Global sensitivity analysis of nuclear district heating reactor primary heat exchanger and pressure vessel optimization

Recently, small modular reactors (SMRs) have received greater interest as a source for clean and affordable district heating (DH). Compared to power plants, the low-pressure, low-temperature design and nearly 100% efficiency reduce the cost of produced energy considerably. However, few practical implementations exist yet, and cost estimates and design principles are subject to uncertainties whose interactions remain largely unknown. In this work, we present a techno-economic optimization and sensitivity analysis of a natural circulation DH SMR primary heat exchanger. A Cuckoo Search variant augmented with a modified Hooke-Jeeves search was used as the optimizer, with SimDec (simulation decomposition) subsequently employed for global sensitivity analysis. The reactor pressure vessel and containment vessel specific costs exhibited the greatest impact on the cost of heat and the optimized configurations. While low-pressure, low-temperature design is central to heating reactor cost-effectiveness, optimized primary circuit temperatures clearly exceeded previous assumptions. In a 5260 full-load hours mid-load application, a 34-41 €/MWh cost range was found for produced heat at 8% interest and 20-year lifetime. For heat exchanger optimization, the results indicate the potential for considerable performance improvement from using deterministic local search for terminal convergence and sensitivity analysis for dimensionality reduction. • Small heating reactor appears an affordable source of clean district heating • Pressure vessel specific costs important for cost of heat and optimized design • Novel approach combines global sensitivity analysis and engineering optimization • Dimensionality reduction and local terminal search for performance improvement • SimDec tool reveals interaction of decision variables and uncertainties