Executive summary

Nuclear batteries (NBs) are an innovative class of nuclear microreactors with a thermal output below 30MWth. They promise to provide energy as a service to customers with minimal staff and site preparation requirements, with rapid deployment – on the order of days to weeks – owing to their compactness (size of one or multiple shipping containers) and autonomous plug-and-play nature. In addition, the possibility for colocation of the NBs with customers and their high targeted reliability (capacity factors upwards of 90%) limits the need for costly transmission and storage systems. This is especially appealing in the context of hydrogen production, as the hydrogen transport and dispensing costs are prohibitively high for the development of a hydrogen economy – in 2017, the hydrogen transport and dispensing costs were 14.4 – 15.6 $/kg [1].

This work aims to identify the requirements NBs must satisfy for decentralized electricity or hydrogen production and their ability to do so. This report outlines the current state of research and identifies key areas of future work.

A set of requirements has been identified in the context of both hydrogen production and the use of NBs in offshore power generation – which is part of the broader scope of this project sponsored by Shell, Equinor and Exxon Mobil – but the set is only treated qualitatively for now. No system requirements have been identified that are unique or more stringent for electrolysis than for grid electricity production. Thus,the limiting requirement for the viability of NB-power electrolysis is its economics, which is why the study has focused mainly on the economics of hydrogen production using NBs.

For the economic analysis, several projects starting in 2030 in California (CA) are considered, see Table 1. The projects either buy electricity from the grid, or produce it using NBs. In the latter case, the more efficient, high-temperature Solid Oxide Electrolysis Cells (SOEC) are considered alongside the mature low temperature Polymer Electrolyte Membrane (PEM) electrolysis. CA is chosen because many hydrogen projects have been developed there already, and hence, data on costs and price projections is readily available. Two types of decentralized hydrogen production are considered, one is a community-scale facility that is close to the demand and the other is on-site production using a single NB – hereafter referred to as centralized and distributed production, respectively. Note that the community-scale facility can be referred to as a semi-centralized facility under the nomenclature or Reddi et al. [2], but is called a centralized facility in this report.

The capacity of the community-scale facilities represents an electrical demand of roughly 60 MWe for PEM electrolysis and 45 MWe for SOEC electrolysis, well within the reasonable range for a multiple-NB project. For the distributed production, a capacity of 1600 kg/d is chosen based on the capacity of the currently-largest hydrogen fueling station in CA [3].

The economic analysis is based on simple levelized cost models to compare the levelized cost of electricity (LCOE) and hydrogen (LCOH2). Reported costs result from Monte Carlo simulations and in addition, sensitivity analyses are performed for each project. Moreover, multiple ways of claiming subsidies under the Inflation Reduction Act (IRA) are considered for each case considered in Table 1. Finally, a doubling of the capacity of the distributed projects to two NBs is also considered to highlight economies of scale in the NB projects.

ABSTRACT

Significant effort is now underway in the United States (US) to develop the licensing basis for advanced reactors, which vary greatly in terms of power rating, physical size, types of fuel and coolant, and civil construction. At the same time, commercial deployment of these reactors need to satisfy capital cost and operations and maintenance targets, for which they can be highly correlated to their licensing basis. On this front, shrinking the reactor building footprint and volume of materials, reducing or eliminating the need for on-site, 24-hour-a-day security personnel, and maximizing the number of possible sites, including urban areas and cities, will be key to the economic viability of smaller MWe advanced reactors. In August 2023, the US Nuclear Regulatory Commission (NRC) ruled on emergency preparedness for small modular reactors and other new nuclear technologies. NRC ruling provides optimism for a performance-based treatment of smaller MWe reactors. However, physical security considerations were not part of the ruling and the guidance in this area remains unclear.

This investigation performs a consequence-based analysis to determine the ability of a microreactor to withstand physical security attacks without the intervention of an on-site security team. The work utilizes a MIT-designed 15MW Sodium-cooled Graphite-moderated Thermal microReactor (SGTR) as the reference reactor. A scope of postulated security threats, organized by category (fire, blast, aircraft crash, sabotage), is considered. A screening process is conducted, using probabilistic and qualitative factors, to identify Design-Basis Threats (DBTs). Damage from each DBT is studied, and mitigation strategies against various threats are identified. Particular effort is devoted to quantifying the robustness of the reactor building’s (RB) to mechanical damage and impacts. Each DBT source term is evaluated through examination of a primary coolant fire and the mechanistic estimation of source term during a Prolonged Loss-of-Heat-Removal (PLHR) scenario. Resulting public dose exposure is assessed using a Gaussian-plume atmospheric dispersion model. Dose exposure goals of 25 and 1 rem (250 and 10 mSv) are expressed in terms of Low Population Zone (LPZ) and Emergency Planning Zone (EPZ) sizes. When relevant, recommendations for offsite security response timing and its impact on accident source term are provided.

From the civil engineering perspective, employing a thick (~2 m) reinforced concrete (RC) radiation shield as a core enclosure is highly effective in preventing mechanical damage to the core from physical security events. Regarding radiological impacts, it is shown that a PLHR event does not pose a threat to public health. However, the analysis highlighted the vulnerability of the SGTR design to high coolant (sodium) activation. Design-Basis aircraft crashes and contact blasts, capable of damaging the primary heat exchanger could be a more limiting accident. Based on this finding, the design of the SGTR was modified such that the the primary exchanger would be moved to the 2m-thick RC

radiation shield enclosure, safeguarding it from the considered DBTs. Overall, since the SGTR was originally designed to have minimal off-site consequences during a PLHR event, this resiliency during a beyond design basis accident translated to high resiliency from a physical security point-of-view while the reactor is in operation. Therefore, we conclude that there is high potential for SGTR to forgo a need for onsite-security guards to protect its reactor building from DBTs under normal operation. This conclusion is design specific, not applicable to other operational modes (e.g., transportation of a microreactor back to a central facility), and subject to uncertainty and scope of the study.