January-March 1999 Issue


Reducing Electric Power Emissions:
Are Kyoto Targets Reachable?

hat mix of policies and technologies will let us produce and use electricity while protecting environmental, economic, and social health for future generations? An international team led by Energy Laboratory researchers has developed a methodology to help decisionmakers address that question. The outcomes of initial case studies are sobering. The methodology combines engineering, cost-accounting, and life-cycle assessment techniques to study potential strategies for specific regions. In each study, an advisory group of governmental representatives, utility executives, business people, and special interest groups provides information on regional needs and constraints and ultimately identifies the strategies that best meet their diverse interests. Studies of New England and Switzerland examined strategies involving expanded use of natural-gas combined-cycle generation and renewable technologies as well as improved end-use efficiency. Even the most aggressive strategies produced carbon dioxide (CO₂) emissions well above the levels prescribed by the Kyoto Protocol. The researchers conclude that achieving substantial and sustained reductions in CO₂ emissions will require changes in the basic electricity supply and demand infrastructure. Target strategies include the use of distributed generation systems including small cogenerators, more efficient transmission systems, novel building designs, and "smart" meters that control electrical devices to maximize operational efficiency. Also critical is the transfer of new technologies to nations whose energy infrastructures are expanding rapidly, especially developing nations. Finally, control of CO₂ emissions must be coordinated with the reduction of other emissions that are more important at local and regional levels.

Most people agree that energy supply and demand systems should be "sustainable." Energy policies and technologies should preserve the environment while supporting economic development and social welfare in both industrialized and developing countries. According to Stephen R. Connors and his colleagues, such broad definitions of sustainability are unlikely to bring about real change because they fail to connect with the more immediate interests of those who can actually implement strategies for electric development. Real change usually occurs in response to near-term or occasionally longer-term regional or local considerations. If climate changes, will our farms experience more frequent floods, droughts, or hurricanes? How many of our neighbors will become unemployed? The first step in developing robust sustainable energy practices is to identify region-specific impacts, needs, and constraints. The next steps are to define cost-effective, responsive, and realistic energy strategies and to communicate them to the relevant policymakers.

Carrying out those tasks is the goal of a new project called Strategic Electric Sector Assessment Methodology under Sustainability Conditions, or SESAMS. SESAMS has been developed by a multidisciplinary team that is led by Mr. Connors and involves researchers at MIT, ETH-Zürich, the Ecole Polytechnique Federale de Laussane, and the Paul Scherrer Institut. The focus is on electricity because it is cleaner at the point of use than is energy produced directly from fuel combustion and because it is versatile: it can power devices ranging from computers and air conditioners to industrial equipment and vehicles.

SESAMS is designed to determine how different strategies for generating, transporting, and using electricity will affect costs, emissions, and other consequences in a given region. Each strategy consists of a variety of technologies, fuels, conservation programs, policy actions, and so on. The SESAMS approach differs from traditional approaches to utility planning in several ways. It uses engineering models in conjunction with economic models, but the economic and environmental analyses are regional or local--a challenge because statistics for critical factors such as economic growth, electricity production, and pollutant emissions tend to be available for nation- and industry-wide groupings rather than for regional ones. Linked with the engineering and economic models are life-cycle assessment (LCA) models that calculate the emissions, resource consumption, and environmental damages associated with a given strategy. Thus, the LCA models take into account the electricity consumed in extracting and refining fuels, building power plants, constructing factories for making high-efficiency light bulbs, and so on.

Finally, the SESAMS approach does not weigh or prioritize different impacts, for example, by placing dollar values on specific emissions or environmental effects. The priorities of various groups of "stakeholders" involved in policy decisions differ, making such weighting schemes inappropriate for the intended policy dialogue. SESAMS therefore permits the stakeholders themselves to incorporate their interests by involving them in the studies directly--an approach that Energy Laboratory researchers developed a decade ago in electric utility planning studies of New England (see e-lab, April-September 1989). Each SESAMS case study involves an advisory group of environmental and economic regulators, business people, and members of environmental and other interest groups. The group's primary goals and constraints--economic, political, environmental, and social--are thereby incorporated into the analysis. With each new set of scenario analyses, the researchers present the advisory group with a graphical representation of the relative costs, emissions, and other implications of the strategies. The stakeholders then work together, discussing the tradeoffs among costs, emissions, and other benefits to identify the strategies that best meet their collective needs.

The researchers have performed several case studies that examine changes in technologies, policies, and practices and their impacts on long-term emissions in well-defined geographical regions. One major study focused on New England. Assisted by an advisory group, the researchers used state-of-the-art engineering models to simulate the operation of the regional power system for thousands of combinations of technology types, environmental regulations, fuel costs, and other factors. Eight of the most interesting strategies, each involving plans for generating the needed electricity and measures to reduce electricity demand, are described here. (Life cycle assessment was not included in the New England project.) The simulations run from 1995 to 2014 and start with the 1995 energy mix used to generate electricity in New England--28% from baseload nuclear power, 16% from coal generation, 40% from oil and natural gas, and the remainder from a mix of hydropower, imported power, and other sources.

The demand-reducing measures used in the strategies involve demand-side management (DSM) programs--utility-implemented programs aimed at increasing the energy efficiency of their customers' facilities. Four levels of DSM are considered. The baseline strategy assumes that current DSM programs continue, cutting electricity consumption by 8% over the 20-year study period. Three other strategies call for increasingly aggressive DSM programs that result in reducing consumption by 15%, 21%, and 26% over the 20-year period. These higher levels of DSM are phased in over time and then stopped so that their longer-term impacts can be assessed.

Two methods of adding new generation are used in the strategies. The first plan uses existing generating capacity to the extent possible and meets any need for additional power using natural-gas combined-cycle generation, a relatively high-efficiency, low-emissions electricity source. The second plan uses the same approach but adds renewable power to the generation mix: 1400 MW of wind generation is phased in during the first ten years of the study period. The result is a total of eight strategies, four DSM levels with and four without windpower. It is important to note that these eight strategies do not represent forecasts of what might happen but rather of what is "technically feasible" -- a distinction that must be emphasized, as discussed by Henry D. Jacoby, William F. Pounds Professor of Management (see The Uses and Misuses of Technology Development as a Component of Climate Policy, MIT Joint Program on the Science and Policy of Global Change Report No. 43, http://web.mit.edu/globalchange/www/rpt43.html). Issues relating to the economic and political feasibility of various strategies as well as policies to promote their adoption must be considered by working with the stakeholder advisory group.

The first step in determining the impacts of the strategies is to calculate the resulting electricity demand under each of the four levels of DSM. The figure below shows the outcome. The top curve shows the impact of the reference strategy: demand increases steadily at a rate of about 1.3% per year. The three lower curves show the effects of the increasingly aggressive DSM strategies. Electricity demand tends to be relatively flat or declining during the first part of the study period, but each curve takes an upward turn when the DSM programs end.

The next step is to calculate CO₂ emissions resulting from meeting those four levels of demand using the two generation plans (with and without windpower). The figure below shows the projected levels of CO₂ emissions resulting from implementation of each of the four strategies excluding windpower. In each case, adding windpower lowers the curve shown here by about 2 percentage points. The horizontal line indicates emissions levels in 1990 -- the historical baseline used in the Kyoto Protocol.

Under the "business-as-usual" reference strategy, CO₂ emissions rise steadily; and by 2014 they are fully 80% higher than the 1990 levels. This increase in CO₂ emissions is caused largely by the long-term growth in electricity demand seen in the first figure. Two events exacerbate the growth in emissions: in 2001, contracts for Canadian hydropower expire; and between 2003 and 2013 aging nuclear units--the source of half the total nuclear power--are retired. Those non-CO₂-producing sources are replaced by new and existing fossil-powered plants, leading to further increases in CO₂ emissions. (These results are somewhat optimistic as the trend in the region has been to retire some of these nuclear units prior to the end of their operating licenses.)

As expected, the more aggressive DSM strategies do better. Those strategies are able to reduce industry-wide CO₂ emissions to or below 1990 emissions--but only temporarily. As soon as the DSM programs are phased out, CO₂ emissions begin to rise. By 2014, the double and triple DSM strategies are 60% and 40% above the 1990 levels. Even the quadruple-DSM strategy (with wind included) is about 25% over 1990 levels--well above the 7% reduction called for in the Kyoto Protocol.

The difficulty of designing a CO₂-reduction strategy is illustrated by one observation: as nuclear units are shut down, CO₂ emissions increase more rapidly under the more-aggressive DSM strategies than under the reference case. In the reference case, demand for electricity grows rapidly; and new natural-gas-fired power plants must be built. Generation from nuclear plants is replaced using natural gas plants, pushing up CO₂ emissions. But the outlook is even worse under the more-aggressive DSM strategies. Demand grows more slowly, so less new natural-gas-fired generation is built. And when nuclear generation is retired, old lower-efficiency, higher-carbon-content plants are used to fill the gap. Overall CO₂ emissions therefore increase faster in later years. Emissions such as nitrogen oxides and sulfur dioxide can actually become greater in the more-aggressive DSM strategies than in the reference strategy.

Another case study using SESAMS is examining strategies for the electric power industry in Switzerland. To perform this study, the MIT researchers teamed up with their colleagues at the Paul Sherrer Institut and ETH-Zurich, who contributed both detailed knowledge of the Swiss situation and sophisticated new analytical tools for assessing the life-cycle energy and emissions implications of the strategies studied. Results from the Swiss study are as discouraging as those from the New England study. Switzerland's electricity now comes largely from hydropower and nuclear plants, so CO₂ emissions are already low. But the SESAMS analysis concludes that future increases in CO₂ emissions are inevitable because natural-gas-fired generation is the most likely candidate to replace the retiring nuclear plants. Strategies combining dramatic increases in end-use efficiency and the use of substantial forestry biomass generation could hold emissions down for several decades. But by 2019 demand will be sufficiently high that natural-gas-fired generation is required, and CO₂ emissions will begin to climb.

The researchers are now continuing their SESAMS case studies. They are expanding the Swiss analysis by including nuclear plant retirement, growing reliance on imported power, and increasing competition with Western European electricity markets. They are also undertaking a major study of electric power strategies for the province of Shandong, China, a region with rapidly growing electricity demand and fewer choices for electricity generation.

Based on their work so far, the researchers conclude that achieving significant, sustained reductions in CO₂ emissions will be difficult. No single technology or class of technologies will do the job; and efficiency improvements cannot be assumed to continue indefinitely, on either the supply or the demand side. The only way to achieve the needed reduction is by improving the entire energy supply and demand infrastructure, including the development and deployment of many new cost-effective technologies (see article below). Such an undertaking requires close coordination among policymakers, technology developers, energy service providers, and the public--a difficult task, especially without the leadership provided in the past by the large utilities.

Stephen R. Connors is a research scientist in the Energy Laboratory and, since 1989, director of the Laboratory's Analysis Group for Regional Electricity Alternatives (AGREA). The New England case study, which ran from 1988 to 1996, was funded by a consortium of New England utilities and the National Renewable Energy Laboratory. The ongoing Swiss and forthcoming Chinese case studies are supported through the Alliance for Global Sustainability, a partnership of MIT, the Swiss Federal Institutes of Technology, and the University of Tokyo. Further information can be found in references.