Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 1
The National Energy Modeling System Executive Summary In response to a request from the U.S. Department of Energy (DOE), in the summer of 1990 the National Research Council (NRC) established the Committee on the National Energy Modeling System (NEMS) to advise DOE and the Energy Information Administration (EIA) on the development and application of a modeling system to support energy policy analysis and strategic planning. As discussed later, this system should have the capability to simulate the effects of various energy policies on U.S. energy supply and demand including economic, environmental and national security impacts. In this report, the committee offers its findings and recommendations on the process of developing a NEMS, the architecture and data needs of the system, and the organizational and management actions that will, in the committee's opinion, help make the system work. The executive summary identifies the most important of these findings and recommendations. The committee was briefed by DOE and EIA on their analytical and modeling activities in policy analysis for the National Energy Strategy (NES). The committee was also briefed on DOE and EIA modeling capabilities that might play a role in future NEMS development. In addition, the committee reviewed state of the art energy modeling and analysis through current publications and invited presentations by individuals and public and private organizations engaged in such work. The report and its findings and recommendations rest on this information and the committee's deliberations. Principal findings of the committee are as follows:
OCR for page 2
The National Energy Modeling System PRINCIPAL FINDINGS The set of EIA models reviewed by the committee at the beginning of this study constitutes a reasonable starting point for developing a National Energy Modeling System. However, considerable development will be needed to attain a modeling system satisfying the requirements outlined in this report. The committee believes that modeling can play a valuable part in energy policy analysis. In its first advisory report issued January 30, 1991, the committee observed that DOE has no comprehensive model or set of models that can respond adequately to the needs of the NES but “the approach taken by DOE in using available models, and off-line supplemental analysis as necessary, was a rational response to the department's need for expedient support of the NES process” (NRC, 1991a; reprinted as Appendix B to this report; also see Appendix E). The committee noted, however, that the set of models used by DOE in the NES had “significant limitations relative to the analytical results reported by DOE,” and that one needs “to appreciate the limited power of the existing set of models used for evaluating policy choices.” Evaluation of the NES effort thus indicated that the NEMS must significantly exceed the capabilities of existing DOE models. The committee recommends a course of action that builds on existing capabilities. The NEMS Program, once established, should complement and interact with a variety of other public and private groups that contribute to policy analysis. The development and use of models is only one aspect of policy analysis. Policy analysis also includes the identification of policy issues and initiatives of importance, development of assumptions, validation of models and data, interpretation of model results, and attention to social values. Additionally, the capabilities of the NEMS will complement many credible public and private models and analysts. Successful development of the NEMS will require the Secretary of Energy and EIA Administrator to establish and foster an organizational environment that is outward-looking and ensures greater intellectual and institutional commitment to its development and maintenance. After more than a decade of budget stringency and less than full support by the DOE at large, DOE/EIA “culture” has clearly not moved forward. For the NEMS to fulfill its potential as a national energy modeling and data resource system, DOE and EIA must create a supporting institutional environment that is sufficiently rich, broad, and interactive, that relates well to the DOE's policy and program offices and to appropriate federal, state, regional, and nongovernment groups. The committee's principal recommendations follow here; additional recommendations are offered throughout the report. The committee's goal is to lay the foundations for
OCR for page 3
The National Energy Modeling System creating a NEMS and associated institutional environment to support energy policy analysis that is objective, useful, and of the highest professional standards. The recommendations are classified by three major issues for the NEMS: the timing of its development, its management, and its design. PRINCIPAL RECOMMENDATIONS. Timing of NEMS Development 1. The EIA and DOE should move quickly to configure an initial National Energy Modeling System within the next one to two years and apply it to policy issues, including the next National Energy Strategy. The most useful and cost-effective approach to begin the development of NEMS is to build on existing models, modified as appropriate, to address the most important national energy policy problems. Using this approach, NEMS can be implemented quickly and provide analysis for decisions in the next NES. NEMS does not need to be fully developed to be a valuable near-term tool. The recommended structure of NEMS will facilitate later system improvements by allowing the substitution or addition of new or modified modules. Through constant use and this evolutionary approach, NEMS will become increasingly more comprehensive, accurate, and useful. The committee is concerned that if the pace of NEMS development doesn't allow its early use, then an important opportunity will be missed to improve the analysis of energy policy issues important to DOE. If priority is given to NEMS in the allocation of budgets and personnel, an initial system can be configured and applied in the next NES exercise. By virtue of its application, the system will receive greater peer review and undoubtedly will be changed and improved and expanded for future use. NEMS Management 2. The Secretary of Energy should designate the EIA Administrator as the chief executive for the implementation of the NEMS, and make this assignment one of the main performance requirements for the Administrator. By law, EIA was established within DOE as an independent, nonadvocacy group to collect, analyze, and report data for the federal government. (In contrast, DOE's Office of Policy, Planning and Analysis advances the philosophy of the Department.) Thus, EIA is insulated from the vagaries of the policy process and political pressures to which the rest of the DOE is exposed. EIA is traditionally looked to for impartial analyses of energy supply and demand, for example, as reported in their Annual Energy Outlook. Such independence helps confer greater objectivity and credibility to EIA activities. For the successful development of NEMS, the committee believes that the Secretary needs to vest the responsibility for its development and implementation directly with the EIA
OCR for page 4
The National Energy Modeling System Administrator. The Administrator should be directly responsible because the development of NEMS will affect most other agency activities, will require new functions, and will overlap with functions of other DOE offices. 3. Toward keeping NEMS outward-looking and sensitive to relevant policy issues, EIA should form a Users Advisory Council of likely NEMS users from within DOE, other federal, state, and regional agencies, and private organizations. EIA should develop and manage the NEMS with the intensive involvement of this advisory group. User advisory groups are vital to the effective functioning of NEMS. An important use of NEMS will be for energy policy analysis by those outside of EIA. To help ensure the responsiveness of NEMS to users' needs, EIA must maintain close working relationships with potential users, both inside and outside DOE, with experience in policy analysis. Various federal and nonfederal organizations have significant data resources and modeling capabilities. These groups, resources, and advice should be sought and used well in NEMS development. 4. DOE should take action to attract and retain highly skilled professionals for the design, development and implementation of the NEMS. The initiation and sustained development of NEMS will require talent. The models developed will continually require updating and maintenance. NEMS will be valuable only if supported on a continuing basis with adequate budgets and skilled personnel. The human resources now devoted to NEMS are seriously limited, especially the number of senior personnel. NEMS Design 5. NEMS should be designed to estimate and display as primary outputs the economic, environmental, and national security implications of alternative energy policies, with priority given first to the mid-term time horizon and then to the longer term. NEMS should produce economic, environmental, and national security measures relevant to the analysis of public policy. Decision makers need more information than energy quantities and prices over time. Economic outcomes include consumers' and producers' surpluses, gross national product (GNP), and the federal budget deficit. Environmental measures include the direct and indirect emissions of major pollutants and their effects. Security measures include the economic impacts of energy supply interruptions. Regional disaggregation and equity measures are also relevant to some policy analyses. Clear accessible graphical outputs are also essential to users. Three roughly defined time horizons are appropriate for the analysis of energy policy issues: the short-term (roughly up to two years), mid-term (up to about 25 years), and longterm (beyond 25 years). The committee believes that NEMS should initially focus on developing mid-term analytic capability: this time horizon is of greatest relevance for most
OCR for page 5
The National Energy Modeling System important national policy issues. Priority should then be placed on the development of long-term models, which do not now currently exist at EIA. 6. NEMS must incorporate the behavioral and policy-driven aspects of decisions about energy use, fuel and technology choices, and energy supply investments. Many of the important energy policy issues for the foreseeable future will concern energy consumption and policy options to influence energy consumption. To evaluate the potential effectiveness of economic incentives and disincentives and regulations, the nature of relevant decision making processes must be reasonably reflected in the models. To a very great extent, decisions about energy consumption and even about investments in energy supply are now being described in models based only upon the surmise of the modeler. Resources should be devoted to capturing available descriptive information on the behavior of energy users and suppliers in model structures and to supporting related research and collection of further information. 7. NEMS should be designed to represent and analyze the effects of uncertainty explicitly. Inherent with energy modeling are many types of uncertainties, which are often not sufficiently addressed in policy analysis. By characterizing such uncertainties and their consequences, NEMS can help to identify and manage the risks associated with energy policy alternatives through the design of insurance or hedging strategies. Two major types of uncertainties that should be considered initially in NEMS are (1) uncertainties in model input parameter values, which determine projections of future energy use (e.g., parameters describing economic growth, energy prices, and technological change); and (2) uncertainties in the basic relationships conjectured between energy use and economic or behavioral activity (e.g., the structural or causal relationships assumed by NEMS components). To address the first type of uncertainty, sensitivity analysis and scenario studies should be performed to assess the implications of alternative assumed situations on key model output measures. To address the second type of uncertainty, NEMS's modular structure can be exploited to analyze the implications of alternative model formulations for various NEMS components (including those developed outside DOE and EIA). Over the longer term, other types of uncertainty analysis should also be incorporated in NEMS. For example, the development of reduced-form modules (simplified representations of more complex model components) will facilitate the use of probabilistic methods that allow uncertainties to be expressed as probability density functions describing the likelihood of particular outcomes (in response to assumed uncertainties in model inputs). Such methods already are in use (e.g., by the Northwest Power Planning Council) and have provided valuable insight. They have also been used in research and development planning.
OCR for page 6
The National Energy Modeling System 8. NEMS should be designed to be modular in structure. The system should readily accommodate substitution of alternative models (modules). The committee proposes a modeling system that would consist of a set of modules linked together by a simple control module. The modules should be designed so that they can be run separately, all together, or in combinations, depending on the analytical need. Each component module would provide the richness of detail and diversity of structure appropriate to the sector it represents. The control module would provide a simple set of switches to govern the operation of the component modules and an algorithm to achieve equilibrium of supply, demand, and price. The structure proposed represents a logical continuation of modeling development within EIA. Modular architecture would facilitate decentralized development and maintenance of the modules. Once the inputs, outputs, and interfaces between models are appropriately defined and specified, then development of the modules could proceed in a decentralized manner. Modular architecture would allow the substitution of alternative modules embodying different conceptual structures, theories, empirical representations, and data bases to examine some of the sources of forecast uncertainty. This would allow the use of existing models from groups outside the DOE and EIA, especially with regard to demand-side modeling. As mentioned in the discussion above on uncertainty, reduced-form versions of the component modules should be developed that will simplify the more complete modules by approximating the relationship between input variables and output variables. Reduced-form modules would increase the transparency of the full modules by allowing users to readily perceive their aggregate behavior. The committee believes that at least the reduced-form version of NEMS should be configured to run on microcomputers or personal computers.1 If the system could run on widely available hardware configurations, such as personal computers, the number of analysts who could use the models would be greatly expanded. Such access would increase the numbers who understand, use, and test the system. This hardware restriction is essential to achieving the goals of transparency and quick turnaround. Alternatively, but less desirable, NEMS could be configured to run on other widely available hardware configurations (e.g. workstations) if it were coded using a machine-independent computer language. 1 The committee uses the terms “personal computer” and “microcomputer” as generic terms for small computers that are widely available to a variety of users. They could include computers made by IBM or others using the DOS software, Apple, NEXT computers, or any other widely available computers.
OCR for page 7
The National Energy Modeling System 9. NEMS architecture should provide for quick turnaround. EIA procedures must support this rapid response capability. Although the EIA will develop, manage, and maintain NEMS, and use it for its own purposes whenever appropriate, the DOE Office of Policy, Planning and Analysis and program offices will need to access the system independently. These offices must run some analyses very quickly (in hours rather than days, including hard copy output). Recognizing tradeoffs between modeling level of detail and run time, analysts can use the reduced-form models mentioned earlier extensively in this process. The committee suggests that for such policy analysis, computer outputs could be generated without an EIA imprimatur; only the results of final runs might be reviewed through the full EIA administrative process before their release in policy documents or reports. 10. EIA should create a group to develop a long-term modeling and analysis capability. The mid-term NEMS modeling effort that the committee recommends be implemented immediately should not simply be extrapolated for long-term modeling (for the time horizon beyond 25 years). Instead, the EIA should create a distinct group to develop long-term modeling and analysis capability. In addition to relating this long-term capability to the medium-term modeling system in a consistent fashion, this group should avail itself of expertise of external individuals and groups who already focus on long-term energy modeling. This effort should be initiated quickly. However, the actual development or acquisition of long-term modeling capability should be given lower priority among the tasks recommended for the NEMS effort. 11. A major effort is needed to collect more extensive data and information on the U.S. energy system, especially on end use. Accurate and relevant data are needed to support the development of models for the NEMS. But current EIA data efforts lean too heavily toward mid-term supply-side data. Considerably more demand-side data collection is needed, as is the collection of long-term supply-side data. Use of DOE and EIA data bases should be supplemented by the use of data bases from other federal agencies, state and regional governments, and private organizations. Where appropriate, existing EIA surveys could also be modified to collect new kinds of data without major additional effort. Finally, new data-collection efforts on a one-time or continuing basis may be needed to meet reasonable modeling requirements. A major effort is needed to collect disaggregated data on (1) the characteristics of both existing and new energy-using equipment; (2) the stock of energy-using equipment; (3) energy use by “service” category (e.g. space heating); (4) activities that underlie the various categories of energy use, such as the number and types of housing units in use; and (5) the cost and savings of technologies that can be implemented or developed to improve energy-efficiency (in modeling, “supply curves of conserved energy”). These data and the models that employ them need to be adequate to evaluate demand-side policies, such as schemes to encourage investment in efficient equipment or regulations mandating efficient performance of equipment. Such data and models are not adequate at present.
OCR for page 8
The National Energy Modeling System To support longer term energy modeling, new approaches and data are needed to (1) assess undiscovered resources in DOE resource assessments, (2) appropriately account for the large unconventional oil and gas resources into the assessment, (3) estimate the effects of continued evolution in extraction technology and production costs, and (4) evaluate the future contributions of coal, nuclear, and renewable energy, taking into account their economics and environmental implications. Data requirements for renewables are different from those for fossil fuels. The difference between discovered reserves and undiscovered resources has no parallel for renewable resources. The modeling problem for renewable resources is lack of data on early adoption potential in small low-cost resource pockets and small high-value market niches. The nation will be spending several trillion dollars over the next thirty years at satisfying our increasing demands for energy services. Federal investments in technologies for both supply and end use will be evaluated based not only on cost effectiveness but also on environmental and national security concerns. If a better modeling system can guide energy investments in a way that addresses environmental and national security concerns and reduces costs by only a percent, it will have paid for itself a thousand-fold. The remaining chapters of this report lay out the requirements, a proposed architecture, and a strategy for management that the committee feels will result in a modeling system that will respond to the nation's needs.
Representative terms from entire chapter: