Carbon-Free and Nuclear-Free: A Roadmap for US Energy Policy


By Arjun Makhijani

A three-fold global energy crisis has emerged since the 1970s; it is now acute on all fronts:

  1. Carbon dioxide (CO2) emissions due to fossil fuel combustion are the main culprit in the buildup of greenhouse gases in the atmosphere, and fossil fuels – coal, oil, and natural gas – provide over 85 percent of the U.S. and world commercial energy supply. Fossil fuels account for about 84 percent of U.S. greenhouse gas emissions.
  2. Rapid increases in global oil consumption and conflicts in and about oil exporting regions have driven prices high, even as supplies become more insecure. The United States imports 60 percent of its petroleum requirements. At the same time, producing oil in sensitive areas like the Arctic National Wildlife Reserve, from tar sands or shale, or turning coal into liquid fuels raises a host of environmental and resources questions that are difficult, including in some cases, increasing greenhouse gas emissions relative to oil imports.
  3. Proliferation of nuclear weapons is being exacerbated partly by the spread of commercial nuclear power technology. If one uranium enrichment plant in Iran poses such vast security challenges, how will the world cope with a situation where one or more new enrichment plants would need to be built somewhere in the world each year?

Yet, the three problems tend to be treated separately in the policy debate. An integrated energy policy that aims at an efficient U.S. economy based entirely on renewable energy sources like wind and solar energy would address them all simultaneously. Further, a zero-CO2 emissions economy in the United States is not only desirable: Something close to it is a treaty obligation under the United Nations Framework Convention on Climate Change (UNFCCC), given the current state of knowledge about global climate change.

Specifically, the Intergovernmental Panel on Climate Change has estimated that it will require global CO2 emissions to be reduced by 50 to 85 percent relative to the year 2000 in order to limit average global temperature increase to 2 to 2.4 degrees Celsius relative to pre-industrial times. The former represents a 15 percent chance of limiting the temperature rise to this range; the latter an 85 percent chance. If a global norm of approximately equal per person emissions by 2050 is created along with a 50 percent global reduction in emissions, it would require an approximately 88 percent reduction in U.S. emissions. An 85 percent global reduction in CO2 emissions corresponds to a 96 percent reduction for the United States.2 China, India, and other developing countries are unlikely to accept anything less than a uniform per capita global norm – though they may argue for a more stringent standard given historical inequities. If the United States adopts a target of 80 percent CO2 emission reductions, and if a similar per capita level (the U.S. per capita level in 2050 after 80 percent reduction in total CO2 emissions) became the norm worldwide, energy sector CO2 emissions in the year 2050 would be about the same as they are today.

Is a reliable energy system constructed entirely from renewable sources of energy that has the same material benefits as would otherwise be available in the absence of climate change concerns technologically and economically feasible? That is the central question addressed in my book, Carbon-Free and Nuclear-Free: A Roadmap for U.S. Energy Policy.

The starting point for such a study is the set of economic projections used by the Energy Information Administration to assess U.S. energy requirements normatively, assuming a business-as-usual approach. The same number and area of residential and commercial buildings, cars and other vehicles, aircraft passenger miles, and GDP growth in industry and commerce are assumed in the renewable energy economy as in the business-as-usual approach. In other words, no changes in values or lifestyles are assumed, even though such changes may accelerate the transition to a renewable energy economy. The analysis is carried out based on delivered energy to residences, business, vehicles, and industry, after which energy processing losses at power plants and biofuel plants are added. This ensures comparability between the assumptions used to model a renewable energy economy with business-as-usual.

It should be noted that business-as-usual requirements are projected assuming the increases in energy efficiency that have been typical since the energy crisis of the 1970s. In recent years, GDP growth of about 3 percent has been accompanied by energy growth of about 1 percent, in contrast to the pre-1973 period, when energy and economic growth rates tended to be about the same. The change has been much more marked in the industrial sector, where energy use has not grown since 1973, than in the residential, commercial, or transportation sectors.

The increases in efficiency in a fully renewable economy are relative to business-as-usual. A detailed analysis shows that a reduction of one percent per year in end use energy in absolute terms, and even somewhat more until 2050, is fully compatible with the same GDP growth as in business as usual. In other words, an increase in end use efficiency of about two percent per year relative to business-as-usual is shown to be feasible by the analysis in Carbon-Free and Nuclear-Free.

The other keys to phasing out of fossil fuels and nuclear power are:

  1. optimizing solar, wind, standby capacity, and storage to produce a reliable electricity sector;
  2. biofuels made from biomass with high solar energy capture efficiency (defined as being considerably greater than one percent).

If hydrogen production from wind-generated electricity using electrolysis and/or direct solar hydrogen production (for instance, using thermal cracking) can be accomplished economically ($3 to $ 4 per kilogram or less of compressed hydrogen), the transition would be considerably eased. Hydrogen fuel would be produced on a distributed basis to be used in industry as a raw material and possibly as compressed fuel in internal combustion engines for a part of transportation fuel requirements, if the tanks for storage at 10,000 psi can be commercialized. Fuel cell vehicles and are not envisioned in this analysis.


Affordable technologies and practices for vastly improved efficiency already exist in the residential and commercial sectors. Figures 1 and 2 show the current average consumption of energy at the point of use in residential and commercial buildings, compared to efficient buildings.

Figure 1: Delivered energy use, Btu per square foot, residential
Figure 1: Delivered energy use, Btu per square foot, residential

Figure 2: Delivered energy use, Btu per square foot, commercial
Figure 2: Delivered energy use, Btu per square foot, commercial

It is evident that reductions of a factor of three to seven of delivered energy are quite possible using sound principles of building design, such as appropriate thermal mass, south-facing windows and the right levels of insulation. The example of Hanover House in New Hampshire, shown in Figure 1, is especially interesting. It uses active solar thermal heating as well, with a large hot water storage tank (4,500 liters) for space and water heating. It is an all-electric house that gets its energy from the grid and uses electric resistance heating to supplement the solar thermal system. Even though resistance heating is among the most inefficient, the overall energy use is very low. The annual electricity consumption averaged about 5,000 kWh. This could be supplied by a grid-connected solar PV system of about 3.5 kW. The cost of house construction was $111 per square foot, with the owner serving as the general construction contractor.3 Overall specific end use in the residential sectors in the reference scenario developed for a renewable energy economy is estimated to be about 39,000 Btu per square foot in 2050, which is still well above the potential for energy efficiency. Commercial buildings can often meet the highest green energy and environmental recognition (the “platinum” level Leadership in Energy and Environmental Design certification) for less than $10 per square foot in investment, which can be generally recovered relatively easily in reduced energy costs.

Lighting efficiency is especially important in the commercial sector, where it is the largest single energy use, if thermal losses at the power plant are included. It can be reduced by a factor of five or more with existing and emerging technologies. One of the best of the latter is hybrid solar lighting, invented at Oak Ridge National Laboratory.4 A four-foot parabolic solar concentrator focuses sunlight onto a four-inch bundle of optical fibers, which are incorporated into especially designed luminaires, which also have electric lights. The electric lighting component automatically compensates for varying solar light availability, maintaining a constant output. An added benefit is reduced air-conditioning electricity use during sunny days, since the thermal loading of the solar part of the luminaire is negligible compared to electric lighting.

Oil use in transportation is even more inefficient than energy use in buildings. The net average efficiency of personal car transport is currently only about one percent, based on the payload transported – the people in the car.5 Besides automobile efficiency standards, one logical place to start is plug-in hybrids – gasoline-electric cars that have extra batteries that store enough charge to enable much or most commuting on electricity only. Depending on the battery capacity, the liquid fuel efficiency is 70 to 100 miles per gallon, plus an input of 0.1 to 0.15 kWh per mile. There is no real obstacle to commercialization of this technology. Efficiency standards set for the year 2020 should reflect this.

Plug-in-hybrids can be charged using renewable energy sources. This will reduce both oil imports and CO2 emissions. They can also be used in a “vehicle-to-grid” arrangement –V2G for short. When the batteries are low, the car is charged from the grid; when grid needs power, it can draw on the electricity stored in parked cars that are plugged in. The first small-scale practical trial of V2G is being prepared in Google’s Silicon Valley parking lot. A V2G system can provide electricity storage to help solve the one real difficulty associated with very large-scale use of solar and wind energy: they are intermittent.

A V2G scheme was economically unthinkable even two years ago. The batteries wore out much faster than the rest of the car, making it prohibitively expensive to use them in a V2G system. But tests show that newly designed lithium-ion batteries will last far longer than the car. With them, V2G can provide one way for solar and wind energy to reliably provide the majority of the electricity we need. The batteries are still being made on a small scale. A cost reduction of about a factor five is needed; it is expected in the next few years as production technology matures and economies of scale kick in. Most cars are parked over 90 percent of the time. Only a few percent of all cars would be needed to provide large-scale back up for renewable electricity sources.

All-electric cars using advanced lithium-ion batteries also appear to be on the horizon. Phoenix Motorcars is making an all electric five-passenger pick-up truck, with a range of ~130 miles and an efficiency of about 3.5 miles per kWh. Tesla Motors is making a sports car that goes from zero to 60 mph in four seconds and has an efficiency of about 5 miles per kWh. Lithium-ion battery pack costs need to come down by about a factor of five before such cars can be commercially competitive. One great advantage of such vehicles, of course, is that they can be charged using renewable electricity sources. The greatest energy efficiency improvements are assumed to occur in the transportation sector, because that has the greatest potential, especially in a transition to electric or mostly electric personal vehicles.

Liquid fuels

What about fueling aircraft, trucks, and industry? Let’s first note that, except for waste cooking oils, using food sources for energy is not a very good idea. The net energy balance of ethanol from corn is poor. The net greenhouse gas emission reduction is modest, at best. Even at moderate levels of production, using corn for ethanol fuel is causing a rise in food prices in the United States, Mexico, and elsewhere. Using palm oil for biodiesel is even worse than ethanol from corn. The emissions of CO2 from the destruction of the peat bogs in Indonesia, where the palms are grown, are much greater than if petroleum were used in transport. Using food crops as a major source of fuel is unsustainable in a world of eight to ten billion people (by 2050) who are acquiring the means to eat well.

Biofuels are important for a renewable energy future, but we must use a sharp pencil to choose the right ones. Some approaches that are being funded, like converting corn stover and prairie grasses to ethanol (“cellulosic ethanol”), are worthwhile. But the most promising approaches are not on the national policy radar yet.

Consider microalgae -- tiny, ubiquitous plants that can even grow in salty water. The right species can provide 5,000 to 10,000 gallons of biodiesel per acre, compared to about 300 gallons of gasoline equivalent for ethanol from corn. The main inputs are water, sunshine, and today’s pollutants – carbon dioxide and nitrogen oxides from power plant exhaust.

The technology has been demonstrated on a small-scale at the Massachusetts Institute of Technology. Larger scale tests have been done at power plants in Arizona and Louisiana. A pilot test over 19 summer days at the Arizona power plant produced an yield of 98 grams of dry matter per square meter per day, indicating an annual potential at present of 200 metric tons per year or more.

Other high productivity biomass includes aquatic plants that are now considered as nuisances or worse. Of special note are water hyacinth and duckweed. The former may be the most productive plant on Earth, with a solar energy capture efficiency of up to 5 percent – yielding up to about 250 metric tons of dry organic matter per hectare per year. Aquatic weeds grow well in high-nutrient content water, such as agricultural run-off and municipal wastewater. They have been used as part of experimental wastewater treatment systems, off and on since the 1970s, but they have never been a significant part of energy considerations. That needs to change.

Overall, the requirements for liquid and gaseous biofuels in the residential, commercial, transportation and industrial sectors present possibly the most difficult challenge to a transition to a renewable energy economy. The main issue is land area requirements. Combing high productivity prairie grasses and very high productivity aquatic plants (including microalgae) with high efficiency would still result in requirements of 5 to 6 percent of the land area of the United States. But with the right choices of plants, the biofuels could be grown on land that is now not suitable for agriculture – even in desert areas. For instance, wastewater from the Los Angeles metropolitan area could be treated in the Owens Valley, where aquatic plants could be grown as part of the treatment. In any case, much of the water for the Los Angeles region comes from the Owens Valley.

It would be highly desirable to reduce land area requirements in a renewable energy economy. One way would be to focus on commercializing direct solar hydrogen production technologies, which are still largely in the laboratory stage. Electrolytic hydrogen using wind power plants is closer to commercial. However, large-scale development of this technology would also require development of a hydrogen pipeline infrastructure, which would be a major undertaking amidst several other major transformations. Such an infrastructure is not envisioned in the analysis. Hydrogen use would be mainly for industry as a feedstock and possibly as compressed hydrogen for use in internal combustion engines, if developmental problems are resolved.6

A Renewable Electricity Grid

The United States is blessed with enormous renewable energy potential. North Dakota, Texas, Kansas, South Dakota, Montana, and Nebraska each have wind energy potential greater than the electricity produced by all 103 U.S. nuclear power plants. Wind energy is already more economical than nuclear. Solar energy is even more abundant. Assuming 20 percent efficiency, 0.1% of the land area of the United States would provide almost all U.S. electricity requirements. Indeed, the area of parking lots and commercial rooftops is large enough to provide most U.S. electricity requirements. Wind energy is presently economical. Solar energy costs are running at about 20 cents per kWh and are declining. The small scale of manufacturing capacity is a major contributor to high cost. That is changing rapidly. The Department of Energy expects commercial competitiveness by about 2015. The land area requirements of wind and solar electric power plants are modest. The footprint of wind turbines, including roads and other infrastructure is ~0.6 hectares per megawatt (though it varies a good deal). A trillion kWh of solar electricity can be generated on less than 1,600 square miles in sunny areas, including a 30 percent allowance for infrastructural land.

Individual technologies aside, can a reliable electricity grid be created using renewables alone? If so, how do we get from here – coal, nuclear, natural gas, and hydropower – to there? One key is to use existing infrastructure to make the transition. For instance, hydropower use can be restricted mainly to times when the wind is not blowing. This is being demonstrated in Washington State.

A large economic miscalculation made in the last 15 years can also be turned to advantage. During that time, natural gas-fired power plants with a capacity three times larger than U.S. nuclear capacity were built, in anticipation of continued low natural gas prices. But prices have tripled and the plants sit idle over 80 percent of the time. This huge capacity can be used to provide backup for wind and solar, at night and times of low wind. In 20 or 30 years, natural gas can be replaced by methane from biomass. With this approach, the fraction of solar and wind energy in the grid can be increased from under one percent at present to 30 percent or more over the next two decades, without resort to storage technologies that are now uneconomical. Intermittency of solar and wind is therefore no bar to meeting electricity growth requirements with a combination of efficiency and an optimized mix of solar, wind, hydro, and natural gas standby power for about two decades. New nuclear and coal-fired power plants are quite unnecessary to a reliable grid. In fact, a distributed grid that uses commercial rooftops and parking lots as a key feature, would be more secure than the one we have today.

In the longer term, some baseload power plants using solid biomass, geothermal energy, and some intermediate load solar thermal power plants with thermal storage will be necessary to anchor the system, unless electricity storage technologies such as ultracapacitors and sodium-sulfur batteries become much more economical than they are today. A possible configuration for a renewable energy grid is shown in Figure 3.

Figure 3: A renewable electricity grid configuration

Figure 3: A renewable electricity grid configuration

The costs of electricity per kWh from such a grid may be in the range of 12 to 18 cents per kWh, if solar energy costs continue to decline for the next few years. This is higher than those prevailing in most parts of the country. However, the overall cost of energy services will not be higher if appropriate investments are made in efficiency.

It will not be easy. Determination, a vigorous research, development, and demonstration program, and sensible overall policies are essential requirements. But we can solve all three problems – climate change, nuclear power as a source of proliferation, and insecurity of oil supply – simultaneously. Done right, it will not burn a hole in the national pocketbook – the amount we spend on energy services, such as heating, cooling, lighting, and personal transportation, as a fraction of Gross Domestic Product would remain about the same as energy expenditures in 2005 – about eight percent. But more will be spent on efficiency and less on fuels and electricity.

The main policy recommendations arising from the analysis are as follows:

  1. Enact a physical limit of CO2 emissions for all large users of fossil fuels (a “hard cap”) that steadily declines to zero prior to 2060, with the time schedule being assessed periodically for tightening according to climate, technological, and economic developments. The cap should be set at the level of some year prior to 2007, so that early implementers of CO2 reductions benefit from the setting of the cap. Emission allowances would be sold by the U.S. government for use in the United States only. There would be no free allowances, no offsets, and no international sale or purchase of CO2 allowances. The estimated revenues – approximately $30 to $50 billion per year – would be used for demonstration plants, research and development, and worker and community transition.
  2. Eliminate all subsidies and tax breaks for fossil fuels and nuclear power (including guarantees for nuclear waste disposal from new power plants, loan guarantees, and subsidized insurance).
  3. Eliminate subsidies for biofuels from food crops.
  4. Build demonstration plants for key supply technologies, including central station solar thermal with heat storage, large- and intermediate-scale solar photovoltaics, and CO2 capture in microalgae for liquid fuel production.
  5. Leverage federal, state, and local purchasing power to create markets for critical advanced technologies, including plug-in hybrids.
  6. Ban new coal-fired power plants that do not have carbon storage.
  7. Enact at the federal level high efficiency standards for appliances.
  8. Enact stringent building efficiency standards at the state and local levels, with federal incentives to adopt them.
  9. Enact stringent efficiency standards for vehicles and make plug-in hybrids the standard U.S. government vehicle by 2015.
  10. Put in place federal contracting procedures to reward early adopters of CO2 reductions.
  11. Adopt vigorous research, development, and pilot plant construction programs for technologies that could accelerate the elimination of CO2, such as direct solar hydrogen production (photosynthetic, photoelectrochemical, and other approaches), hot rock geothermal power, and integrated gasification combined cycle plants using biomass with a capacity to sequester the CO2.
  12. Establish a standing committee on Energy and Climate under the U.S. Environmental Protection Agency’s Science Advisory Board.

Arjun Makhijani is president of the Institute for Energy and Environmental Research. In November 2007 he was elected a Fellow of the American Physical Society "for his work to provide the public with accurate and understandable information on energy and environmental issues," having been nominated by the Forum on Physics & Society.

Based on a forthcoming book of the same title, published by RDR Books. The book can be downloaded free at Details of the analysis and references can be found there.

Based on a global population of 9.1 billion and a U.S. population of 420 million in the year 2050.

More details about this house are on the Internet at The solar PV installed capacity requirement would, of course, be lower in sunnier parts of the United States.

Details are available at

Vehicle efficiency = 15%; average vehicle weight of 3,240 pounds; occupancy of 1.64 person-miles per vehicle mile. Average weight (all ages) ≈ 130 pounds.

The APS’s Panel on Public Affairs, The Hydrogen Initiative, March 2004 concluded that one to two orders of magnitude improvements in technology and discovery of a new material for vehicle storage tanks would be needed for a fuel cell car to be able to compete with gasoline cars. See Hydrogen Initiative link at