14.53. Many small developmental reactors in a number of countries have been decommissioned without problems. A small (58-MWt) BWR with a 118-ft-high, 85-ft-diameter containment at Elk River, Wisconsin was shut down in 1968 and completely dismantled from 1972 to 1974, with the site converted to a parking lot. The cost was $5.7 million.

14.54. The 6-year decommissioning of the Shippingport, Pennsylvania Station, starting in 1984, one year after shutdown, served as a demonstra­tion of the prompt dismantling option [10]. Although the reactor operated at only a modest rating of about 500 MW(t), the pressure vessel had a diameter of 10 ft, which is comparable to the 15-ft-diameter of present large PWR vessels. Other comparisons indicate that the decommissioning operation was indeed relevant.

14.55. Since the reactor pressure vessel is the largest and the most highly irradiated component in a reactor plant, its removal and disposal is the most challenging decommissioning operation. In the case of Shippingport, it was possible to prepare a monolithic package including the vessel, in­ternals, and concrete shield, remove it from the containment, and ship it by barge to a burial site at Hanford, near Richland, Washington. For another reactor, rail transportation might be necessary. This is likely to require segmentation, which would lengthen the project. Although steam generator removal is also a challenge, there has been ample experience with removal and replacement in existing plants.

14.56. The approximate $100 million cost of the Shippingport project provides a useful reference point for estimating decommissioning costs for presently operating reactors and those to be ordered. However, cost es­timates often show large variations, with site-specific considerations and regulatory uncertainties playing a factor [11]. A logical approach is to include an expense for future decommissioning as part of the annual energy generating costs. For example, in Table 10.2, we have arbitrarily included an annual allowance of $5 million for future decommissioning in the listing of energy generation costs. Considering the time value of money at an annual compound interest rate of 4 percent, this will yield $149 million in 20 years, $280 million in 30 years, and $475 million in 40 years for decom­missioning. Of course, escalation (inflation) could add to the future cost of decommissioning. On the other hand, should this become significant, a higher compound interest rate may be appropriate for the accumulating funds. Since such cost projections are in the range of most estimates, it is reasonable to assume that nuclear power plants will not become an eco­nomic or environmental burden after they have completed their useful life.

GENERAL REFERENCE

Rahn, F. J., et al., “A Guide to Nuclear Power Technology,” John Wiley & Sons, 1984, Chap. 12.

REFERENCES

1. R. G. Rosenstein et al., “Proc. Topical Meeting on Advances in Fuel Man­agement,” Pinehurst, NC, American Nuclear Society, 1986.

2. К. M. Taylor and A. Williams, Trans. Am. Nucl. Soc., 63, 388 (1991).

3. Y. Fujita et al., Nucl. Technol., 95, 116 (1991).

4. R. E. Uhrig, Nucl. Safety, 32, 68 (1991).

5. “National Energy Strategy, 1991/1992,” U. S. Dept, of Energy, 1991.

6. “Final Rule on Nuclear Power Plant License Renewal,” U. S. NRC Report SECY 91-138, 1991.

7. T. E. Murley, Nucl. Safety, 31, 1 (1990).

8. T. J. Griesbach, Trans. Am. Nucl. Soc., 64, 263 (1991).

9. R. Bardtenschlager et al., Nucl. Eng. Des., 45, 1 (1978); R. I. Smith et al., U. S. NRC Report NUREG/CR-0130, 1984.

10. F. P. Crimi, Trans. Am. Nucl. Soc., 56, 72 (1978).

11. K. W. Sieving, Trans. Am. Nucl. Soc., 67, 219 (1993).

INTRODUCTION What Is Needed

15.1. “Next generation” reactor concepts have been developed to pro­vide some enhanced safety and economic features for new plants that may be ordered during the 1990s and subsequent years. However, before con­sidering such designs, we should examine the need for additional generating capacity and the likely competitiveness of nuclear plants in meeting this need.

15.2. New generating plants are needed to replace plants that are retired from service and to meet growth in demand. Need projections have been the subject of numerous studies [1]. In the United States an annual growth rate of electricity sales of about 3 percent during the late 1980s and early 1990s indicates a likely need for additional generating capacity after the year 2000.

15.3. A factor in energy planning is the effectiveness of demand-side management. In such management, utilities attempt to influence customer use patterns in various ways so that the efficiency of utilization is improved.

The term efficiency is preferred to conservation to avoid the impression that energy is being saved by sacrificing the quality of life. For example, lower rates during hours of low demand provide an incentive to shift elec­tricity use away from peak hours, thus making better use of generating capacity. Aggressive programs to help individual customers improve their utilization efficiency are being carried out by many utilities. However, despite such efforts, it is reasonable to assume that the demand will outgrow supply in the near future and that there will be a need for new generating capacity. Such a need must be anticipated some years in advance to allow time for construction. Whether a significant number of nuclear power plants are constructed to meet the energy demand will depend on the resolution of several challenges.

15.4. New nuclear power plants will be built only if both the public and electric utility management are convinced that such new construction is in their own best interest. Public attitudes vary from strongly negative by a militant small fraction to various levels of uneasiness by other groups as a result of the Three Mile Island and Chernobyl accidents and delays in managing high-level wastes. However, motivated by concerns regarding atmospheric pollution and the greenhouse effect caused by fossil-fueled power plants, there appears to be an increasing willingness by those who hold reservations to accept the construction of new plants provided that they would be extraordinarily safe.

15.5. The safety of present LWR reactors is considered completely ad­equate, particularly following numerous systems improvements made as a result of the lessons learned from the Three Mile Island accident. However, the poor economic experience of some utilities has raised some questions regarding the wisdom of new investment for nuclear power plants. Al­though a high rate of inflation during the 1970s, as well as orders unjustified by energy demand, contributed to plant cancellations, poor management, a lack of standardization, and regulatory inconsistencies were factors. Also, for many U. S. operating reactors, plant capacity factors were less than those for similar plants in other countries. Therefore, from the utility viewpoint, there is a need for financial risk reduction.

15.6. New reactor designs with “passive” safety features do not require prompt operator intervention in the event of a wide variety of accidents. In addition, such features permit simplification of many of the safety sys­tems, with consequent cost savings. Therefore, these designs tend to meet both the safety concerns of the public and the utility need for systems of reasonable cost. We will examine the important concepts as well as the utility financial risk matter. As provided by 10 CFR 52 (§12.239), design certification procedures for several of these concepts are in progress.

13.56. The pressure tubes are supported by tube sheets at the two ends of the cylindrical reactor vessel, called a calandria, with its axis horizontal. The calandria contains the heavy water at essentially atmospheric pressure. Some heat (about 5 percent of the reactor total) is generated in the mod­erator by the slowing down of fast neutrons. The moderator is consequently circulated through a heat exchanger and also through a purifier.

13.57. A fairly complex piping system serves to connect the inlet and outlet ends of each pressure tube to coolant manifolds (see Fig. 13.12). Pressurized heavy-water coolant flows between and around the fuel rods in each bundle; the flow is in opposite directions in adjacent channels. The heated coolant leaving the reactor is collected in a header (or manifold) from which it passes to a steam generator of the conventional type. The coolant is then pumped back to the reactor. A simplified representation of the CANDU cooling system is shown in Fig. 13.13. Two coolant exit headers are provided, one at each end of the calandria, and each header is attached to a steam-generator-pump loop. A single pressurizer, of the same type as is used in PWRs (§13.16), maintains the coolant system pressure.

Power Thermal Electrical (net) Specific power Power density

Length

Diameter

Rod, OD Clad thickness Spacing between rods Rods per bundle Bundle length Bundle diameter U02 per bundle Fuel loading, U02

Ave. burnup 650 GJ/kg

2064 MW 600 MW 24 kW(th)/kg U 7.46 MW(th)/m3

Core 5.94 m (19.5 ft) 7.7 m (25.2 ft)

Fuel

13.1 mm (0.515 in.)

0. 38 mm (0.015 in.)

1.02 mm (0.0047 in.)

37

0.495 m (19.5 in.)

0.102 m (4.0 in.)

21.3 kg (46.5 lb)

97 x 103 kg (214 x 103 lb)

(7,500 MW • d/t)

15.35. The low power density of gas-cooled necessitates very large cores for reactor plants having electrical generating ratings in the 600 MW or higher range. Prestressed concrete reactor vessels (PCRVs) to contain such systems are proven and safe, but are expensive. By dividing the desired capacity into smaller units, steel pressure vessels can be utilized. Also, the components and subsystems for such small modular units lend themselves to factory fabrication with resulting economies.

15.36. The plant capacity factor of the Fort St. Vrain plant was dis­appointing as a result of component problems, particularly with the water — lubricated circulator bearing seals. Although the modular HTGR (MHTGR) uses a proven magnetic bearing system for the circulator, which should be trouble-free, the Fort St. Vrain experience demonstrated the economic sensitivity of a large plant to excessive shutdowns. In a modular plant built up from small units, maintenance on individual unit components can be performed more easily than in a large plant. A multiunit plant also permits considerable fuel management flexibility.

15.37. A typical MHTGR plant consists of four identical modular re­actor units housed in separate adjacent reinforced concrete structures lo­cated below grade, but under a common roof. The reactor units, each producing 350 MW(th) of steam, are paired to feed two turbine generators in a separate energy conversion area to generate a total of 538 MW(el).

Introduction

12.237. We have already mentioned several aspects of the licensing and regulation of nuclear plants, which in the United States are the legal re­sponsibility of the Nuclear Regulatory Commission (NRC). Consistent with the need to assure a high level of safety, the procedures required are lengthy, with much documentation required. In the past, two stages of approval have been required, the first for a construction permit, the second after the plant was built, for an operating license. Since public input op­portunities at each stage permitted various groups who were opposed to nuclear power for philosophical and other reasons to obstruct the licensing process, long delays were common. These delays, particularly after the plant was built, added greatly to the financial burden to both the owning utility and the eventual ratepayers. An additional complication has been the need to satisfy requirements at the state level. For example, as a result of objections by New York State authorities regarding emergency planning, the Shoreham Plant, a new 800-MW(el) BWR, was dismantled, starting in 1992, without ever being operated.

12.238. Such experiences have discouraged investment in new nuclear power plants. In fact, the establishment of a stable and predictable licensing environment is considered essential by industry before a new plant is likely to be ordered. There is general agreement that reforms are needed, but they are still evolving during the early 1990s. However, we will describe the essential features of likely licensing reforms that will improve the in­vestment environment and yet satisfy the need to “assure public health and safety.”

12.239. In 1989, the NRC issued a landmark rule, 10 CFR 52, which provides for the certification of standardized plant designs, the issuance of early site permits, and the issuance of a combined construction and op­

erating license. To avoid delays after construction was completed, the pro­cedures were clarified in 1992 to provide for an informal post-construction hearing but limit it to issues of nonconformance to the previously approved combined license. We will describe these steps in the following sections.

14.25. The Three Mile Island accident focused attention on the short­comings of most nuclear power plant control rooms existing at that time since operator response during the accident was complicated by confusing displays, many alarms, and an “overload” of information. Diagnosis of the problems during the stress of the accident was difficult. As a result, practically all control rooms were upgraded to meet new NRC require­ments. Human factors engineering, which is concerned with the interface between people and machines, was recognized as an essential contribution to control room design.

14.26. Control room designs for new nuclear power plants have been developed with important features that we will examine. Much attention has been given to providing the operators with an easily understood struc­tured information display arranged in order of importance, taking advan­tage of extensive computer modeling of plant systems. Expert system and neural network features are used in some systems to provide guidance for remedial actions [3].

14.27. The primary control room objective is to help the operator make the decisions that are necessary, particularly under unexpected circum­stances. To accomplish this, designers have made the most important plant safety and power production information readily apparent on fixed indi­cators, while reducing the number of distracting displays. Less vital infor­mation is made available by a data processing system on color graphic and other video displays. A third level of information is provided on cathode ray tube (CRT) monitors by calling up suitable “pages.” Similarly, alarms are categorized by priorities so that operators are alerted to problems but not overwhelmed with many signals. A “safety panel” informs the operator of the status of the plant during an accident with data processed so that guidance to corrective action is provided. Feedback indicates the effect of such actions.

14.28. Another new feature is a large screen, visible throughout the control room, which provides a graphical color display of the status of the plant and critical functions. This display overview links the vital fixed indicator information with that provided by CRT “pages.” Past major dependency on such CRT access to a great deal of information diminished the operator’s “feel” of the plant condition. Thus, the large display aids coordination and diagnostics. The development of an optimum balance between the various informational and operating tools has required much experimentation using computer-driven plant simulators.

14.29. A typical advanced design control room layout is shown in Fig. 14.2. The operation of major plant functions, such as the core, primary coolant system, and turbine generator, is centered at the master control console (MCC). Other panels serve the safety and auxiliary systems. The MCC is engineered using anthropometric standards for a seated operator

A. Master Control Console

B. Integrated Process Status Overview

C. Auxiliary System Panels

D. Safety System Panels

E. Control Room Supervisor Area

F. Technical Support Center G> Shift Supervisor Office

H. Assistant Operator Work Stations Computer Room

Оліплі/і бкнЫлшп Danole /Кілі

as shown in Fig. 14.3. Anthropometry deals with the study of human body measurements.

14.30. All new control rooms use digital multiplexing for signal trans­mission. In multiplexing, many different signals are sent over the same wire using different frequencies. Individual sensors are wired to remote terminals where they are sequentially sampled for transmission to the control room area, where the signals are separated and distributed to information­processing equipment. In this way, it is possible to save hundreds of miles of hardwiring.

15.63. The four concepts previously described incorporate passive fea­tures in basically proven systems that have been converted to a smaller size amenable to modular construction. For the two LWRs, the passive features permit substantial system simplification with accompanying eco­nomic benefits. An improvement in public acceptance is also expected.

15.64. Other concepts have been proposed which have an improvement in public acceptance as a major objective. Probably the most interesting of these is the process inherent safety (PIUS) reactor developed in Sweden. Also of interest is the safe integral reactor (SIR), which has evolved from a system originally proposed for maritime use. We will briefly describe each of these.

13.18. An evolutionary reactor design is one that draws heavily upon a successful previous design but incorporates many improvements that are desirable as a result of technological development, experience, and possibly new requirements. The basic features are retained so that there would be a high level of confidence that a new plant would operate reliably without the need to first build a demonstration plant.

13.19. An example of this approach for a PWR that is to be available whenever orders for new reactors are forthcoming is the Combustion En­gineering System 80 Plus™ reactor model [4]. Several general needs were addressed in developing this design for a so-called “next-generation plant” which provide some background for new reactor trends. A European PWR design meets similar needs.

13.20. The opinion of utilities regarding desirable characteristics of fu­ture plant designs was determined by EPRI. Of major importance was the need to satisfy public acceptance with regard to safety, economics, and reliability. Design simplification, standardization, increased safety margins, and various technical improvements were felt to help satisfy public con­cerns. However, other issues also affect the attractiveness of utility in­vestment in new plants. Predictable licensing and a stable regulation climate were considered necessary to avoid nontechnical and expensive delays that have plagued the industry in the past. Also recognized was the need to have fair treatment by state regulatory agencies on financial and planning matters.

13.21. Design simplification is accomplished by eliminating unnecessary duplication of functions performed by separate systems or in some cases combining component functions. Both simplification and improved plan­ning should lead to shorter construction schedules, which are essential for improved economics. It is also desirable to include in a new plant features that will mitigate the effects of a severe accident. For example, changing the reactor vessel geometry to deal with molten core material can be ac­complished in a new design at relatively little cost.

15.7. Passive reactor plants have about one-half the electrical generating capacity of present large units. It is felt that such smaller units are more practical for medium-sized utilities that wish to increase their capacity than the large units. Generally, a generating unit should be no larger than about 15 percent of a utility system capacity. Therefore, worldwide market con­siderations favor a smaller standard plant. Since smaller units lend them­selves to a greater degree of standardized factory manufacturing than large units, cost savings are likely. Another consideration is that smaller plants contain less radioactive material and are easier cooled. Some level of risk reduction therefore results.

13.58. Since there is only slight neutron absorption in the coolant and the moderator is kept cool, there is little decrease in reactivity from shut­down to power in contrast to that for LWRs. Continuous refueling also avoids reactivity burnup changes. The core reactivity inventory requiring compensation by the control system is therefore small. The coolant tem­perature and void coefficients are slightly positive; nevertheless, the power coefficient is negative, although small, as a result of the overriding effect of the negative Doppler fuel coefficient. Thus, the system is inherently stable.

13.59. Several types of control elements are provided in the CANDU reactor. Primary shutdown capability is provided by 28 vertical absorber (shutdown) rods with secondary shutdown, if necessary, by high-pressure injection of gadolinium nitrate solution into the moderator. Since the core is large and the negative power coefficient is small, xenon oscillations, which are slow, require control (§5.76 et seq.). The neutron flux level is therefore measured continuously at many points (—100) in the core and the flux level adjusted as necessary. Flux flattening is provided by 21 absorber rods called “adjusters.” In addition, there are 14 “zone control” absorbers for bulk reactivity control and local suppression of flux oscilla­tions. These absorbers consist of vertical chambers which can be filled with ordinary water to any desired level. Four vertical (solid) absorber rods supplement the zone control absorbers.

15.38. The use of refractory-coated fuel microspheres is a key feature of the MHTGR as a result of their ability to retain fission products under severe conditions. There are two types of microspheres: fissile, containing a uranium oxycarbide (UCO) kernel with the uranium enriched to 20 percent, and fertile, containing a kernel of Th02. The fertile material enhances the conversion ratio.

15.39. Each fissile kernel, of about 350 (xm in diameter, is first coated with a porous graphite buffer, followed by three successive layers of pyr­olytic carbon, silicon carbide, and pyrolytic carbon, with an outer diameter of about 800 |xm achieved. The fertile particles, are similarly coated, but slightly larger. In earlier reactors, only the fissile particles, designated as TRISO particles, contained a silicon carbide layer, to permit fuel recycling, which is not a current option. Both types of coated particles are mixed with an organic binder and graphite filler, then fired to form fuel rods, which are inserted into holes in hexagonal graphite core blocks (§15.41).

15.40. The inner layer of silicon carbide prevents the escape of most fission products, while the outer layer of dense pyrolytic carbon provides mechanical support and acts as backup containment. Extensive testing has confirmed the ability of the coated particles to maintain their integrity and retain fission products up to sustained fuel temperatures of 1760°C. There­fore the safety design philosophy of the MHTGR is that control of radio­nuclide releases can be accomplished by their retention within the fuel particles rather than by active features or operator actions [5].