Dr. Edwin Lyman, Senior Scientist, Global Security Program, Union of Concerned Scientists, has published a paper titled Small Isn’t Always Beautiful: Safety, Security, and Cost Concerns about Small Modular Reactors that poses many questions about the development of SMRs. I’ve seen Dr. Lyman taking notes and asking questions at a number of SMR related conferences; this paper is apparently one of the products of that effort.
Since I have been advocating for smaller nuclear power plants since 1991 and helping to develop one particular brand of SMR since 2010, I read his report with a great deal of interest.
I will address specific concerns and disagreements with his analysis — along with pointing out the subjects on which we agree — but first I want to state my general impression. This paper seems to be one more example that supports my long standing assumption that the UCS is fundamentally opposed to the use of nuclear energy. Despite its frequent protestations that it is not antinuclear, the UCS is predictably quick with an automatic negative reaction to any attempt by nuclear energy technologists to improve its viability in the competitive energy market.
The next time I see Dr. Lyman I should ask him directly what kind of power systems he would like to see us building today and in the near future. Based on his dislike of both small and large nuclear power plants, I assume that he and his employer favor coal and natural gas. (Wind and solar are incapable of the task of supplying the kind of reliable, industrial strength power that most members of our society expect to be available 8760 hours per year.)
Dr. Lyman is partially correct in asserting that there are challenges associated with producing smaller reactors that can be operated profitably without reducing safety margins, but he is giving the wrong impression by implying that the task requires cutting corners or that the obstacles are so great that the effort is not worth government investment and encouragement. Let me explain the basis for my dissenting professional opinion.
In addition to my first-hand experience in operating and maintaining submarine nuclear propulsion plants, which are far smaller than commercial power stations, one of the main reasons that I became interested in designing and building smaller commercial reactors was the experience of reading I. C. Bupp’s Light Water: How the Nuclear Dream Dissolved. That book, sharply critical of the nuclear industry of the 1970s and 1980s, pointed out how the quest for “economy of scale” had driven reactor designers to create systems that were several times larger than any in operation at the time that they did the design work.
Bupp showed how the rapid scaling of designs from Gen I to Gen II depended on unproven assumptions and models rather than full scale testing of the structures, systems and components. His work, and the early work of the UCS, led to the expensive and time consuming loss-of fluid test (LOFT) large break experiments that eventually proved that the emergency core cooling systems could work to keep cores from being damaged and releasing harmful radiation — with some significant design changes. Here is a key quote from NUREG/IA-0028, which is titled Review of LOFT Large Break Experiments.
The principal finding from the large break experiments is that, for the degrees of severity in initial and boundary conditions, the measured fuel cladding temperatures remained well below the peak cladding licensing limit temperatures.
Even though I knew that the experimental results eventually validated the codes and showed that safe large reactors could be built, I recognized that there were a number of diseconomies of scale that combined to reduce the assumed economic benefits of ever larger nuclear power plants. I also realized that the focus on very large power plants pushed nuclear fission energy into a tiny niche of the overall energy market; it is never good for any product to be dependent on a single type of customer. I began my research into smaller reactors with the idea that it was possible to change the prevailing paradigm that “bigger is cheaper”.
My research on the topic of scale economies has spanned the past 22 years. It has taken some unusual paths that included a three-year stint as the general manager for a small manufacturing company. The company that I managed, J&M Industries, Inc., produced a wide variety of products with a large range of quality requirements and production run volumes. We provided custom product development services that helped inventors produce low volume prototypes of new ideas. We manufactured packaged products for the consumer market, bulk products for the medical market, and high quality marine components for other companies whose products included large luxury yachts. That experience gave me direct understanding of the economies associated with requirements, material variations, order timing, inventory, delivery, and various sizes of production runs.
Our company was in a cost-competitive market; the only way to make money was to have a complete understanding of the many components of product cost so that prices could be set at a competitive, but profitable level. The owner of the factory had learned that lesson the hard way. He had developed a couple of simple rules of thumb; I built product costing models based on his experience and refined them over the years we worked together. One of my most important take aways from that job was that the cost of anything includes far more variables than most people want to think about.
There is little academic doubt about the scaling equations associated with electricity production machinery; if you assume that smaller systems are just scaled down versions of larger systems, the cost associated with the larger system does not scale linearly with output. The key to producing economical smaller systems is to take advantage of opportunities to simplify the system design and to eliminate systems and components that are no longer necessary because they were initially added to overcome a scale diseconomy. Of course, it is also possible to break the scaling laws by a complete redesign that puts the technology on a different cost curve altogether.
If, for example, a reactor designer chooses to create an integral design in which the reactor core, steam generator, and pressurizer are directly connected in a single vessel, that designer can eliminate the cost of connecting pipes, isolation valves, controls and indications for the isolation valves, supports for the piping and valves, routine inspections of the piping, and the delays associated with fitting and welding large piping on a construction site.
Another thing to realize about scale economies is that they are not only applicable to nuclear energy, they are applicable to all other energy system competitors. That means that smaller nuclear plants do not compete in cost against the very largest nuclear or coal fired power plants; they must compete in cost against the other options for providing reliable power in 100-500 MWe chunks.
Dr. Lyman seems quite adamant in his assertion that smaller reactors should get no credit for their enhanced safety and smaller cores and that they should be forced to continue planning for a 10 mile emergency planning zone (EPZ). He states that any reactor greater than 250 MWth will contain enough radioactive material to be able to produce a release that would require evacuation up to and beyond 10 miles.
However, that assertion requires the physically incredible assumption that the entire core is somehow vaporized and distributed into the atmosphere. Based on the results of the State of the Art Reactor Consequence Analysis, I believe it is time to reevaluate the need for a 10 mile EPZ for our existing fleet of reactors. Though the report’s executive summary contains a lot of words apparently designed to obscure the key result, here is the important paragraph:
The unmitigated versions of the scenarios analyzed in SOARCA have lower risk of early fatalities than calculated in the 1982 Siting Study SST1 case. SOARCA’s analyses show essentially zero risk of early fatalities. Early fatality risk was calculated to be ~ 10-14 for the unmitigated Surry ISLOCA (for the area within 1 mile of Surry’s exclusion area boundary) and zero for all other SOARCA scenarios.
There is certainly no public health reason for asserting that the already obsolete EPZ requirements should be applied to newer, safer, and smaller power plants.
Of course, applying the obsolete EPZ requirement to all reactors, regardless of size and design features, would support the apparent UCS mission of forcing nuclear energy to be uncompetitive with its preferred sources of power. It would limit the ability of smaller reactors to be considered as emission-free replacements for older, smaller coal stations that are often located within the boundaries of smaller cities in the United States and abroad. It would also limit the ability to capture more value and thermal efficiency from a nuclear power plant by locating it near an industrial facility or campus/district heating system that could make use of the waste heat that is an inevitable part of a Rankine cycle steam plant. Lyman and UCS apparently do not want any nuclear cogeneration facilities, despite their potential advantages for the climate.
I also disagree with Dr. Lyman’s assertion that the security force requirements should be the same for a modern SMR as they are for large light water reactors that were designed and built more than 30 years ago. Facility vulnerabilities can be mitigated in creative ways when starting with a clean sheet of paper compared to retrofitting security onto an existing site that was not fundamentally designed to resist attack. I have had the pleasure of working with one of the leading site security experts in the country and have seen far more of the details of his security plans than Dr. Lyman. I am confident that he and his team will be able to convince the most skeptical regulators that his design will adequately protect the plant.
Unfortunately, site security is one of those areas where there must be a point at which the public agrees to turn over its information rights to trusted agents — like government regulators. It is not possible to allow the public to have access to all of the details since the bad guys are part of “the public.”
Dr. Lyman’s discussion of the importance of security costs (pages 14 and 15 of his report) contains a glaring analytical error. He makes the following statement:
The nuclear industry’s preoccupation with reducing security staffing is somewhat surprising. Even though security labor costs are significant, they are far from being a dominant contributor to overall O&M costs. Security staffing costs range from 15 to 25 percent of total O&M costs.
He follows that statement up by showing that an armed guard force of 120 people is a fairly small portion of the O&M cost for a large nuclear plant.
In total, considering the number of shifts per week, a typical reactor site would need approximately 120 security officers. For comparison, typical total plant staffing is between 400 and 700 personnel per site, so the security force is roughly 20 to 30 percent of the total workforce.
That statement ignores the fact that holding the security staff at a constant number while reducing all other staffing through system simplification and improved control systems makes the portion of O&M devoted to security increase. SMR designers are designing their systems for staffs that are considerably smaller than the 280-580 non security personnel per site that Dr. Lyman assumes is typical.
I agree with some Dr. Lyman’s points. Without redesign, smaller systems have a cost disadvantage over larger systems. Going smaller is not a magic bullet that will suddenly make nuclear energy competitive. The investment required to build a manufacturing infrastructure that enables series production techniques to overcome scale disadvantages might be a real hurdle that prevents SMRs from ever becoming competitive.
Aside: There is a solution to that challenge in the United States that should be more fully explored. Contrary to popular assumption, the US has been building smaller nuclear power plants with some regularity for the past 60 years. The infrastructure to manufacture and assemble complete power plants exists, but much of it is off limits to commercial endeavors. The assumptions that cause that to be true should be open to questioning in today’s political environment. End Aside.
Dr. Lyman expresses well-justified skepticism about the benefits of building nuclear plants underground. It seems to me that the choice of burying power stations comes with at least as many additional risks as the ones it eliminates. The key cited advantage of going underground is reducing the plant’s vulnerability to aircraft impact. I personally think that the Greg Jaczko-initiated Aircraft Impact rule should be discarded as being an unnecessary barrier to building new nuclear power stations. Almost every other component of our national infrastructure is more vulnerable to attack by jet airplanes than an above ground nuclear power station.
Dr. Lyman correctly points out that going underground raises questions about access in emergencies and resistance to flooding. Building plants below grade also adds enough site specific design and construction requirements to negate most, if not all, of the projected advantage of manufacturing the nuclear portion of the power plant in a factory. It is a construction truism that the deeper the foundation, the more difficult the site assessment and the more expensive the construction.
The height of the tall units that all of the integral light water reactor designers are proposing is a fundamental part of their natural circulation, passive safety design, but putting them almost completely underground requires installing foundations that are deeper than the foundations for the world’s tallest buildings. Building a power plant starting at the bottom of a very deep hole adds many hours to the construction process and requires the use of some of the world’s largest cranes, especially if the power plant is to be built of large, heavy modules.
In my opinion, encouraging SMR designers to choose to bury their plants underground may be part of a strategy to add enough cost and make the construction timeline long enough to prevent smaller reactors from competing with other, more immediately lucrative power plants that burn the hydrocarbon products of some of the world’s most politically powerful and wealthy corporations.
The bottom line is that, although it is not automatically true, small can be beautiful in nuclear power plants. Smaller plants can be built with greater predictability, improved safety, increased reliability, and can serve a larger number of potential customers. The Department of Energy is on the right track in its program to support small reactor design and licensing, though the 5-year, $452 million (total) program is incredibly tiny in comparison to the technology’s potential and in comparison to the investments DOE is making in technologies like technically unproven carbon capture and sequestration, large scale solar thermal power stations, and wind turbine deployment.