17. U.S Reactor Types, Suppliers, Architect/Engineers, and Constructors 20. Early Operating Experience of U.S. Commercial Light Water Reactors 21. Classifications for INPO Evaluations. Figures Figure No. 28. Comparison of Commodity Requirements for Coal and Nuclear Powerplants ... Management of the Nuclear Enterprise INTRODUCTION In the previous chapter, alternative reactor types were reviewed in terms of safety, operability, and economics. While light water reactors (LWRs) lack some of the inherent safety features that characterize the alternatives, this comparison yielded no compelling reason to abandon LWR technology in favor of other, reactor types. The excellent performance records of some of the pressurized water reactors (PWRs) and boiling water reactors (BWRs) indicate that LWRs can be very reliable when properly managed. Large and complex nuclear units can also be built within budget and on schedule, as proven by some recent examples. These cases indicate that it is possible to construct and operate nuclear powerplants efficiently and reliably. Unfortunately, not all utilities perform to the same high standards. Numerous examples of construction malpractices and operating violations have surfaced in recent years, and many of these problems are serious enough to have safety and financial implications. Utilities evidently need to depend on more than government safety requirements and the conservatism of nuclear designs to compensate for errors. As the accident at Three Mile Island so vividly illustrated, LWRs are not entirely forgiving machines; they are susceptible to certain combinations of human error and mechanical failure. Although LWRs are built to accommodate to some problems in construction, maintenance, and operation, there is a limit to the extent of malfunctions and operational error that can be tolerated. The construction and operation of nuclear powerplants are highly sophisticated processes. Because nuclear technology is very complex and has the potential for accidents with major financial and safety implications, management of the nuclear enterprise must be of an intensity that is seldom required in other utility operations, or indeed, in most other commercial endeavors. Many utilities readily grasped the unique characteristics of nuclear technology and devoted their best management resources to its development. Others, unfortunately, seem to have misjudged the level of effort required to manage nuclear power operations successfully. This is not surprising, considering the variability in the nuclear utility industry. Forty-three utilities operate 84* nuclear powerplants, and 15 additional utilities are in the process of constructing their first nuclear units (40). Among these various organizations can be found a wide variety of management structures and philosophies, experience, commitment, and skill. While utilities are not the only organizations that seem to have underestimated the difficulties involved with nuclear powerplants, they must assume the ultimate responsibility for the safety and financial success of their plants. The diversity of the utility industry has not created major difficulties in managing nonnuclear generating plants. Many different organizational styles and structures have been used successfully to construct and operate fossil fuel stations and distribution systems. With the advent of nuclear technology, however, several new questions can be raised: • Is the technology so sensitive to its management that it is not adequately safe or reliable when poorly managed? • If so, can the quality of management be improved to a uniformly acceptable level? • Alternatively, can the technology be modified so that it is less sensitive to its management? Management and quality issues will be addressed in this chapter by illustrating the sensitivity of nuclear power operations with a few recent examples, a look at factors that contribute to such problems, and a review of current efforts to ensure uniformly high levels of performance. *Includes all plants with operating licenses, even though some (Three Mile Island Units 1 and 2, Dresden 1, and Diablo Canyon 1) are not currently in operation. VARIATIONS IN QUALITY OF CONSTRUCTION AND OPERATION The following discussion will address variations in quality of construction and operation on three levels: 1. An overview of the nuclear industry will be presented to demonstrate that various projects differ significantly from one another. This will provide a qualitative basis for assessing the sensitivity of the technology to its management. 2. Some of the more successful plants will be examined to identify the conditions under which it is possible for nuclear powerplants to be constructed and operated to the highest standards of quality. 3. Some less successful plants will be examined to identify the factors that contribute to poor management and to understand the cost and safety implications. The examples that have been selected for discussion are not intended to fully span the range of good and bad practices; they are, however, useful in illustrating the differences in the ways in which nuclear power has been implemented in recent years. Construction The construction of nuclear powerplants in the United States is far from being a standardized process. As shown in table 17, a utility must choose among several reactor types and vendors and among an even larger selection of architect/ engineers (AEs) and constructors. Wide differences in design and construction practices can result from these various combinations. A utility can superimpose additional changes on the basic design to customize its plant according to its special needs or to accommodate to specific site requirements. Such factors partially explain the variations in construction time and quality discussed below. There are no simple measures of quality in construction, and no attempt will be made to develop comprehensive measures. But efficiency in construction can be partially indexed by construc Table 17.-U.S. Reactor Types, Suppliers, Architect/Engineers, and Constructors Reactor types: Pressurized water reactors (PWR) High temperature gas-cooled reactors (HTGR) Babcock & Wilcox Co. (PWR) Combustion Engineering, Inc. (PWR) Westinghouse Electric Corp. (PWR) Bechtel Power Corp. (AE, C) J.A. Jones Construction Co. (C) Stone & Webster Engineering Co. (AE, C) United Engineers & Constructors, Inc., (AE, C) Wedco (a subsidiary of Westinghouse Electric Corp.) (C) SOURCE: "World List of Nuclear Power Plants," Nuclear News, February 1983. tion time and cost, which differ widely among the utilities shown in table 18. Only plants beginning commercial operation after the accident at Three Mile Island were included, so all of these units were affected to some degree by the regulatory changes that have occurred since 1979. These data should be interpreted with some care. Several of the longer construction times may reflect inordinate licensing delays or a utility's decision to delay construction in response to slow growth in the demand for power. In addition, some of the projections for very short construction times may be overly optimistic. It is also difficult to make direct comparisons of construction costs since they are based on different accounting schemes. Furthermore, both estimates and actual expenditures are reported by the utilities in "current dollars." Annual expenditures are then summed without accounting for the time value of money, with the total construction costs ex 50 Table 18.-Construction Records of Selected U.S. Light Water Reactors 8 8 8 8 Longest construction times: pressed in terms of "mixed current dollars." This accounting system tends to further distort actual costs. It is interesting to note that the best and worst construction schedules from table 18 differ by an average of about 6 years. In fact, the plants with the longest construction times took twice as long to complete as those with the shortest schedules. Dramatic differences also can be observed in the costs, even when the construction schedules are similar. For example, the Callaway 1 unit is projected to cost $2,500 per kilowatts electrical (kWe) after 8 years of construction, while the Byron 1 plant is projected to cost only 60 percent of that with the same construction schedule. A recent study by the Electric Power Research Institute (EPRI) attempts to identify the reasons for the variations noted here (3). In a statistical analysis of all nuclear powerplants, it was found that 50 to 70 percent of the variation in leadtime could be accounted for by regulatory differences, deliberate delays, and variations in physical plant characteristics. EPRI ascribed the remaining variation to management practices and uncontrollable events. To more fully understand the importance of utility management in the construction phase, it is valuable to examine a few specific examples. Two of the more notable nuclear powerplant construction projects are Florida Power & Light Co.'s St. Lucie 2 unit at Hutchinson, Fla. and the Palo Verde 1 plant at Wintersburg, Ariz. owned by Arizona Public Service Co. As shown in table 18, both units are projected to be completed with relatively short construction schedules. Neither utility has encountered significant regulatory difficulty nor much opposition from intervenors (6). Both units had to accommodate to the wave of backfit and redesign requirements of the Nuclear Regulatory Commission (NRC) that followed the accident at Three Mile Island, and yet no significant delays have been experienced at either site. These examples indicate that nuclear powerplants can be constructed expeditiously, even in the most difficult regulatory environment. In contrast to these examples, other plants have had a long history of problems. Quality control in nuclear powerplants, as in other commercial endeavors, is important in assuring consistency and reliability. In industries such as nuclear power and aerospace, where the consequences of failure can be severe, quality is guaranteed by superimposing a formalized, independent audit structure on top of conventional quality control measures in design, procurement, manufacturing, and construction (30). Deficiencies in the quality control procedures at nuclear reactors are cause |