It’s not clear whether the Department of Defense should continue to develop new types of microreactors — primarily because they have uncertain costs and regulations. However, we can use a systematic and repeatable framework for plugging in such information as it becomes available to pave the way for effective DoD investment decisions. We can even use presently available data and information to identify preliminary findings.
For example, the smallest microreactors, which have a power capacity of 1-5 megawatt electric, are too expensive for almost all DoD locations in the United States when comparing average electricity costs to the status quo energy costs — even when a cost of carbon, value of energy resiliency and manufacturing efficiencies are included.
On the other end of the spectrum, larger, advanced nuclear reactors (80-300 MWe) have a lower average cost of electricity but produce too much power for almost any DoD location to consume independently.
In between these two extremes, a medium-sized microreactor (17 MWe) could be a worthwhile investment for up to 21 different DoD locations in the United States.
Regardless of their size, microreactors have significant capital costs. The DoD recently announced a $300 million contract to design, construct and test a microreactor prototype. This is between four and 20 times greater than construction cost estimates from a Nuclear Energy Institute report published in 2019. Although the contract includes more than just construction, the significant cost increase aligns with themes from research out of the Massachusetts Institute of Technology — that nuclear power plant construction cost estimates tend to increase as the microreactor matures from a concept to a prototype.
Rather than just looking at construction costs, a more effective way to compare different electricity-generation sources is their levelized cost of electricity, which is the average cost of electricity over the entire life span. For microreactors, this is largely dependent on how much power can be produced in an instant — the power capacity, or MWe — the percentage of full power capacity delivered over time — the capacity factor as a percentage — and how manufacturing efficiencies decrease costs for each subsequent unit.
Microreactors are more cost-effective if they have a large power capacity (MWe), if they are used consistently for their entire life span (a capacity factor near 100%) and if numerous units are implemented.
The preliminary findings above focus on siting microreactors at DoD locations in the United States. This allows them to have a larger power capacity (MWe) than portable or mobile reactors. The analysis also required DoD locations to have an average power demand greater than the microreactor’s power capacity (MWe) to ensure the microreactor is used consistently for its life span. This increases the capacity factor closer to 100% and keeps the average cost of electricity as low as possible.
Because most DoD locations have relatively low average power demand, there is a trade-off between microreactors with a larger power capacity (MWe) and the number of DoD locations that can use them with a capacity factor near 100%. This partially explains why a medium-sized, 17-MWe microreactor could be a worthwhile investment for the DoD.
In order for a DoD investment in a 17-MWe microreactor to be worthwhile, the DoD must account for a cost of carbon and a value of energy resilience; the microreactor cost estimates cannot increase; and all 21 locations must be viable hosts — all so that manufacturing efficiencies decrease the microreactor costs with each unit.
Therefore, DoD decision-makers may want to first invest in more accurate cost estimates and perform a more targeted investigation of DoD locations to determine if the investment will be worthwhile.
Today, the DoD is developing a stationary and transportable microreactor, but this is not the first time the DoD invested in this technology. The Army Nuclear Power Program, or ANPP, designed, constructed and operated microreactors from 1954-1976 — when nuclear energy was in its infancy. ANPP demonstrated eight microreactors with different designs; three were stationary, three were portable (for transport and reassembly), and two were mobile (on a truck/barge). The program was ultimately discontinued because none of the reactors provided a unique operational capability — each could have been replaced with an alternative energy source — and because they were likely more expensive than those alternative energy sources.
An archival analysis suggests that the energy produced by ANPP microreactors was three to five times more expensive than if diesel generators produced the same amount of energy with a fully burdened cost of diesel fuel, which accounts for costs on top of the fuel itself, like transportation, protection, personnel, etc. These historical examples highlight how important costs are when considering an investment in a substitutable technology.
Ultimately, the decision to invest in a microreactor is complex; numerous assumptions and uncertainties could interact with each other to determine if the investment is worthwhile. Therefore, DoD decision-makers should use a systematic and repeatable decision framework to help manage these complexities; one possible framework can be found in my doctoral dissertation. The framework can equip DoD decision-makers to effectively communicate with each other about the decision to invest in a microreactor and can be reapplied in the future, as microreactor capabilities and DoD location requirements inevitably change.
U.S. Air Force 1st Lt. Kyle D. Haak has a doctorate in policy analysis from Pardee Rand Graduate School. The views expressed in this commentary do not necessarily represent the views of the think tank Rand, the U.S. Air Force, the U.S. Defense Department nor the U.S. government. For more details, watch a presentation online and read the dissertation on this topic.
