but optimizing such systems is hindered by high simulation costs and a lack of systematic waste-heat recovery. This thesis addresses both via a framework combining thermodynamic analysis, waste-heat integration, and surrogate-assisted optimization (SAO). A VVER-1000 secondary cycle is integrated with HTSE modules. Three steam return scenarios are compared: mixing with turbine inlet steam (Scenario 1), condenser discharge (Scenario 2, baseline), and return after the last high-pressure preheater (Scenario 3). Scenario 3 achieves the highest cogeneration efficiency (38% at 8 kg/s H2), lowest power loss factor (36.5%), and 26.9% heat cost reduction, yielding a levelized hydrogen cost of 1.74 $/kg. A strong linear relationship between standalone and cogeneration efficiencies is identified. Using Wilks’ 95/95 criterion, 59 HYSYS simulations train a Gaussian Process Regression surrogate (R2=0.999). Four metaheuristic algorithms converge to a global optimum of 35.17% standalone efficiency (back-validated with 0.34% error), reducing LCOH further to 1.70 $/kg. The modular framework provides actionable guidelines for retrofitting reactors, showing that strategic waste-heat recovery and SAO significantly enhance thermodynamic and economic performance.