Nuclear exit,the US energy mix,and carbon dioxide emissions

Abstract

If the United States were to adopt a policy to phase out nuclear generation, as has happened recently in other developed countries, what would the environmental and energy-mix implications be? Based on alternative scenarios of nuclear exit that consider the influence of potential policies to limit greenhouse gas emissions, a model of the US and global economy indicates that, under current policy, a US nuclear exit would increase carbon dioxide emissions, and likely raise electricity prices and reduce gross domestic product by relatively small amounts. Those economic impacts would be increased by additional measures to limit carbon dioxide emissions.

A number of issues have combined to stymie nuclear plant investment in the United States, importantly including its high per-kilowatt-hour cost compared to coal (MIT Energy Initiative, 2009). But even with no new plants starting operation since the mid-1990s, license extensions and upgrades in the capacity of existing units have maintained nuclear generation in the neighborhood of 20 percent of US electricity supply for the past quarter century. In the last decade or so, there was talk of a nuclear renaissance, in the United States and elsewhere, stirred by the expectation of penalties on carbon dioxide emissions, improved capacity factors and safer designs, and (in the United States at least) more supportive federal regulations and subsidies.

But then came Fukushima. Whatever one’s view of the reality of the earlier renaissance, the accident appears to have substantially dimmed the future of the industry. ¹ The effect on Japan has been particularly dramatic, creating difficulty in restarting units down for maintenance at the time of the accident or shut in its aftermath and initiating an on-and-off policy of complete phase-out by 2040. There also has been a strong response in Europe—notably in Switzerland, Italy, and Germany, which have imposed phase-out plans—and there is even talk of a partial phase-out in France (France 24, 2012).

Reaction in the United States has been muted. A Gallup poll after the Fukushima accident found that US views of nuclear safety had not changed since a similar assessment a decade earlier (Gallup, 2011). And although there is continued opposition to the approval of new plants or the extension of operating licenses for existing ones, there is no national plan to force a US exit from nuclear generation, nor is any such proposal widely discussed in policy circles. Still, given the evolving attitudes in other developed countries, it is useful to consider the implications should support for such a policy emerge. Our simulations of various scenarios of the US nuclear future consider the context in which an exit might occur—with a particular focus on potential measures to limit greenhouse gas emissions—but deal with only a subset of the consequences. Even our results are sensitive to input assumptions—about natural gas prices, emissions limits on new coal-fired power plants, and subsidies for renewable energy sources, for instance—that are not explored here. Considering these uncertainties, however, it is clear that a US exit from nuclear generation would impose costs on the environment, electricity consumers, and the national economy.

US nuclear exit scenarios

The implications of a policy-driven US exit from nuclear electricity are widespread. On the one hand, nuclear supplier firms will feel the effects through lost business opportunity, and workers in the nuclear industry will see a progressive disappearance of their jobs. On the other hand, demand for competing technologies and their supporting industries and specialized labor will increase. While there will be benefits from the reduction of both the risks associated with nuclear generation and the difficulties of radioactive waste management, there will also be increased costs from the non-climate-related environmental effects of replacement technologies like natural gas and coal—plus an increase in the cost of electricity.

Though there are no official proposals on the table, three hypothetical nuclear futures illustrate the implications of varying governmental policies on nuclear power.

The first such future assumes that the country does not pursue a nuclear exit. The US nuclear fleet currently is composed of 104 units with a total rated capacity of 102 gigawatts. In this base case, this system is sustained, and even expands somewhat, as projected by models of the US economy and its energy system. This case implies both further life extension of existing plants and some additional nuclear construction. By 2050, the fleet will have grown to 110 gigawatts, while falling to about 15 percent of generation in a growing US electricity sector. Through 2035, this projection is essentially the same as the reference case of the Energy Information Administration’s Annual Energy Outlook (Energy Department, 2012), and capacity is held constant at the 2035 level through 2050.

In a second scenario, the government would freeze nuclear licensing. Seventy-three US nuclear units have received license extensions, while 13 have filed for license renewal and another 17 are expected to apply (Nuclear Energy Institute, 2012). In addition, three nuclear units (Southern Company’s Plant Vogtle units 3 and 4 in Georgia and the Tennessee Valley Authority’s Watts Bar Nuclear Plant Unit 2) are under construction. For this case, no additional license extensions or uprates of capacity are granted to existing plants, and no new ones enter service beyond those in operation today. Figure 1 shows the decline in the US nuclear fleet under these assumptions. US nuclear capacity is reduced by some 60 percent by 2035, and is phased out entirely by 2048.

Figure 1.

Freeze and forced exit

The third scenario would have the government force a nuclear phase-out on an accelerated schedule. In this case, US nuclear generation is reduced to zero between 2015 and 2030, a profile also shown in Figure 1. The scenario is somewhat less draconian than current German policy, where nuclear currently plays a similar role in the power sector, which will phase out nuclear generation completely by 2022.

Although there are many possible impacts of each of these possible nuclear policies, the analysis presented in this article looks primarily at the effects of these policies on US emissions of greenhouse gases, the domestic price of electricity, and national economic performance, measured in terms of gross domestic product.

The analysis

Several assumptions about future conditions will greatly influence the projected results of a US nuclear phase-out. Here, we focus on one specific condition: the potential for a national policy penalizing the carbon dioxide emissions of the principal replacement technologies. To explore the range of outcomes of the ongoing debate over US climate policy, we examine three policy options: The current carbon policy, in which the United States adopts no additional national policies on greenhouse gas emissions beyond those already in place; ² a regulatory option, in which measures similar to some being pursued at present are imposed, including a national renewable energy standard mandating 25 percent of US generation by renewable sources by 2025 and retirement of 50 percent of US coal-fired generating capacity by 2030; and the adoption of a target of reducing US carbon dioxide emissions by 50 percent from the 2005 level by 2050, implemented by an emissions price (i.e., applying a tax or a cap-and-trade system).

These cases are close to those applied in analyses of US natural gas development (Jacoby et al., 2012; MIT Energy Initiative, 2011; Paltsev et al., 2011), which offer a basis for comparison with still other scenarios.

This analysis of nuclear exit scenarios applies the MIT Emissions Prediction and Policy Analysis (EPPA) model, a comprehensive, 16-region, multi-sector representation of the global economy (Paltsev et al., 2005). ³ The model identifies sectors that produce and convert energy, industrial sectors that use energy and produce other goods and services, and the various sectors that consume goods and services (including energy). Energy production and conversion sectors include coal, oil, and gas production, petroleum refining, and an extensive set of alternative generation technologies, including representations of conventional and advanced nuclear generation.

The economic impacts of a nuclear exit will be affected by assumptions about the costs of substitute generation technologies; if, for instance, natural gas prices are high, it would cost more to replace the electricity previously produced by nuclear plants than if gas prices were low. ⁴ US gas prices are now (and are projected to remain) substantially below those in Europe and Asia, which will lead to lower US exit costs. We apply the median gas supply costs for the United States, as analyzed in the MIT study of the Future of Natural Gas (MIT Energy Initiative, 2011), and we adopt costs of nuclear and non-nuclear generation technologies documented in that study.

All models have limitations, of course. Influential input assumptions—e.g., about population and economic growth, resource prices, and the ease of an economy’s adjustment to price changes and technology costs—are subject to uncertainty over decades. Also, some details of market structure and the behavior of individual industries are beneath this model’s level of aggregation. Considering these strengths and weaknesses, the results presented here should be viewed not as predictions conveying confidence in particular numbers, but rather as illustrations of the role of nuclear power in the US electric power system, and as a basis for forming intuitions about the direction and rough magnitude of the effects of policies determining its survival.

Environmental and economic consequences of a US exit

It’s axiomatic that an elimination of nuclear power generation will change the mix of generation sources of US electricity. How and when that mix changes will depend on a variety of factors. If, for example, current US national policy on greenhouse gas emissions remains unchanged until 2050, under the license freeze scenario the lost nuclear generation would be replaced mainly by natural gas, and to a lesser degree by coal. ⁵ With a forced exit, the same substitution occurs, only earlier in the study period, as shown in the left-hand column of Figure 2.

Figure 2.

The examined alternatives

Under the assumed regulatory measures, however (as indicated in the right-hand column of Figure 2), and with no nuclear exit, the mandated reduction in coal generation is only partially replaced by the mandated renewables, with the rest made up by natural gas. As will be explored in more detail below, the increase in electricity price caused by carbon emissions regulation would likely not be great; the change in electricity demand would therefore not be significant. With a license freeze, the lost base-load generation would have to be replaced, and as the figure shows, this replacement power would come mainly from gas-fired plants, with a contribution by renewables over and above the mandated level. With a forced exit by 2030, the outcome by the end of the period is the same, but the increases in other generation sources must come more quickly.

A target to reduce US carbon dioxide emissions to 50 percent below 2005 levels by 2050 will, even with no nuclear exit, drive conventional coal from the US energy system by 2045 and require a very large increase in electrical generation from natural gas and more than a doubling of the contribution from renewable energy sources. The resulting increase in electricity price will lead to an increase in the efficiency of energy use. With a nuclear freeze, the task requires still more significant changes in the generation mix. Even with a further increase in renewables, in 2040 the emissions from gas generation will begin to press up against the national emissions limit, and gas with carbon dioxide capture and storage will enter the picture. By the end of the period, coal with carbon capture will appear as well. With forced exit by 2030, these adjustments again must begin earlier, so that still more carbon capture and storage will be in place by 2050. Also, nuclear exit scenarios lead to even higher, price-driven improvements in energy efficiency.

The effect on carbon dioxide emissions

If nuclear generation continues to supply US electricity as it does under the current policy assumption, the nation’s carbon dioxide emissions are projected to increase. ⁶ With no additional climate policy, carbon dioxide emissions from electricity are projected to roughly stabilize over the coming decade, but then rise due to an increase in generation from natural gas and coal. Imposing the coal shutdown and renewables mandate projected in the regulatory measures case reduces emissions from electricity by about 50 percent below the 2005 level by 2050. But this result only relates to electricity generation; total US carbon dioxide emissions will initially fall slightly, but then start to rise and reach 2010 levels by 2035. The assumed regulatory measures affect only the electricity sector, and emissions from other sectors drive the increase in the total emissions.

Regardless of which scenario of climate policy is in play, a phase-out of nuclear generation will increase national carbon dioxide emissions. If current policy on greenhouse emissions is maintained, a freeze on nuclear licenses will increase US annual carbon dioxide emissions by about 430 million tons, or 5 percent, by 2050. Under the assumed pattern of a forced exit, the ultimate increase is the same, but this level is reached as soon as 2030, when the nuclear contribution to the national energy budget ceases. If nuclear generation is phased out in the context of the assumed regulatory measures on carbon emissions, the increase is smaller, as most nuclear generation is replaced by natural gas and renewables. National carbon dioxide emissions rise by about 150 million tons by 2050, ultimately about a 2 percent increase, with this level reached more quickly under the quicker rate of phase-out. In the forced phase-out scenario, additions to emissions are reduced after 2030 as electricity price increases make renewable generation more economic, and it expands beyond its 25 percent mandated level.

In all these cases, the increase in emissions comes mainly from additional natural gas or coal burning to replace diminished nuclear generation, but adjustments are being made outside the electricity sector, as well. Natural gas and coal prices respond to the changes in their use in electricity generation, leading to adjustments in other sectors of the economy and their carbon dioxide emissions.

The effect on electricity pricing

The phase-out of nuclear generation also will increase the price of electricity. Prices are projected to increase even with no nuclear exit. Assuming a continuation of the current policy on greenhouse emissions, the price will rise (in real terms) from 10 cents per kilowatt-hour to around 16.5 cents by 2050, in response to rising costs of fuel (mainly natural gas) and other inputs. Imposition of regulatory measures, driving out cheap coal and mandating more expensive renewables, will further increase the 2050 price to around 18.5 cents. Meeting the 50 percent national emissions-reduction target will drive the price even further, to more than 26 cents per kilowatt-hour, because of the growing requirement of carbon capture and storage for both gas and coal and the pushing of renewables into less cost-effective applications.

Under current climate policy, the effect on the electricity price of a nuclear phase-out is very small, because relatively cheap coal and natural gas can fill the gap it leaves. With policies that drive out coal or penalize the carbon dioxide emissions of power generation based on fossil fuels, the nuclear exit leads to an increase in price over and above what would otherwise exist under the particular climate policy. Under regulatory measures the price increase is about 1 cent per kilowatt-hour by 2050, as relatively cheap natural gas will replace the nuclear generation. If the 50 percent national emission target is imposed, the impact of nuclear exit is greater, adding some 2 cents to 2.5 cents per kilowatt-hour.

There are 126 million residential electricity customers in the United States, with an annual average consumption in 2011 of 11,300 kilowatt-hours, so an increase in electricity prices of 1 cent per kilowatt-hour yields a $90 increase in annual electric bills. It is reasonable to assume that most of the effects of this price increase will be passed forward, to show up in the cost of other goods. To consider the larger impact, therefore, note that, for 2011 US consumption of 3,860 billion kilowatt-hours, a cost increase of 1 cent per kilowatt-hour would yield to a national increase of up to $39 billion overall. Spread across 115 million households, a 1 cent increase would impose an additional annual cost of approximately $336 per household.

The German nuclear phase-out has been the subject of a number of studies. Any comparison of results is necessarily approximate, because different estimation methods are applied, and their conditions are different from the United States; the United States, for example, has cheaper gas, but Germany can import electricity from other European counties, and expectation about the carbon dioxide price in the European Union Emissions Trading System are built into all the German analyses. Still, it is worth noting that the results are similar, with the price effect of the German forced exit reaching in the range of 1 cent or less per kilowatt-hour (e.g., Fürsch et al., 2012; Knopf et al., 2012; Matthes, 2012).

The cost to the economy

If the US were to phase out nuclear power, there would also be some cost in terms of reduced gross domestic product. This cost is not quite the same as the increased expenditure on electricity, because the adjustment spreads across the economy, and the increased resources devoted to electricity in early years reduce investment somewhat, lowering economic growth in future years. As with the effect on electricity prices, the impact of nuclear exit on gross domestic product is very small if there is no carbon dioxide constraint beyond current policies. However, exit will impose a burden on the economy if climate policies are put in place. If the regulatory policies are adopted, the cumulative GDP loss for the period of 2015 to 2050 is projected to be around $250 billion for a license freeze, and around $500 billion for forced exit. ⁷ Under a 50 percent target for emissions reduction, the cost would be higher: around $500 billion for the license freeze and around $800 billion for a forced exit.

Despite uncertainty, there will be a cost

A US exit from nuclear electricity generation could have many consequences; our analysis illuminates only a subset of them. And even these results are sensitive to input assumptions not explored here. Natural gas prices higher than our projection would increase the cost of exit, as would the imposition of carbon dioxide emissions limits on new coal-fired power plants. Also, lower costs to the electricity sector of renewable generation—perhaps through subsidies unrelated to nuclear policy—would lower the costs attributed to measures to push nuclear out of the system. And, of course, the pace of US economic growth and electricity demand will have an effect on the cost of the electricity system and of policies that encourage or diminish the contribution of particular generation sources. Even considering these uncertainties, however, the short answer to the large question is that a US exit from nuclear generation will impose costs on the environment, on electricity consumers, and on the national economy.