Researchers Paul Seurin and Dean Price, affiliated with the University of California, Berkeley, have expanded their previous work on optimizing heat-pipe microreactors (HPMRs), compact nuclear power systems designed for remote areas. Their latest study, published in the journal Nuclear Engineering and Design, explores multi-objective optimization of these reactors, balancing both economic and operational goals.
In their prior research, Seurin and Price focused solely on minimizing the levelized cost of electricity (LCOE) using a design optimization framework that incorporated techno-economic considerations. This time, they’ve extended that framework to include a second objective: minimizing the rod-integrated peaking factor ($F_{Δh}$), which is a measure of the reactor’s power distribution and safety. To tackle this multi-objective problem, they employed the Pareto Envelope Augmented with Reinforcement Learning (PEARL) algorithm.
The researchers evaluated three different cost scenarios for the reactor’s components: high-cost axial and drum reflectors, low-cost axial reflector, and low-cost axial and drum reflectors. They found that reducing the solid moderator radius, pin pitch, and drum coating angle—while increasing the fuel height—effectively lowers $F_{Δh}$. Across all scenarios, four key strategies consistently emerged for optimizing LCOE: minimizing the axial reflector contribution when costly, reducing control drum reliance, substituting expensive tri-structural isotropic (TRISO) fuel with axial reflector material priced at the level of graphite, and maximizing fuel burnup.
While PEARL shows promise in navigating trade-offs across diverse design scenarios, the researchers noted discrepancies between surrogate model predictions and full-order simulations. They anticipate further improvements through constraint relaxation and surrogate development, which will be the focus of their ongoing investigations.
For the energy sector, this research highlights the potential of advanced optimization techniques like PEARL in designing cost-effective and safe nuclear power systems. HPMRs, with their compact and transportable nature, could provide a viable alternative to costly fossil fuels in remote regions, contributing to a more sustainable and reliable energy mix. The insights gained from this study could guide future designs and deployments of small modular reactors, enhancing their economic competitiveness and operational safety.
This article is based on research available at arXiv.