Sowell's Law Blog: The Truth About Nuclear Power

Subtitle: Nuclear plants are huge to reduce costs

In this series of articles on the truth about nuclear power plants, one focus area is the economics. Others include safety, financing, water usage, different technologies, and a few more. This article addresses a part of the excessive cost issue: why are nuclear power plants so huge? One description of these plants is “cathedrals”. For perspective, modern designs are approximately 1100 MWe per reactor, however a French design has 1,600 MWe per reactor.

Another reason for the articles on nuclear plant costs is the argument by nuclear proponents that the only reason the plants cost so much is the opposition and lawsuits brought by anti-nuclear groups. That is simply not true, as these articles demonstrate. The very nature of a nuclear power plant design, its inherent features as part of using nuclear fuel for heat, requires more equipment, larger equipment, more costly alloys, longer construction times, and attendant costs for equipment inflation and financing interest. Even without lawsuits, nuclear power plants remain the highest cost plants for baseload power production. Several studies reach the same conclusion on this point.

This article delves into attempts to reduce costs by three aspects of economy of scale,

1) where bigger is cheaper if a manufacturing process is based on a circle or sphere;

2) where mass production reduces costs; and

3) where a learning curve makes future projects more efficiently constructed, in theory, at least.

Those are three of the elements of gains due to economy of scale. Each is addressed in turn.

Economy of scale based on a circle or sphere

This takes advantage of the fact that a pipe or tube has a circular cross-section. A pipe with double the diameter can carry four times the quantity compared to the original pipe. Also, if one doubles a sphere’s diameter, it holds eight times the volume. These basic facts of engineering are employed all over the world in process plants, refineries, power plants, chemical plants, and any other process where fluids are moved through pipes or stored in spheres. To a lesser extent, material storage in cylindrical tanks also achieve economy of scale as the tank diameter increases.

Over time, nuclear power plant designs have increased in output from 600 MW to 1000 MWe to 1200 MWe to 1600 MWe, all as measured in electrical output, MWe. These increases in size were attempts to reduce costs through economy of scale. It is appropriate to pause here, and consider that statement. If, as nuclear proponents assert, nuclear power really is as cheap as 4 cents per kWh, or 6 cents, why would it be necessary for successive designs to be bigger and bigger, trying to drive down the costs of the power produced? The very fact that modern designs are bigger than previous designs puts the lie to the cheap power argument.

Since most of a nuclear power plant has fluids, steam or water, flowing through a pipe or some similar item based on a circular cross-section, the first economy of scale applies. As examples, the major equipment includes cylinders for the reactor vessel, steam generator, steam condenser, and containment structure. Steam turbines also are based on a horizontal cylinder. All the pipes, pumps, fittings, and valves also are based on a cylinder. To illustrate, a pipe carrying water at 7 feet per second, approximately 2 meters per second, can convey 22 cubic feet per second. If the pipe diameter is doubled, with velocity maintained at 7 feet per second, the water volume goes up by a factor of four to 88 cubic feet per second. Similarly, if one wishes to double the plant size, for example from 600 MWe to 1200 MWe, pipes can achieve double the flow with an increased diameter of only 1.4 times the original size. If the initial pipe is 2 feet or 24 inches in diameter, the pipe size that is required for double the flow is only 34 inches in diameter. This is significant because a 34 inch pipe usually will not cost double that of a 24 inch pipe, but will cost only about 30 percent more than the 24 inch pipe. The cost savings occur throughout the plant, wherever equipment is based on a circular cross-section.

Similarly, the third level of containment required by the NRC, the containment structure, typically has a dome for the roof. A dome is half of a sphere, and economy of scale applies here, too. The result is a plant with twice the production capacity, but a containment structure that costs approximately 1.3 times that of a smaller, half-sized plant.

However, one factor works against the gains afforded by economy of scale. That factor is the low steam pressure and temperature produced in a modern nuclear reactor system. This requires that more steam must be circulated to produce the same amount of power. Therefore, all equipment must be larger: reactor, steam generators, steam turbines, condensers, pipes, pumps, cooling towers, everything except the generator and electrical transmission equipment. This is due to the low-pressure, saturated steam that is inherent in the PWR design. The steam has no superheat and cannot use supercritical pressures due to risk of large pipes bursting, or the pipes must have prohibitively expensive thick walls. There are proposals to overcome this problem by burning natural gas in supplemental boilers to provide superheat to the steam. However, nuclear proponents are very much against natural gas and shudder at the thought that their nuclear power plant must allow natural gas on the premises.

Economy of scale from mass production

In this aspect of economy of scale, the cost of each unit of production decreases where a large number of units are produced. This is a very old concept, perhaps starting with the assembly line and manufacturing cars. Certainly, the concept applies in many manufacturing processes. For nuclear power plants, with only approximately 400 plants in the entire world, built over several decades, there has never been much opportunity to achieve mass production. One reason for this has been the lack of standardized designs. If each design is unique, the unique parts must each have its own design, drawings, manufacturing process that may require special jigs, possibly its own transportation system, installation system, etc. This includes the major equipment such as the reactor, steam generator, steam condenser, steam turbine, power generator, auxiliary systems, containment structure, and other items. The cost of each plant, therefore, suffers from a lack of mass production. In the US, this problem has long been recognized and an attempt was made to standardize the design of new plants. The Westinghouse AP-1000 design is supposed to solve the unique design aspect, so that each new plant will use the same design. This has yet to occur, as the first such plants are currently under construction in Georgia at the Vogtle power plant. These are, by definition, first-of-a-kind in the US. The Vogtle plant has many drawbacks and serious issues, which will be addressed in one or more future articles. It is notable that the Vogtle plant is building two reactors adjacent to each other, where there is at least a minimal opportunity to employ a small form of mass production. However, building two of an item is not much better than building only one. It requires building many items to reduce costs substantially.

Economy of scale from a learning curve

The cost reductions from a learning curve applies to the second and subsequent projects, which attempt to apply the lessons learned from building the first project. This aspect of economy of scale works fairly well at times, but only when the lessons learned are effectively transferred to the subsequent projects. In some cases, even when the lessons are communicated, the subsequent project has unique aspects that cannot apply the lessons. Different geography, site conditions, climate and weather, all are examples of potential reasons why lessons cannot be applied. There may also be different construction companies with different business philosophies, different equipment used in construction, and many other reasons why lessons learned will not be applied. As above, using the Vogtle plant as the example, the two reactors are being built next to each other, and on a schedule so that one plant should follow about two years behind the first. This will allow the second plant to take advantage of lessons learned, from the learning curve experienced in building the first plant.

Conclusion

Nuclear power plants cost far too much to build due to their inherent design for use of nuclear fuel. The plants are huge to try to reduce the costs by economy of scale.

It can be seen that nuclear power plants have attempted to reduce the very high costs of construction by designing and building larger plants, by building multiple reactors at the same site, and to apply lessons learned from recent plants to the new plants. It is still an interesting question, why should any of this be necessary if nuclear power plants truly produce electric power at the lowest cost of any type of baseload plant? Some nuclear advocates go even further, with the assertion that nuclear plants can also be used as load-following plants. If the costs of nuclear power were truly as low as the advocates maintain, there would be no reason ever to employ the well-known strategies of economy of scale.

Benefits and drawbacks from building smaller, modular plants will be addressed in a future article in the series.

Previous articles in The Truth About Nuclear Power series can be found at the following links.

Part One – Nuclear Power PlantsCannot Compete

Part Two – Preposterous PowerPricing if Nuclear Power Proponents Prevail

Part Three – Nuclear Power PlantsCost Far Too Much to Construct

Part Four – Nuclear Power PlantsUse Far More Fresh Water

Part Five – Cannot Simply Turn Offa Nuclear Power Plant

Part Six – this article

Part Seven -- All Nuclear Grid Will Sell Less Power

Part Eight – No Benefits from Smaller Modular Nuclear Plants

Part Nine -- Nuclear Plants Require Long Construction Schedules

Part Ten -- Nuclear Plants Require Costly Upgrades After 20 to 30 Years