FREE

Small Modular Reactors Part I

Small nuclear reactors were the building blocks of the early nuclear industry. Small modular nuclear reactors (SMRs) have been a mainstay of military propulsion for over 50 years. Their entrance in the civilian market is, however, a recent phenomenon. Specific conditions such as investor interest and ample pre-existing research facilitated the appearance, throughout the past half-decade, of several such commercially-oriented reactor designs, from both established industry names and fresh start-ups. This paper attempts to show the SMRs particular advantages compared to large nuclear reactors, along with several challenges to their deployment, particularly from the investors side.

The early push for nuclear power

For over half a decade, the prevailing doctrine in nuclear power plant (NPP) construction has been “big is beautiful”. In an evolutionary process, nuclear reactors and their auxiliary systems (the so-called “balance of plant”) have been steadily increasing in size: from the few dozen megawatts-electric (MWe) of early experimental reactors of the 1950s, through the several hundred MWe of Generation II designs (which compose the majority of existent generating capacity at global level) to the 1,000+ MWe of Generation III/III+ current large reactor (LR) designs. This continuous ‘supersizing’ was postulated as necessary due economies of scale, the idea that a larger reactor would be cheaper to build, on a per-megawatt basis, than a smaller one based on the same design. For a period of time, this concept seemed to be validated, as plants were brought online on schedule. However, as the initial enthusiasm started to disperse, problems started surfacing. Most notably, initial industry estimates regarding construction costs were proven to be extremely optimistic^[1] with final figures being roughly twice the originally quoted amounts for the 75 plants being built in the United States between 1966 and 1986. Alongside cost, construction delays (not just in the U.S., but also the rest of the world) pushed the average completion time from 60 months for plants built between 1965 and 1970 to 80 months in the case of those built from 1977 to 1982^[2], reaching a record 116 months for the 1995 – 2000 interval.

Undoubtedly, an important factor in ballooning costs and construction time overrun was the impact of the Three Mile Island incident (March 1979). As a side comment, Three Mile Island is positively the only ‘unavoidable’ nuclear incident of the three.^[3] The lessons drawn from its investigation served as starting points for the inter-related concepts of information hierarchy, operator workload and human-computer interface, and also introduced the notion of “inherent (later ‘passive’) safety”.^[4]

The stricter (and “supercharged”, as Kessides calls it^[5]) regulatory environment brought with it increased oversight on nuclear activities, causing costs to go up (as structural reinforcements required more materials) and permits to be issued over lengthier intervals (due to impact analyses that required more time).

The March 1979 incident represented, according to Daniel Ingersoll, the end of the “first nuclear era”^[6], one which had begun in 1957 with the successful criticality of the first commercial reactor design. In the aftermath of Three Mile Island, the civilian nuclear market in the United States cooled down quickly, with no new nuclear reactors being approved for construction until 2012, when the Nuclear Regulatory Commission issued a combined license for units 3 and 4 of the Vogtle NPP in the state of Georgia, U.S.A.

In fact, the slow down in the nuclear market had started before, and Three Mile Island just happened to be the “last straw”. The economic upheaval caused by the Vietnam War expenditure and President Nixon’s 1971 decision to unilaterally repudiate the Bretton Woods Agreement^[7] caused financing problems for utilities and forced many of them to raise prices for the first time in a quarter-century.^[8] The conversion of the old Atomic Energy Commission created by President Truman in 1946 into the more ‘sceptical’ Nuclear Regulatory Commission in 1974 further decreased nuclear enthusiasm. Of the 197 plants on order in 1974, fewer than half eventually proceeded to completion.^[9]

Reactor construction in the rest of the Western world generally followed similar patterns of slowdown/cessation, with the Chernobyl incident causing some European countries to curtail or even outright abandon (Italy, for instance) their nuclear power programmes.

Despite the near collapse of the new-build sector post-1980, nuclear research continued at a steady pace, albeit on a different tangent. The first sign of this new direction was a 1982 study conducted by the Electric Power Research Institute (EPRI) that interviewed personnel from 11 American utilities that owned NPPs, from all branches of power production (management, operations, maintenance). The three most important conclusions gleaned from the survey were that: “(1) 1200–1300 MWe plants were felt to be too large, (2) the plants respond too rapidly to transients, and (3) nuclear plants need to be less sensitive to events in the secondary systems.”^[10]

According to Rodney Adams, nuclear industry commentator and former U.S. Navy submarine engineer, the issues affecting large plants would have likely been discovered earlier had nuclear energy research and power plant design/deployment evolved organically. Instead, it was forced to undergo a ‘Sovietisation’^[11] process by a small “task force” of government officials, contractors and government-held utility representatives. Ingersoll concurs with this view, noting that, in 15 years’ time, an order-of-magnitude increase in size and power rating took place without a proper internalisation of lessons from long-term operational trials of the first wave of power plants.^[12]

The results of the EPRI survey were further expanded upon in 1985 by Alvin Weinberg and his team at the Institute for Energy Analysis.^[13] As corollaries they noted that:

• large light-water reactors pose low safety risks to the public but high financial risk to investors;
• large reactors are difficult to operate, since they are complex and delicate;
• small, inherently safe designs are possible if they can be deployed in an economical fashion.^[14]

It was in this study that the founding principles of the small, ‘inherently’ safe reactor were first enunciated. Nevertheless, these findings were not expanded upon at the time. The general apathy towards nuclear power grew after Chernobyl, with the 1990s as a low point in terms of overall disinterest. Attempts to revive new nuclear build in the West were undertaken in the early-mid 2000s, but with mixed results due to longstanding issues with respect to established large reactor-based nuclear plant design and construction paradigm.

Large reactors and the pitfalls of “economies of scale”

The most common reactor design in use today is the regular (‘light’) water cooled reactor. Its development and widespread expansion as main reactor type wasn’t an accident of history. Rather, it is a continuous testament to the “original sin” of nuclear power, namely its roots in the military programme. The Westinghouse-built S1W water-cooled reactor (installed on USS Nautilus) was selected by Admiral Hyman Rickover over its erstwhile competitor (the General Electric-sourced S2G sodium cooled reactor, installed on USS Seawolf^[15]) for standardisation as a submarine power plant. Later on, similar (enlarged) designs were to serve as power sources for the United States’ aircraft carrier fleet. In his choice of a water-based system, Admiral Rickover leveraged 50 years’ worth of U.S. Navy experience in the field of high-pressure steam engineering and the considerable industrial base remaining from the wartime gear-up. Furthermore, their specific fuel constraints were deemed irrelevant, uranium enrichment for propulsion being but a minor sideshow to the vast nuclear weapons programme of the era.^[16] The designs that were to power the Navy’s submarines (and other ships) could be considered (at least partially) the very first implementation of the Small Modular concept. By necessity, they were compact (space inside a submarine is at a premium). Given the need for a standardised engineer training process, they were single-type units, assembled en-bloc in a factory, and they were used to produce electricity and fresh water^[17], the two main applications envisioned today for SMRs.

However, as explained in a June 2014 Breakthrough Institute brief on nuclear costs, transferring the light-water reactor blueprint directly into the civilian engineering sector as its mainstay was, in retrospect, “less than optimal”.^[18] Light-water reactors use water both as a coolant (keeping core temperature within safe parameters) and as a moderator (slowing down and controlling the fission reaction inside the fuel). As reactors operate well over the boiling point of water, it is necessary that water be kept liquid in order to maintain moderation. This entails the use of specialised systems that maintain high pressure inside the primary coolant loop and the reactor vessel itself. Due to failsafe requirements, secondary systems that inject additional coolant in case of an emergency are also needed.

The greatest and concurrently most common danger for a water-based reactor is a Loss Of Coolant Accident (LOCA). In the event of a pressure loss, the coolant starts boiling off, potentially causing the fuel assemblies to become exposed. Their Zircaloy cladding reacts with water vapour creating free hydrogen, which then escapes the pressure vessel (possibly via the same path as the lost coolant) and reacts explosively with external oxygen. In the particular case of Boiling Water Reactors (BWRs), such as the ones found at Fukushima Dai-ichi, which are designed to operate at lower pressures with the provision for a steam void at the top of the reactor vessel, a similar situation is encountered if the water removed from the reactor vessel isn’t replaced via pumping, causing a similar drop below the safety limit and exposing the fuel bundles.

As noted above, large reactors rely overwhelmingly on active cooling via pumps to maintain safe coolant levels. This is in many cases a result of the design environment of the time, when engineers relied on slide rules and Curta “pepper mills”^[19] to create the blueprints for the most commonly encountered generations of nuclear reactors (II/III). The solutions chosen for heat management, coolant flow, radionuclide inventory control, plant management etc., while cutting edge at the moment of their implementation in the early 1960s, were nevertheless conceived during an era when computer simulation didn’t exist even at theoretical level. And, in any case, providing for more pumping capacity would have been simpler and quicker than having to run sophisticated computer simulations^[20] in still rudimentary computing labs (particularly if the reactor being simulated was a 1,200+ MWe Gen. III+ and the timeframe was the early 1990s, when computing power, while respectable, was still rather far from capable of full-plant virtualisation). In addition, in the case of large reactors, with their considerable volume-to-surface ratio (high thermal inertia) and their ample radionuclide inventory (high amounts of decay heat) attempts at using natural convective circulation^[21] as a means of passive safety are technically unfeasible. The one experiment where an attempt was made (in 1986) to run a full-size NPP on passive cooling ended in the worst civilian nuclear disaster in history. ^[22]

As plant sizes increase, so do the number and dimensions of their components. This often leads to the necessary rerouting of pipes, rearrangement of valves, pumps and other balance-of-plant elements owing to geographical limitation of the physical site. However, due to stringent regulations enacted after Three Mile Island, any modification to an existing design was analysed in aggregate, as an entirely new reactor, which lead to considerable amounts of time being spent on justificatory paperwork. This regulatory reset, coupled with repeated overhauls (increased restrictions) of the codes themselves, has, in practice, served to eliminate what was initially supposed to be an important mechanism for decreasing power plant construction cost, namely “learning-by-doing” (or “learning economies”). In other cases (such as the French nuclear industry), politically motivated decisions, for instance preferring national parts suppliers over international ones, regardless of price along with continuous plant upscaling (from an initial 900 MWe to 1,500 MWe) served to eliminate any efficiency gains generated by industrial learning.^[23] According to Davis, the overnight cost^[24] (expressed in dollars per kilowatt of nameplate capacity) increased from $1,000 in the 1970s to $2,300 in the 1990s.^[25] This example shows that costs can escalate even in the presence of reasonable regulatory stability and basic layout standardisation.^[26]

Another reason for cost escalation is the reduced pool of potential component suppliers. Nowhere is this more evident than in the specific case of the largest single part of a nuclear power plant, namely the reactor pressure vessel (RPV). Safety-wise, the best way to forge such a vessel is from a single large ingot, since assembling one from several smaller casts via welding can lead to seam fatigue and potential leaks (necessitating expensive and time-consuming repairs) later on in the plant’s operational life (welded pressure vessels require additional continuous inspections, increasing operating costs). For several decades, the only foundry able to accept the 600-tonne ingots used in single-piece forging was Japan Steel Works located in Muroran, Hokkaido. The demand for reactor vessels, primarily from China, has led to an expansion in heavy forging capacity in China (China First Heavy Industries, Shanghai Electric), South Korea (Doosan Heavy Industries), Russia (OMZ Izhora) and France (Le Creusot) with further plans in India (Larsen & Toubro) and the U.K. (Sheffield Forgemasters).^[27] However, the RPV isn’t the only power plant component requiring adequate heavy industrial capacity. Steam generators and turbines, condensers, electrical generators, pressure valves and other components need to be suitably upscaled to fit the requirements of gigawatt-range NPPs.

Delays are another common yet critical economic weakness in NPP construction. The reason for this, according to Locatelli et al. is twofold:

1. Fixed daily cost: a large-reactor nuclear construction site retains considerable numbers (in the thousands) of people and expensive equipment (high-capacity cranes). Each working day has a (high) minimum cost of employ.
2. Interest on upfront loans: “Each year of construction postponement (or delay) of inbound cash flow for the utility increases the interest to be paid on the debt. Consequently, for each year of delay the revenue has to be considered as lost.”^[28]

Such delays have an amplifying effect on final cost. According to an International Energy Agency 2006 report, a construction delay of 24 months will increase the levellised cost of electricity (LCOE)^[29] for a KWh produced from nuclear by 9.6%, as opposed to 2.6% for a gas combined-cycle plant and 6.6% for coal.^[30] Since the cost profile of a large nuclear power plant is considerably front-loaded (65% of it is capital deriving mostly from construction expenses^[31]) it needs to run at as high a capacity factor^[32] as possible in order to amortise its investment and start turning profit. Any issue that causes decreases in capacity factor, such as load-following operation or outages (planned or unplanned) will push the amortisation point further into the future, increasing the LCOE even more. If the utility which contracted the plant (and served as one of its financiers) has to sell electricity at a price below the plant’s LCOE, it will be unable to recoup its original investment. Consequently, ratings agencies have started examining any companies engaging in NPP construction in a more critical light. For instance, analysts at Moody’s Investors Service do not consider the construction of NPPs as a “bet-the-farm” endeavour any more.^[33] Therefore, in a recent official report the company has stated that it considers taking a more negative view of the effect NPP investment has on a participating utility’s financial outlook.^[34]

The recent experience with new units under construction in Finland and France showcases yet another, more subtle reason for cost overruns and delays: deprofessionalisation. Unit 3 at Olkiluoto and unit 3 at Flamanville NPPs are First-Of-A-Kind (FOAK) level projects^[35] involving a Generation III+ large reactor design, the EPR (European Pressurised Reactor). In many respects, this was an ambitious undertaking, given that it was not only the first European ground-up build in nearly two decades, but it was also a very large (1,600 MWe) and completely new design. Unfortunately, the over 20 years that passed between projects of such amplitude have had a harmful effect on the available pool of European experts and the specific aptitudes of heavy construction companies. In retrospect, given the odds against the projects, the unfolding events were unsurprising. From the very beginning, construction was repeatedly marred by delays due to quality control: improperly assembled rebar necessitating complete replacement, porous concrete casts having to be jackhammered out and re-poured. These were just some of the many problems that caused costs to balloon over 100% of initial estimates^[36] and start-up dates be pushed back from the originally envisioned 4 years to over a decade. The inexperience exhibited in the initial construction measures was followed by more inexperience in cost-cutting attempts, creating a snowball effect of continuously accumulating unintended consequences.^[37]

The case for Deliberately Small Reactors^[38]

As noted by Ingersoll and others, the International Atomic Energy Agency (IAEA) classifies all reactors below 300 MWe as ‘small’ and all those between 300 and 700 MWe as ‘medium’. This is, however, somewhat of a ‘grab-bag’ arrangement, since it does not take into account individual design particularities. Most of the small reactors currently in operation are merely miniaturised versions of large ones. In fact, many of them are, on average, older than larger ones, having been built as technological demonstrators and First-Of-A-Kind units at the beginning of the supersizing drive mandated by engineering ‘orthodoxy’.

Designers of the new crop of small reactors have undertaken a conscious effort to eschew the train of technological decisions that led to ever-increasing plant size, complexity and ultimately cost. The reasons for doing so are the same as those described in the EPRI study and Weinberg’s follow-up of over two decades prior: compactness and technical simplicity enabling higher safety, more control, lower upfront expenditure and wider distribution.

Since no Small Modular Reactors are currently in commercial operation, researchers have opted to perform analytical extrapolations starting from acknowledged Large Reactor weaknesses. According to their findings, there are notable areas where SMRs hold definite advantages which could allow for their market deployment, assuming a number of preconditions are met. The main areas where SMRs have been found to present tangible benefits compared to LRs are:

• Siting and grid appropriateness;
• Financing and construction;
• Operational safety.

Siting and grid-appropriateness

Compared to large reactors, the compact nature of SMRs and their decreased prerequisites for heat removal allow them to be emplaced in areas where available water sources would provide insufficient flow for a large-scale NPP.^[39] Furthermore, some designs (such as Holtec’s SMR-160, or many metal-cooled models) allow for air-cooling, either by design or as an option, potentially eliminating completely the requirement of siting a plant near a body of water. Another siting advantage is related to the seismic resilience of small plants. Owing to the reduced part inventory and the small size of individual components, SMRs exhibit fewer points of potential failure compared to large reactors. In addition, their dimensions allow them to be mounted on regular-sized seismic shock absorbers^[40] (instead of the highly customised, therefore expensive ones used for gigawatt-range NPPs) and some designs (light-water-cooled) are submerged in a large pool of water which serves as an additional buffer against earthquake energy. According to a 2012 UK report, these siting benefits could increase the percentage of land availability for SMRs compared to LRs from 13% to 24%.^[41]

Small reactors are better suited to small or vulnerable electric grids compared to large ones. A simple electrical engineering axiom is that no single plant should compose more than 10% of a grid’s generating capacity lest its stability be endangered.^[42] This is particularly important in the case of small or developing countries whose grids are unable to safely accept 1,000+ megawatt LRs.^[43] Moreover, some grids favour localised power plants instead of large concentrated generating capacity, either due to geographical features (considerable distances between population centres) or demographic ones (many scattered remote areas).^[44] In such locations, constructing (or upgrading) high-voltage transmission infrastructure could prove uneconomical given the distances involved or the number of final customers.

Financing and construction

Financing is what proponents of large power plants and supporters of small modular reactors disagree the most. According to classical theory, economies of scale would invariably make large reactors more financially efficient in terms of dollars per kilowatt-electric. This was the basis for power capacity increases in large reactor designs throughout the 1960s and 1970s. However, experience has shown that, in many situations (at least in the case of power plants built in the West), the increase in power corresponded to a disproportionate increase in costs and construction time. The complexity and added expense of safety systems necessary to prevent accidents also increased alongside plant size.^[45] There is a possibility that large reactor proponents underestimate or use rather optimistic figures for the real overnight costs of gigawatt-plus-range NPPs. Similarly, the estimates for SMR construction could be erroneously inflated due to the mistaken belief of large reactor proponents that they are comparing “like for like”.^[46] In reality, SMRs are different in several fundamental respects from LRs such as: modular design, integrated systems (reactor pressure vessel, primary loop, pressurizer and steam generator in a single unit), passive safety (natural coolant circulation), decreased parts inventory and serialized factory construction. Their ability to be mass-produced in a factory creates a different kind of economy of scale, that of mass production. Being able to assemble reactors in a dedicated facility (as opposed to in situ construction) improves quality control^[47] and the smaller unit size makes inspections easier to conduct. For instance, should welding defects be detected in a 500-tonne large RPV, the solution would be expensive and time-consuming as repairs require regular follow-up inspections over the RPV’s lifetime. In the case of annealing flaws in a 60-tonne small modular reactor pressure vessel, on the other hand, the solution would be to send it back to the smelter and use another one from storage. Owing to the requirements of serialised production, SMR parts would be ordered in bulk, decreasing unitary costs and increasing standardization.^[48] Also, SMRs use smaller components, allowing a larger pool of potential vendors^[49], increasing competition and decreasing acquisition prices (in addition to bulk ordering). Each successive small reactor built through this standardised process would finally bring to fruition the cost-decreasing effect of industrial learning (“learning-by-doing” or “learning economies”), which has frustratingly eluded the nuclear industry for decades (despite being a well-known factor in branches such as shipbuilding^[50] or aircraft construction).

Construction economies can become manifest not just in the factory mass production stage, but also in cheaper transport costs: large reactors need high-capacity barges (decreasing available sites due to lack of access) to carry them and expensive super-heavy-duty cranes capable of hoisting hundreds of tonnes’ worth of Reactor Pressure Vessel to get into place. On the other hand, modules can be transported by rail flatcar or even by lowboy road trailer (depending on size) and set in position with cheaper and easier-available equipment. The number of personnel on an SMR construction site is lower, thus incurring a smaller daily fixed cost.

On the financiers’ part, SMRs provide both investment scalability and flexibility.^[51] As shown by Carelli et al. the modular nature and short construction time of deliberately small reactors offers the utility considerable adaptability in meeting market needs. A multi-module plant may be scaled-up sequentially, each SMR being deployed after the previous was brought online, decreasing the financial impact of each successive new unit via the cash flow from electricity sold.^[52] Or, they can be built concurrently, allowing the plant to reach full capacity as quickly as possible.^[53] Flexibility manifests itself in situations of market uncertainty, when SMR operators have the ability to more accurately size their investment (build on an ‘as-needed’ basis) according to market evolutions than LRs, due to the latter’s long lead times.^[54] The shorter construction times for SMRs (particularly after industrial learning has been internalised) compared to large reactors make them an attractive proposal for operators wary of financing costs.^[55]

According to Kessides and Kuznetsov, the cumulative effect of all these savings and cost reductions is that, instead of a 74% higher hypothetical overnight cost (for a 4-module SMR plant compared to a LR plant of equal total capacity), the capital cost difference decreases to a mere 4% compared to the single large reactor. The authors conclude: “Thus, the economics of SMRs challenges the widely held belief that nuclear reactors are characterized by significant economies of scale”.^[56]

Nevertheless, even if SMR economics compared to LR economics were worse than the (tentative) posted figure, the fact remains that SMR plants require lower upfront investment compared to LRs. As Ingersoll notes, “This price tag (for a LR) may be prohibitively high for many potential owners such as smaller countries or private utilities in the U.S. For these customers, the cost per kilowatt is irrelevant if they cannot afford to purchase the plant.”^[57] The reduced total cost of SMRs compared to LRs increases their appeal to small power companies or “nuclear newcomers”^[58] operating in deregulated energy markets or to utilities in countries where the multi-billion cost of a large reactor plant would be unaffordable from the start. Another draw for utilities in small/developing countries is the potential for SMRs to replace old fossil fuel (coal or bunker oil) power plants that have either reached the end of their operational lifetimes, or need to be decommissioned due to pollution issues.^[59] Since SMRs tend to be similar in power to older, smaller fossil fuel plants, they act as a drop-in, without the need to uprate the high-voltage lines, as would be the case with a large reactor.

In terms of personnel costs, plants using SMRs can attain savings compared to LRs due to their lower requirements for high-tier engineers. In general, a plant’s size determines the number of total engineers necessary, while its complexity determines the necessary average knowledge and training level. Deliberately small reactors employ considerably more passive, automatic safety features compared to large ones, thus the potential competence bottleneck of high-level operators can be avoided.^[60] Moreover, the multiplexed nature of modular plant control rooms bears the promise of fewer administrative chains.^[61] This is an attractive feature for (small or developing) countries whose technological educational systems lack the capacity to produce high-tier engineers or for whose nuclear training and relevant experience pools have atrophied due to lack of practical application.^[62]

Operational safety

The main separating factor between operating large reactors and deliberately small ones is the different approach to plant safety. As explained previously, the ever-expanding core size, heat output and radionuclide inventory, combined with the sensitivity to power fluctuations noted in the 1982 EPRI study has caused regulatory agencies to mandate the addition of successive layers of multiply-redundant engineered failsafes. As the forces (pressure, temperature, radioactivity) exerted in a gigawatt-range light-water-reactor are considerable, so too are the safety systems. Worse still, the basic safety profile of a LR is actively managed. In basic terms, systems need to be maintained in operation in order to keep the reactor’s parameters within limits after shut-off and avoid meltdowns, either of its fuel bundles (followed by the inevitable thermo-chemical hydrogen formation) or (admittedly much rarer) its RPV. Unsurprisingly, anti-nuclear critics latch onto this demonstrable weakness when proclaiming that nuclear reactors are barely-contained time-bombs.

In rejecting the “big is beautiful” design orthodoxy, SMR designers have attempted to implement Alvin Weinberg and Edward Teller’s concept of “walk-away safe”, a design where the core’s reactions ‘coast’ to a standstill without direct operator intervention.

According to several authors who studied the specifics of SMR safety, the three main routes adopted by designers in enhancing plant safety are:

1. Eliminating certain design features vulnerable to potentially initiating accidents;
2. Reducing the probability of accident occurrence;
3. Mitigating consequences of potential accidents.^[63]

Due to differing operating principles, the actual implementation of these concepts tends to take different forms depending on reactor type. However, they all share a number of common features which set them apart from large reactors:

Box 1: Common features of SMRs

• Reduced core size.
• Increased coolant inventory in their primary reactor vessel relative to core size compared to LRs.
• Elimination of large coolant piping. The entirety of primary coolant is located in the reactor vessel, while heat is transferred to the steam generators via a secondary loop through pipes of considerably smaller diameter. Since narrower pipes are cheaper to build in higher thicknesses, this also improves economics.
• Incorporating primary system components (pumps, steam generators etc.) into a single vessel.^[64] This enables the reduction of total pipe inventory, which decreases the likelihood of pipe cracks or seal failures.
• Facilitating the passive convection cooling of the reactor core and vessel. A large vessel surface and ample coolant volume allows natural temperature differentials to create density differentials within the coolant, enabling the formation of currents which carry decay heat away from the core without the need of active pumping. Furthermore, the integral reactor vessel is either partly submerged in a pool of water (in some designs, notably LWRs) or provided with air channels and heat pipes (in the case of helium, sodium, lead or salt-cooled reactors).
• Improved resilience to seismic events.

Sources: INGERSOLL (2009, pp. 592-593); INGERSOLL (2011 Conference presentation, p.5); FAIRHALL (2012, p. 7); CARELLI, M.D., P. GARRONE, G. LOCATELLI et al., (2010, p. 405), all cited in this paper.

It is necessary to mention, however, that not all SMR designs are fully passive. In fact, many light-water-based designs feature a number of primary coolant pumps in the integral reactor vessel.^[65] In layman’s terms, as power increases, it starts outpacing the vessel’s capacity to passively disperse heat. This is particularly problematic in LWR-based designs, since temperature increase can outstrip the pressuriser’s ability to keep water liquid (the primary loop contains water at over 100 degrees Celsius, above its boiling point) leading to a Loss Of Coolant Accident (LOCA). The designers of LWR SMRs have to choose between maximising passive functionality (with a loss in power and increase in vessel size), or improving output (accepting the decrease in passive heat removal, higher cost and increased complexity) via active pumping. In a sense, this dichotomy between passive safety and power output is a remnant of the “economy of scale” concept.

Investor wariness towards SMRs

There is ample statistical evidence supporting the point adopted by many scientists that purported “economies of scale” transformed into “diseconomies of scale” in the case of large plants. However, concrete evidence for deliberately small reactor economics has been largely unavailable (owing to the first few small reactors build being either pilot or LEAD^[66] plants). So far, the position rests on extrapolations based on their special design characteristics (en-bloc factory construction, standardised bulk part orders, modular emplacement, inherent safety measures, passive cooling, simplified functionality) compared to known issues in large reactors. Also, due to a still strong established belief in “economies of scale” from parts of the engineering community, the specific advantages of SMRs are disregarded or given lesser weight.

Another important issue is the high cost for initial adopters. Before learning economies and serial manufacturing bring per-unit prices down, the financial burden of building small reactor plants, even though lower in absolute terms than for large ones, will be higher than for subsequent ones. The First-Of-A-Kind (FOAK) effect disproportionally impacts initial adopter(s), posing a ‘catch-22’ issue to potential investors: companies are unwilling to invest lest costs are low enough, but in order for costs to decrease, several plants must be built until the Nth-Of-A-Kind (NOAK) point^[67] can be reached. Thus the classical industrial adage: “everyone wants to be second”. In this case, companies operating under government stewardship or in state-regulated markets have an advantage: costs for deploying LEAD/FOAK plants are (partly or fully) subsidised by the government. Western firms are likely to suffer stronger FOAK effects compared to their Eastern counterparts, due to lack of direct government stake in the R&D and construction process. Moreover, public opinion in many Western countries veers against the idea of incentivising nuclear energy, in general. It must be noted, however, that being first isn’t an entirely disadvantageous position. Completing a “first-to-market” build allows tangible evidence of the proposed technology for showcasing and promotion. It offers the company the option of organising tours of the facility for prospective clients, thus improving their negotiating position regarding their product’s merits.

Regulatory issues can cause uncertainty regarding the final acceptance of a particular reactor type. In turn, this uncertainty leads to financial circumspection from potential investors, notably in the case of designs deviating from the classical light-water-based technology that has been the yardstick in nuclear plant construction for the better part of 50 years. Furthermore, even though in the U.S, the Department of Energy (DoE) provides grants to companies researching SMRs, the actual sums are quite small on an individual basis, given the hundred-plus million dollar yearly costs for extensively testing materials, fuels and procedures. For instance, in 2012 the DoE earmarked $450 million, over the course of a five-year period, for supporting the R&D of up to two SMR designs.^[68] Consequently, each of the two SMR design receives $45 million per year, but the government support has to be cost-matched by the designer, otherwise the money is de facto inaccessible. This caused Babcock & Wilcox, who had initially won a bid for funding in the 2012 round to considerably downsize their mPower SMR R&D department.^[69] Westinghouse, who lost the bid, stopped their efforts altogether.

Instead of a conclusion

Today, over 400 reactors meet nearly 13% of the electricity demand of a 7 billion people planet. They do it 80+% of the time, instead of just when the wind blows or the sun shines. However, the classical power plant based on a few large reactors is no longer a panacea for all energy poverty ills. Due to its size, it can cause grid imbalance and it may be simply be too expensive, thus unattainable, for some pockets. Deliberately Small Reactors represent the road not taken during the first age. They are a conscious rejection of the engineering choices which led to the gigawatt-plus behemoths that ended up over time and over budget. They are smaller, simpler, and faster to build. Are they unequivocally better in every single instance compared to their large brethren? The jury’s still out. The evidence is still too sparse to tell either way. There are still points of contention between supporters of the old “economy of scale” and the as-of-yet unproved (though widely theorized) new “smaller is better” paradigm. Do SMRs have a place in the grid? Definitely. Not necessarily in all grids, but in most. How quickly will that happen? Probably not soon, but the promise these technologies hold for a good part of humanity is likely worth the effort.

 

 

---

FOOTNOTES:

1. KESSIDES, Ioannis N, “The future of the nuclear industry reconsidered: Risks, uncertainties, and continued promise”, Energy Policy 48 (2012), Elsevier, p. 191. ↑
2. Idem. ↑
3. Three Mile Island, Chernobyl, Fukushima. ↑
4. FAIRHALL, Graham, “Small Modular Reactors: Their potential role in the U.K.”, National Nuclear Laboratory, July 2012, p. 5. ↑
5. Ibidem 1. ↑
6. INGERSOLL, D.T., “Deliberately small reactors and the second nuclear era”, Progress in Nuclear Energy, 51 (2009), Elsevier Ltd., p.589. ↑
7. It can be argued that this particular point in time marks the end of the United States’ postwar economic boom. ↑
8. DAVIS, Lucas W., “Prospects for Nuclear Power”, in Journal of Economic Perspectives, vol.26, no.1, Winter 2012, p.51. ↑
9. Idem. ↑
10. Ibid. 6, p. 591. ↑
11. “if a piece of machinery is not economically competitive, make it bigger. This matched the economy of scale concept that the utility companies had been taught by Samuel Insull.”; ADAMS, Rodney, “The First Atomic Age: A Failure of Socialism”, Foundation for Economic Education, January 01, 1995, accessed on 05.08.2014 via http://www.fee.org/the_freeman/detail/the-first-atomic-age-a-failure-of-socialism ↑
12. Ibid. 6, p. 590 ↑
13. Part of the Oak Ridge Associated Universities (ORAU). It’s worth noting that, at the time Weinberg published his study, the Institute for Energy Analysis had disbanded following his retirement the same year. http://www.orau.org/about-orau/history/alvin-weinberg.aspx ↑
14. Ibid. 4. ↑
15. The S2G suffered from constant steam generator issues. Coolant safety was also a factor in its eventual elimination. ↑
16. For Canadians, who did not possess a national nuclear weapons programme, the ability to use natural uranium was a decisive factor in adopting CANDU technology. ↑
17. They also continuously scrubbed air, allowing submarines, for the first time in naval history, to stay submerged at will, without the need to snorkel. ↑
18. NORDHAUS, Ted, Jessica LOVERING, and Michael SHELLENBERGER, “How to Make Nuclear Cheap: Safety, Readiness, Modularity, and Efficiency”, The Breakthrough Institute, June 2014, p.13. http://thebreakthrough.org/images/pdfs/Breakthrough_Institute_How_to_Make_Nuclear_Cheap.pdf ↑
19. The Curta mechanical calculator, thus nicknamed due to its resemblance to an old-fashioned hand grinder. ↑
20. LOCATELLI, Giorgio, Chris BINGHAM and Mauro MANCINI, “Small modular reactors: A comprehensive overview of their economic and strategic aspects”, Progress in Nuclear Energy, 73 (2014), Elsevier Ltd., p.76. ↑
21. The natural tendency of heated, thus less dense, fluids to move upwards; in the case of a reactor specifically designed to take advantage of the effect, coolant loops around the primary circuit without pumping. ↑
22. Chernobyl’s deadly design flaw (high positive void coefficient) could have been kept under control had that particular experiment never been attempted in the first place. ↑
23. Ibidem 1. ↑
24. The hypothetical upfront cost of building a plant, assuming its construction would take a single day, without loan interest, monetary depreciation or escalation. ↑
25. Ibid 8, p. 54. ↑
26. Idem, p. 62. ↑
27. KIDD, Steve, “New nuclear build – sufficient supply capacity?”, 3^rd March 2009, Nuclear Engineering International, accessed on 05.05.2014 via http://www.neimagazine.com/opinion/opinionnew-nuclear-build-sufficient-supply-capability ↑
28. Ibid. 20, p. 79. ↑
29. The price at which electricity has to be generated over the operational lifetime of the plant for project break-even. ↑
30. LÉVÊQUE, François, Jean-Michel GLACHANT, Julián BARQUÍN, Christian VON HIRSCHHAUSEN, Franziska HOLZ and William J. NUTTALL – editors, “Security of Energy Supply in Europe: Natural Gas, Nuclear and Hydrogen”, Loyola de Palacio series on European Energy Policy, Edward Elgar, Cheltenham, UK and Northampton, MA, USA, 2010, p. 121. ↑
31. Ibid. 1, p. 188. ↑
32. The actual percentage of its installed capacity which ends up as generated electricity. Nuclear plants tend to have the highest capacity factors due to their status in the grid as baseload plants. Renewables have very low capacity factors due to intermittency and narrow-band operation. ↑
33. ROSNER, Robert and Stephen GOLDBERG, “Small Modular Reactors – Key to Future Power Generation in the U.S.”, Energy Policy Institute, Chicago, 2011, p.11. ↑
34. Idem, p.13. ↑
35. According to engineering experience in other industrial product areas, learning economies associated with the transition from FOAK to N^th-Of-A-Kind (NOAK) become noticeable after roughly 5-6 units. ↑
36. “Le coût de l’EPR de Flamanville encore revu à la hausse”, Le Monde, 4^th Dec. 2012, available at http://www.lemonde.fr/planete/article/2012/12/03/le-cout-de-l-epr-de-flamanville-encore-revu-a-la-hausse_1799417_3244.html ↑
37. It is worth mentioning that the myriad problems afflicting the two European plants did not surface in the case of similar ones currently under construction in China. Whether this showcases the existence of a learning effect, and a more streamlined regulatory environment, or, on the contrary, the consequence of a laxer one (where difficulties are not reported) is hard to ascertain, owing to the considerable informational opacity of Chinese sources. ↑
38. Ibidem 6. ↑
39. GUINESSY, Paul, “Small nuclear reactors raise big hopes”, Physics Today 63(8), 25 (2010), p.26. ↑
40. Ibid 38, p. 593. ↑
41. Ibid. 4, p. 7. ↑
42. KESSIDES, Ioannis N., and Vladimir KUZNETSOV, “Modular Reactors for Enhancing Energy Security in Developing Countries”, Sustainability 4 (2012), www.mdpi.com/journal/sustainability p. 1822. ↑
43. GOLDBERG, Stephen M., and Robert ROSNER, “Nuclear Reactors: Generation to Generation”, American Academy of Arts and Sciences, 2011, p. 21. ↑
44. CARELLI, M.D., P. GARRONE, G. LOCATELLI et al., “Economic features of integral, modular, small-to-medium size reactors”, Progress in Nuclear Energy, 52(2010), Elsevier LTD., p. 403. ↑
45. Ibidem 1. ↑
46. Ibid. 20, p. 76. ↑
47. ABDULLA, Ahmed, Inês LIMA AZEVEDO, and M. GRANGER MORGAN, “Expert assessments of the cost of light water small modular reactors”, Proceedings of the National Academy of Sciences, vol. 110, no. 24, June 2013, www.pnas.org/cgi/doi/10.1073/pnas.1300195110 pp. 9686-9687. ↑
48. ,Ibid. 20, p. 78. ↑
49. Ibid. 4, p. 6. ↑
50. Ibid. 33, p. 39. ↑
51. Ibid. 44, p. 404. ↑
52. VUJIĆ, Jasmina, Ryan M. BERGMANN, Radek ŠKODA and Marija MILETIĆ, “Small modular reactors: simpler, safer, cheaper?”, Energy, 45 (2012), Elsevier Ltd., p. 294. ↑
53. Ibid. 51. ↑
54. Ibid. 51, p. 405. ↑
55. Ibid. 52. ↑
56. Ibid. 42, p. 1825. ↑
57. Ibid 6, p. 594. ↑
58. Ibid. 20, p. 83. ↑
59. Idem. ↑
60. LOCATELLI, Giorgio and Mauro MANCINI, “The role of the reactor size for an investment in the nuclear sector: An evaluation of not-financial parameters”, Progress in Nuclear Energy, 53/2011, Elsevier Ltd., p. 218. ↑
61. Since each control room has its own “command structure”, there is considerable administrative duplication in the case of plants siting several large reactors. Given the safety issues of LRs, this is considered an acceptable financial burden. ↑
62. Ibid. 8, p. 58. ↑
63. Ibid. 4, p. 7. ↑
64. Ibidem 62. ↑
65. In a reactor, its steam generator is the primary heat exchange unit. The vessel is a secondary one. ↑
66. The “first-of-the-first” SMR plant; ref. ROSNER, Robert and Stephen GOLDBERG, “Small Modular Reactors – Key to Future Power Generation in the U.S.”, Energy Policy Institute, Chicago, 2011, p. 17. ↑
67. When the construction process has been cleared of “teething problems” and efficiency improvements (“learning economies”) have been implemented. ↑
68. “Obama Administration Announces $450 Million to Design and Commercialize U.S. Small Modular Nuclear Reactors”, March 22, 2012, available at: http://energy.gov/articles/obama-administration-announces-450-million-design-and-commercialize-us-small-modular ↑
69. “Funding for mPower reduced”, World Nuclear News, 14 April 2014, available at: http://www.world-nuclear-news.org/C-Funding-for-mPower-reduced-1404141.html ↑