Special Viewpoints

A Realizable Renewable Energy Future

See allHide authors and affiliations

Science  30 Jul 1999:
Vol. 285, Issue 5428, pp. 687-689
DOI: 10.1126/science.285.5428.687

This article has a correction. Please see:


The ability of renewable resources to provide all of society's energy needs is shown by using the United States as an example. Various renewable systems are presented, and the issues of energy payback, carbon dioxide abatement, and energy storage are addressed. Pathways for renewable hydrogen generation are shown, and the implementation of hydrogen technologies into the energy infrastructure is presented. The question is asked, Should money and energy be spent on carbon dioxide sequestration, or should renewable resources be implemented instead.

Interest in renewable energy has depended on the perceived risks of using fossil fuels. During the energy crisis of the mid-1970s, the perceived risk of running out of conventional fossil fuels led to crash programs in developing renewable sources and energy conservation measures, including higher vehicle fuel economy and energy-efficient buildings and homes. These programs were scaled back as supply once again met demand. In the 1980s, the risks associated with pollution spurred work to avoid or remedy environmental damage from fossil fuel extraction, processing, and transport and led to measures to burn fossil fuels more cleanly (the catalytic converter is a product of this movement). The United States's dealings with these risks were akin to the analogy of a smoker who, having perceived the risk of heart disease, takes up running—but continues to smoke. More recently, the risks associated with CO2 emissions and global warming have again spurred interest in renewable energy. However, we cannot continue to burn fossil fuels and somehow sequester the produced CO2 efficiently enough to actually address global warming, as the processes of concentrating and burying or transforming the CO2 are themselves energy intensive; as in the above analogy, there is no activity that can clean our lungs while we are still smoking. Although CO2 sequestration, where it is generated near depleted gas fields and aquifers or used for tertiary recovery, can be done at low cost, building pipelines to transfer CO2 to sequestration sites can quickly become an expensive endeavor. Pipeline costs can range from $1 million to $2 million per kilometer. The safety of CO2 sequestration has also not been fully addressed. Should a CO2 sequestration reservoir break and the CO2 make it to the surface, it would effectively displace O2, because CO2 is heavier than air.

Thus, if global warming issues require us to quit our fossil fuel use, are there in fact practical renewable alternatives? More generally, is there a sustainable energy system that can supply a growing population with energy without destroying the environment within which it is used, providing energy for the present without compromising the ability of future generations to meet their needs (1)?


The United States is the largest user of energy in the world (∼50% of total consumption), and any change in global energy use would require a change in U.S. production and consumption of energy. Is it then actually feasible to supply all of the U.S. energy needs from renewable energy? The major renewable energy systems include photovoltaics (PVs) (or solar cells), solar thermal (electric and thermal), wind, biomass (plants and trees), hydroelectric, ocean, and geothermal (2). Solar cells, perhaps the most recognizable solar energy converters, directly convert the sun's energy to electricity with no moving parts, whereas solar thermal systems generate heat, and range from the simple solar hot-water heating system to the multimegawatt power plant outside of Barstow, California. Wind energy represents the nearest term cost-competitive renewable energy source. Produced by the heating of Earth, wind as an energy resource is possible over the entire United States and presents a dual-use technology: The land can still be used for farming, ranching, and forestry. Biomass power ranges from burning wood chips in power plants to burning biogas from waste treatment plants to the generation of methanol and ethanol, which can be used as fuels. The ocean is Earth's largest collector of solar energy, and ocean thermal platforms have a large potential for electricity generation.

Figure 1 shows the ability of PV technology alone to provide all of the energy needs of the United States. This calculation assumes a 10% solar-to-electrical system efficiency and the use of fixed flat-plate collectors; tracking to follow the sun would lower the area required. A square ∼161 km (∼100 miles) on a side would, during 1 year, produce the energy equivalent to that used annually in the entire United States (3). Although 25,921 km2 (10,000 square miles) is a large area, it is less than one-quarter of the area that this country has covered with roads and streets. If wind is added to the energy mix, this area for PV is reduced (in fact, the United States also contains enough usable wind resources to produce all of the electricity used by the nation); if geothermal energy is added, the PV area is even smaller, and if hydroelectric energy is added, the area is again smaller. The point is clear—we can gather more than enough renewable energy to power our society.

Figure 1

(above). Calculated area required for a PV system to produce the entire U.S. yearly electrical needs.

More important than the area required to supply this amount of energy is the way in which the electricity is generated. If we look at solar irradiance data (how much sunshine is available per day), we see that the sun only shines in that area for an annual average of ∼6 hours/day (4). One of the major drawbacks to many forms of renewable energy is their intermittency, which I will discuss in more detail below.

Paybacks and Costs

A persistent belief is that renewable resources require more energy in their manufacture than they produce in their lifetime; however, actual calculations show a very rapid payback. For example, the energy payback for current PV systems has been calculated to range from 3 to 4 years, depending on the type of PV panel (thin-film technology or multicrystalline silicon, respectively). This energy payback time includes the energy costs for processing the semiconductor and assembling a module, frame, and support structure (5–8) and is expected to be reduced to 1 to 2 years as manufacturing techniques improve. Wind energy has an even faster payback of 3 to 4 months (9). During their lifetime (30 years for PV and 20 years for wind), these technologies not only pay back the original energy investment, but also the emissions produced from their own manufacture.

The CO2 savings from displacing fossil fuels with PVs depend on the regional fossil fuel mix and the solar irradiance; values range from 270 g of CO2/kilowatt-hour (kWh) to >1050 g of CO2/kWh. Other pollutants are avoided, including NOx, SO2, and particulates. With the average value of 662 g of CO2/kWh and an average of 5.5 hours of sunlight per day, a 1-kWh PV panel would give a yearly CO2 savings of 1330 kg. The CO2 payback time from avoided emissions also depends on the local energy mix and the panel efficiency. Assuming an energy cost for a crystalline silicon panel of 600 kWh/m2 (5) and the average of 662 g of CO2/kWh, the manufacture of a 1-m2 panel produces ∼400 kg of CO2. If we assume 10% efficiency and a solar irradiance of 1 kW/m2, it takes 10 m2 of collector to produce 1 kW. Manufacturing 10-m2 panels produces 4000 kg of CO2, which is paid back in avoided emissions at 1330 kg/year for a total time of 3 years. Higher efficiencies lower both the energy and CO2payback time, as do manufacturing techniques that are more energy efficient. From this, one can easily envision a renewable energy breeder plant, a manufacturing plant that completely relies on renewable energy for the manufacture of additional renewable energy resources.

An Infrastructure Roadmap

Two pathways can lead to a renewable-based energy infrastructure while reducing CO2 emissions and oil imports. One pathway focuses on the U.S. electricity-producing sector and the other on the U.S. transportation sector.

Any renewable energy system that produces electrons should be connected to the infrastructure to directly reduce the CO2 emissions from current fossil fuel generation and to avoid construction of additional fossil fuel power plants. PV, of course, comes very close to matching the early afternoon peak in energy use. There is some debate as to how much intermittent power the infrastructure can accept, but even 10% would constitute a large amount of renewable energy.

Intermittency and H2 Use

The missing links required for a sustainable energy system are an energy storage scheme (a way to store the renewable energy for times when it is not being generated) and an energy carrier (something to replace gasoline and other fossil-derived energy carriers). Energy storage technologies include H2, batteries, flywheels, superconductivity, ultracapacitors, pumped hydro, and compressed gas. The most versatile energy storage system and the best energy carrier is H2.

Hydrogen can replace fossil fuels as the energy carrier for transportation and electrical generation when renewable energy is not available. Because H2 is transportable by gas pipelines or can be generated on site, any system that requires an energy carrier can use H2. The conversion of the chemical energy of H2 to electrical energy by a fuel cell (10) produces only water as waste.

Currently, H2 is manufactured in large quantities from steam reforming of natural gas. However, H2 can be generated by solar energy. Figure 2 shows a number of pathways by which solar energy can be used to generate H2. Thermolysis, the direct splitting of water at high temperatures, suffers from the rapid back reaction of H2and O2 at these temperatures, preventing this pathway from being a viable approach. Thermal cycles, in which O2 and H2 are generated in separate steps, are well known, and although these were initially developed to use the waste heat from a nuclear reactor, some have been adapted for solar concentrator systems. The conversion of biomass to H2, although fairly straightforward, has a low conversion efficiency from sunlight to H2, and any system designed to generate substantial amounts of H2 must be rather large. Nonetheless, if the biomass used is a waste by-product, then this is perhaps the least expensive of the H2 generation technologies. Wind energy and PV systems coupled to electrolyzers (11) are perhaps the most versatile of the approaches and are likely to be the major H2producers of the future. These systems are commercially available but are very expensive.

Figure 2

(right). Sustainable pathways from solar energy to H2.

The photolysis systems combine the two separate steps of electrical generation and electrolysis into a single system. These direct conversion systems include photoelectrolysis and photobiological systems and are based on the fact that visible light has a sufficient amount of energy to split water (12).


Fuel cell vehicles have the possibility of being highly efficient, reaching an equivalent fuel economy of over 42 kilometers/liter (100 miles/gallon), with performance and range equivalent to current vehicles (13). Additionally, when fueled with H2 generated from renewable resources, these vehicles have no emissions. However, there is a catch—a H2-fueling infrastructure does not exist. Manufacturers do not want to build a vehicle that cannot be easily refueled, and energy companies do not want to build an infrastructure that has limited or no use.

There appear to be two pathways to addressing these issues. The one currently receiving the greatest funding is a fuel cell vehicle that uses an on-board chemical process plant to reform (convert) gasoline or other liquid fuel to H2. Miniaturizing reformers for small-scale intermittent operation appears to be unreliable, however. Although this approach would allow the continued use of the current energy infrastructure, it does not move us toward a necessary, more environmentally sound H2 economy.

H2 as the energy carrier is a much better approach. On-board H2 would allow a vehicle to maximize the energy-efficient aspects of fuel cells and the environmental attributes of H2. The near-term approach would be to build stationary reformers at current compressed natural gas filling stations. Because it would run continuously, putting the same small-scale reformer on the ground would spread the cost over a larger number of consumers and would minimize development costs (14). Initially, the H2 would be used in modified internal combustion engines. Although not “pollution free,” the exhaust from a H2 internal combustion engine would be much cleaner than that from gasoline engines. This engine technology is already well developed, and the entire automotive industry would not have to retool to build these more environmentally sound vehicles. This approach minimizes the risk to the consumer and allows a H2infrastructure that deals with safety issues to be built simultaneously with the development of fuel cells. Markets that can afford the higher initial costs for new technologies would be used to develop commercial applications of fuel cells. These markets include buses, mining vehicles, buildings, and distributed and remote stand-alone power systems. Fuel cell vehicles would be available when the costs of fuel cells come down to an acceptable amount for automotive applications. Again, one can easily show that renewable energy can supply all of the H2 necessary for cars (15).

Research and Development Issues

The key to the above infrastructure scenario is energy storage. Because the energy storage system is likely to be very dependent on the local environment, there needs to be a capability for matching the energy storage system with the energy generation system. Pumped hydro, for example, is perhaps the least expensive large-scale energy storage system but is very site specific. Hydrogen storage systems are particularly important. Hydrogen stored as compressed gas or as a hydride works well for stationary applications; however, H2 storage for transportation is problematic because of its low volumetric density. Some automakers (for example, BMW and DaimlerChrysler) are looking at liquid H2, but one must pay a heavy energy penalty to liquefy it. Advanced automotive concepts that reduce weight and resistance (16) are very important because they reduce the amount of fuel necessary. Nonetheless, research is necessary on the development of compressed gas tanks made of lightweight advanced composites and on new H2 absorbents such as carbon nanotubes.

For the PV industry, increasing module efficiency to 15% is critical, especially for thin-film technologies that are less expensive to manufacture. For the crystalline solar cell industry, refining plants must be built to produce solar-grade silicon. Currently, the crystalline silicon industry uses microelectronic silicon scrap as feedstock. That feedstock will soon be exhausted as demand increases and the microelectronics industry becomes better at recycling their scrap. Inexpensive systems that concentrate sunlight could minimize the costs and manufacturing demands of PV material and greatly increase our capability of producing electricity from PVs (17).

Two key components of this scenario that require additional research and development are fuel cells and electrolyzers, key components in the load-leveling scheme of renewable energy (electrolysis and H2 storage) fuel cells. These electrochemical systems have been studied for over a century and yet have some distance to go before they can be brought into the energy infrastructure mainstream. The cost of electrolyzers and fuel cells will be critical. Electrolyzers currently cost about $3000/kW, and for large-scale use, they should be less than $200/kW. Fuel cells are even more expensive, and for automotive applications, they need to be less than $50/kW.


The question for our time is whether governments should subsidize building, manufacturing, and implementing renewable energy technologies, or should governments instead develop and build CO2 sequestration technologies. By developing and implementing renewable energy technologies and manufacturing capabilities, we would build a sustainable energy infrastructure to carry us well into the next millennium. We should embark upon this path with all due speed.


View Abstract

Navigate This Article