In the mid-1980s, US utilities were still fully regulated by state Public Utility Commissions (PUCs) and they were still implementing a host of Nuclear Regulatory Commission (NRC) imposed backfits, as a result of the accident at Three Mile Island. It was an atmosphere in which operating and maintenance (O&M) costs were high and regulatory uncertainty was a major concern.
For new nuclear plants, the primary source of competition was expected to be coal-generated electricity. Natural gas was deregulated in 1985 and no one could have foreseen that its price would drop so low that it would become the overwhelmingly preferred fuel for new power plants. It was also assumed that new plants would be ordered by regulated utility companies.
It was at this time that industry and government launched the Advanced Light Water Reactor (ALWR) programme, intended to develop a new generation of reactor designs for the US. The original intent was to have several new designs completed and certified by the NRC by the early 1990s, although the programme was not actually completed until late in that decade.
During the ALWR programme’s early years the NRC implemented a new regulation for certifying standardised reactor designs (10CFR52), and industry developed an ALWR Utility Requirements Document, which spelled out the criteria that potential customers would specify when they began ordering new nuclear plants. With DOE support, the reactor suppliers developed designs to meet the Utility Requirements Document and submitted them for certification by the NRC.
The utilities developing the ALWR Utility Requirements Document were most concerned with reducing O&M costs and resolving all potential NRC safety issues. Because projected costs for new coal-fired plants (which would have to meet stringent clean air standards) were so high, the pressure to reduce nuclear capital costs was minor, when compared to the other issues.
To minimise the introduction of new uncertainties and new regulatory issues, a guiding principle of the ALWR programme was to consider only ‘evolutionary’ changes. Little new technology was introduced. Except for a handful of optimisation issues, the ALWR designs were developed by taking existing NRC requirements and extending them to address the NRC’s list of unresolved safety issues, and the severe accident issues that emerged after the Three Mile Island accident. Probabilistic risk analyses were used to identify weaknesses in the designs, which were then corrected. On top of this, there were a number of ALWR requirements for design features that would make O&M easier.
Designs intended to have lower capital costs were under pressure to implement new requirements that would in fact increase capital costs. The philosophy was that higher overnight capital costs would be offset by shorter licensing and construction schedules and reduced O&M costs. This was acceptable because the primary economic criterion was based upon total life-cycle costs – the economic criterion of regulated utilities.
Another industry initiative launched at the end of the 1980s, the Strategic Plan for Building New Nuclear Power Plants, comprised fourteen building blocks that needed to be addressed in planning for new plant designs. The building blocks were related to institutional issues. Because ALWRs were based on evolutionary changes, there were no major technology issues that required resolution.
THE EFFECT OF DEREGULATION
The US marketplace now looks very different. Deregulation is dramatically affecting the way plant buyers evaluate generation technologies. They are far more averse to investing in capital-intensive technologies. Meanwhile, projections for generating non-nuclear electricity predict substantial cost reductions over the next 20 years.
As recently as 1995, industry studies concluded that new nuclear plants would be competitive with coal and natural gas-based alternatives in the US market if life cycle costs were below 4.3 cents/kWh. This became the central economic goal of the ALWR programme. (It is worth noting that, at the beginning of the ALWR programme, it was projected that new nuclear plants would compete against coal-fired plants with life cycle costs near 7.9 cents/kWh.)
New studies now show that to be competitive in the long term (10 to 20 years), new nuclear plants in the US may need to produce electricity at less than 3.0 cents/kWh, assuming that the government does not step in and impose a carbon tax (or other disincentive) on fossil fuels.
O&M and fuel costs for existing nuclear plants have declined dramatically in recent years, as power companies began preparing for competitive markets, and are now quite competitive with other generation sources. Therefore the major impediment to long term competitiveness is the capital cost component, which may need to be reduced by 35% or more. Achieving such an ambitious goal will require a fundamental re-evaluation of the industry standards and regulatory bases under which nuclear plants are designed and licensed, as well as the introduction of advanced technologies.
RISK-INFORMED REGULATION
In the past, it would have been unthinkable to suggest that the NRC and industry carry out a fundamental re-evaluation of industry standards and regulatory bases for future plants. But the NRC and industry are now making great progress in streamlining the regulatory process for operating reactors and implementing risk-informed regulation to modify or eliminate requirements that increase operating costs without increasing safety. (Risk-informed regulation is the use of probabilistic risk assessment (PRA) as a tool to evaluate the effectiveness of existing deterministic requirements.)
There is a fundamental change in culture going on at the NRC, but so far 99% of this change has been focused on operating plants. The NRC and industry will need to look at expanding the use of risk-informed regulation to re-evaluate the approach for licensing new nuclear plants. PRA could then be used as a design tool to determine how systems and structures can be simplified, while maintaining the reliability of the original design.
CATCHING UP WITH THE COMPUTER REVOLUTION
The revolution in computer technology has dramatically changed the way in which industries do business. But compared to other industries, the nuclear industry has not fully applied computer technology – because of a lack of new plant orders in the US and the regulatory hurdles that must be overcome to implement any rapidly advancing technology.
Computer technology can now be used to make equipment in future nuclear plants smarter (eg pumps and valves with self-monitoring, self-diagnostic features built in). Smart equipment would be more reliable and improve the safety of the system in which it is used. The impact on nuclear plant design could be dramatic. By addressing the question of reliability at the root-cause level, it should be possible to re-evaluate the need for so much redundancy, diversity and quality assurance. In the case of quality assurance, it may be possible to adopt standards from other industries, which have improved substantially in recent years. A graded quality assurance approach may be appropriate. The resulting capital and operating cost savings could be dramatic. Best of all, they could be achieved without sacrificing safety.
As well as being included in the plant design, computer technology can be applied to all stages of design, fabrication, construction, operation and maintenance. For example, a computer-based information management system (IMS) with open architecture could be developed and used, industry-wide, by designers, plant owners, equipment suppliers and regulators. A standard IMS would allow plant designers in different companies (even in different parts of the world) to work together on a design – instantly sharing information as it is being developed. The costs and schedule for managing, sharing, and providing quality assurance of design data would be dramatically reduced. Very importantly, the standard IMS would allow plant buyers and operators to later have complete access to design information. Even NRC regulators could have access to the databases.
Computer technology can also be applied to construction. Besides actions to reduce the cost of each construction activity, anything that shortens the overall schedule will provide substantial savings in interest charges. A 3D model of the plant will be constructed on computer, with problems (eg hardware interference) identified and corrected before construction begins.
During the design phase, the model will provide feedback to identify design changes that would reduce the construction schedule and cost. Critical path scheduling for construction activities will be developed and maintained on computer, helping plant constructors to recover from unanticipated delays.
NUCLEAR ENERGY RESEARCH INITIATIVE
A US Department of Energy (DOE) R&D programme, the Nuclear Energy Research Initiative (NERI), is already funding 46 R&D projects that are seen as a chance to generate new ideas for nuclear energy. Three of the largest projects will lay a foundation for a revolutionary plant design effort that includes the issues above.
The three projects (see NEI, December 1999, p35) are in the following areas:
• Risk-informed assessment of regulatory and design requirements for future nuclear power plants.
• Smart equipment and systems to improve reliability and safety in future nuclear plant operations.
• Development of advanced technologies to reduce design, fabrication, and construction costs for future nuclear power plants.
These projects are intended to begin laying a foundation for larger scale programmes by laying out methodologies and building an infrastructure. A larger, more comprehensive programme will be required to actually change the regulatory basis.
In the next several years government and industry must fundamentally re-think the process of designing, licensing, and constructing new plants. We must make use of the new computer tools that have been employed in other industries We must systematically review all the industry standards and regulatory requirements and consider what changes need to be made – even if legislation is required to revise regulatory requirements and processes.
THE NEED FOR AN R&D ROADMAP
Although the three DOE NERI projects provide a good starting point, there is a need to develop a science-based technology road map for identifying and developing all of the technologies needed for deployment of the next nuclear plant in the US. The result would be an overall R&D plan (including budget and schedule) for the work needed over the next few years to develop the processes and technologies needed to build a competitive new plant.
A fundamental re-evaluation of design and regulatory criteria and advanced computer-based technologies will require the consideration of hundreds of new technology applications in nuclear plant design. The complexity and interactions of these applications mean that the process for evaluating and selecting them must be rooted in a systematic, computer-based technique – eg system dynamics modelling.
Too many nuclear plant design requirements have resulted from subjective judgements, considering single issues. If plant designers are to eliminate the costly, unnecessary over-conservatisms that have been built into regulatory and industry requirements, it is essential that they have the proper tools, not only to systematically evaluate the alternatives but also to understand their interactions. For example, ten different design improvements, looked at independently, might indicate cost savings. But if they are interdependent, the savings may be substantially lower or higher. Just as important, considering these interdependencies might show that only a few of the ten improvements provide ninety percent of the savings. System dynamics modelling of the technology evaluation process would allow interdependencies to be evaluated.
During preparation of an R&D roadmap, input should be sought from other stakeholders, industries and countries. Public workshops should be held, to get input from stakeholders, including special interest groups, from the beginning of the project. Continued openness to all interested groups, from its very beginning, might be combined with limitations on the ability to delay the licensing processes later. This would require fundamental changes for government and industry, not just in the methods, but in the culture of the design, licensing, and construction processes.
Once again, modern technologies can help. Design information could be placed in the public domain and published on a website (with firewalls where necessary). This would not only make the licensing process transparent, it would allow submissions and alterations to be dealt with quickly. Voluminous paper submissions to the regulator could be replaced by freely available and searchable website information.
IT’S A RAPIDLY CHANGING WORLD
The pace of change in technology and business is much greater than it was even a decade ago and nuclear energy development must accelerate, if it is to keep up with its competitors.
Meanwhile, a streamlined, risk-informed regulatory infrastructure must be developed that allows simplified new plant designs to be developed and licensed. This too will require R&D: for example, elimination of the double-ended guillotine pipe break as a design basis event may require new test programmes, new analytical modeling techniques, and new approaches to monitoring piping systems.
The need for nuclear power to become competitive is recognised throughout the industry. Recently the International Atomic Energy Agency set up a consultancy group to consider new approaches to provide safety in new nuclear plants in a more cost-effective manner. The DOE recently held an international workshop on R&D collaboration to address the needs for next-generation nuclear plants.
The target is clear: the nuclear industry must offer competitive power – and that means 3 cents/kWh.