In my estimation of the renewable energy that we must harvest, I will assume that total U.S. energy consumption will remain flat. I base that prediction on U.S. energy consumption patterns over the last 30 years as well as the prospect for achieving major gains in energy efficiency moving forward.
Raising Energy Efficiency
To begin, we need to distinguish “energy efficiency” from “energy conservation.” Energy conservation entails making do with less—for example, maintaining room temperature during winter at 65 degrees Fahrenheit rather than 70 degrees, or cutting back on average automobile travel from the current figure of 13,000 miles per year to, say, 8,000 miles. By contrast, improvements in energy efficiency lower the amount of energy that we use without reducing our quality of life. Realistically, the opportunity to level off our future energy consumption levels rests far more with raising energy efficiency standards than assuming people will make major sacrifices in their living standards.
For example, switching from incandescent light bulbs to LEDs makes lighting more efficient, but does not reduce the amount of light that we use. Air conditioner efficiencies have dramatically improved since the 1990s and are continuing to improve. Improvements in building insulation decrease the energy required to maintain comfortable indoor temperatures. The number of miles that one can drive on a gallon of gas has increased dramatically and is continuing to increase. Hybrid cars consume less energy than fossil fuel–based cars, partly because fossil fuel–powered cars waste energy when they brake, whereas when hybrid cars brake, part of the energy is returned to the car battery. The energy stored in the car battery can then be used to drive the car forward again. Electric cars are even more energy-efficient than hybrid cars, with equivalent mileage efficiencies that can exceed 100 miles per gallon.
Amazingly, heat pumps can be more than 100 percent energy-efficient because they transfer heat energy from cold places to hot places. The energy required to operate heat pumps is the energy equivalent of a shipping and handling charge. That charge can be much lower than the value of the energy transported. When the temperature difference between the cold and hot places is small, the heat energy delivered can be much larger than the electrical energy. In contrast, the heat energy delivered by fossil fuel burning cannot exceed the energy contained in the fossil fuel. Since the heat delivered is typically three times the electricity used, using heat pumps instead of burning fossil fuels can dramatically reduce energy consumption without requiring any change in indoor temperatures.
Given all of these demonstrated and projected increases in energy efficiency, it might be justifiable to assume that we can maintain our current lifestyle while our energy use is able to decline in absolute amounts over time. However, I will assume, conservatively, that gains in energy efficiency will allow us to stabilize overall energy demand at approximately our current level.
Energy flows from the sun to the Earth, just as rain falls from the sky to the Earth. We often say that one inch of rain fell in the last hour. If we collect water in a bucket, after one hour that bucket would contain one inch of water at the bottom. If there were two identical buckets side by side, each bucket would have an inch of water.
The story is equivalent with solar panels. That is, for a solar panel of a particular size, the total amount of solar energy delivered per hour is given by the flow of energy per hour times the area of the solar panel. In standard units, the energy per hour delivered to one square meter is expressed in watts per square meter (approximately three feet by three feet). The total power delivered to any particular solar panel is therefore exactly the wattage of sunlight per area reaching the Earth’s surface times the size of the solar panel. The average amount of solar power per unit area that reaches the Earth’s surface varies somewhat with location, with sunnier places obviously receiving more solar power than cloudier locations. For example, in Phoenix, New Orleans, and Miami, the average solar power per unit area is about 200 watts per square meter. In Boston, Seattle, and Chicago, it is about 145 watts per square meter.
Reducing our future energy consumption levels rests far more with raising energy efficiency standards than assuming people will make major sacrifices in their living standards.
Solar panels are not 100 percent efficient, so the solar energy delivered to the panel is greater than the electrical power produced by the solar panel. The best solar cells available today approach a 50 percent rate of converting solar energy to electricity. But the typical units that are commercially available today operate at between 15 percent and 18 percent conversion rates. Thus, on average, a square meter solar panel in Phoenix, Miami, or New Orleans would deliver 30 to 36 watts of electrical power, whereas, on average, the same solar panel in Boston, Seattle, or Chicago would deliver 22 to 26 watts.
How this general solar energy delivery system will play out in practice will then depend on the specific setting, for example, whether we are considering individual family homes, apartment buildings, or business operations. On average, American households, both in their own homes and in apartment buildings, consume about 1,250 watts of electricity per hour. Therefore, providing all of the electrical energy for an average household would require about 40 square meters in Miami or about 60 square meters in Boston. For people living in a detached house, providing all of the household’s energy would also require replacing fossil fuel heating and gasoline-powered vehicles.
On average in the U.S., about half of the total energy consumed by single-family homes is used for heating. Fortunately, since the heat energy delivered by a heat pump is on average about three times larger than heat delivered by electricity, supplying the same amount of heat to homes through heat pumps would only require increasing the home’s solar panel area by about 16 percent. If one adds two electric cars, each of which is driven 35 miles per day (i.e., about 13,000 miles per year), that adds an additional 1,000 watts (1 kilowatt) to overall electricity demand. Overall, providing 100 percent of this household’s energy using solar panels would require, roughly, 40 times 20 feet of solar panel area in Miami or 50 times 25 feet in Boston.
The discussion above is only directly relevant for the population living in single-family detached houses. For the U.S. population that dwells in apartments, mobile homes, or multifamily homes, the solar power delivered to the roof of the apartment building, for example, could not provide 100 percent of daily use. Densely packed urban populations would therefore have to use solar or wind energy that is harvested by large commercial farms and then delivered to cities. Those sources would also be needed to supply the energy for industrial use, commercial transportation, and commercial applications other than big-box stores.
Wind delivers about two watts of energy per square meter, on average, i.e., only about one-tenth of what an average solar panel on a rooftop in Chicago can produce. In addition, wind power varies with location much more strongly than solar because the wind energy delivered depends on the cube of the wind speed. The average wind speeds in the central United States are approximately twice as large as the average wind speeds in the mid-Atlantic region. As a result, a wind turbine in Iowa delivers more than eight times as much electrical power as the same wind turbine would deliver if it were located in Virginia. This allows Iowa to generate more than 35 percent of its electrical power from wind while the wind turbines also provide farmers with a significant additional stable income that augments income from raising crops and animals. However, since wind power varies so strongly with location, most inland wind power is and will continue to be generated in regions that are sparsely populated and do not consume much electrical power.
Wind speeds off the coasts of the United States can be quite high, but there has been great political resistance to building wind farms that are visible from coastal properties, and at present offshore wind power is much more expensive than onshore wind. As a result, in the near future, exploiting wind power will require the transport of electricity from rural regions in the central United States to urban regions with high population densities, most of which are located near the coasts.
Importantly, the entire U.S. average energy consumption could be provided through siting wind turbines on about 7 percent of U.S. land area, or one-sixth of existing agricultural land. However, if all Americans who live in single-family homes got 100 percent of their energy from solar panels on their roofs, the total remaining land area requirements from large-scale solar and wind farms would obviously fall dramatically. Most of this land use requirement could be met, for example, by placing solar panels on rooftops and parking lots, then operating wind turbines on about 7 percent of current agricultural land. The remaining supplemental energy needs could then be supplied by geothermal, hydro, and low-emissions bioenergy. Recent work also suggests that under some circumstances solar power can enhance crop growth, so there may be benefits to siting solar panels on some agricultural land. This scenario includes no further contributions from solar farms in desert areas, solar panels mounted on highways, or offshore wind projects, among other supplemental renewable-energy sources, though all of these options are viable possibilities, if handled responsibly.
Remaining Liquid Energy Needs for Airplanes and Boats
Rotating propellers move boats and small airplanes, including drones. Many propellers are already powered by electric motors. For short trips batteries of a reasonable size can store enough energy to complete the trip; however, for long-haul ocean transport and rapid high-altitude air transport, batteries of a reasonable size cannot store enough energy for the voyage. Thus, in the foreseeable future these systems will continue to be powered by liquid hydrocarbons.
Fortunately, it has already been demonstrated that liquid hydrocarbons can be generated using entirely renewable resources. Biofuels already provide a liquid hydrocarbon (ethanol) that is combined with gasoline and fed into all of the cars in the U.S. There are also non-biofuel sources of liquid hydrocarbons. For example, the Fischer-Tropsch process uses chemical interactions to convert carbon monoxide and hydrogen into liquid hydrocarbons. The carbon dioxide in the atmosphere can be harvested to provide the carbon monoxide, and renewable electrical energy can free hydrogen from water. Burning of such liquid hydrocarbons would be carbon-neutral since the carbon originally came from carbon dioxide in the atmosphere. Importantly, if the liquid hydrocarbon is buried in the ground instead of burned, then the amount of carbon dioxide in the atmosphere can be decreased. So, certainly within a 30-year time frame, even these last remaining liquid energy requirements can be met without burning fossil fuels.
Intermittency, Storage, and Transmission
Meeting instantaneous local demand for energy is different from meeting average aggregate demand. A system designed to meet peak power demand will usually provide much more electric power than is required, whereas a system that is designed to provide the average required power will not be able to deliver when demand peaks. Even without renewables, during peak demand times, it has been necessary to fire up fossil fuel plants that are rarely used. The cost of electricity from such plants is very high because they usually sit idle. Geothermal energy and hydropower are renewable-energy sources that do not fluctuate with time. They can help stabilize renewable-energy systems. But in the U.S., hydropower cannot be significantly increased, and geothermal power is limited to certain favorable locations. Matching fluctuating demand to supply can be difficult as well as wasteful.
Importantly, in some cases fluctuations in renewable energy actually correlate with fluctuations in demand. For example, air conditioning demand is high during times when solar power delivery is also high. Thus, solar power can already deliver energy during peak air conditioning times without firing up rarely used fossil fuel resources. In general, in the morning electricity use and solar power show similar fractional increases with time, so in the morning changing to solar power can reduce costly mismatches between the supply and the demand for electrical power. Of course, in the late afternoon solar power begins to decrease before electrical power use does.
Science and technology are not preventing us from achieving a 100 percent renewable-energy economy in the United States.
These problems can be greatly reduced through energy storage. Charging electric cars offers one example. During the night, of course, solar power delivers no power; while during the day, the power delivered is more than the household consumes. Electric cars with 100 kilowatt-hour (100,000 watt-hour) batteries can be bought today. Thus, a household could completely disconnect from the grid if the two car batteries stored excess solar energy generated during the day. In reality, households will probably remain tied to the grid, so that the cars can be driven during the day and charged wherever the car happens to be parked.
At the same time, it is not clear this approach to storage, relying on lithium-ion battery technology, is optimal at a large scale. One issue with lithium-ion batteries is their limited lifetime, which is typically a few years and less than 1,000 charge cycles. Considerable research is being devoted to other battery storage systems.
Excess renewable energy can pump water uphill efficiently as a large-scale energy storage system. But siting options are limited and generally are not located near population centers. Excess electrical energy can also perform hydrolysis that separates water into oxygen and hydrogen. Burning the hydrogen produces energy and water, so hydrogen storage is possible. However, to date, large-scale hydrogen storage systems that are safe and economical do not yet exist. We already have an extremely well-developed system for storing and transporting liquid hydrocarbon fuels, so liquid hydrocarbon fuels created using renewable-energy resources could provide safe large-scale energy storage; however, this process is not yet economically viable.
An alternate strategy for addressing mismatches in energy supply and demand is to build enough renewable energy to meet peak demand, and then re-allocate energy during lower-demand periods. For example, during periods of low demand, renewable energy can create reservoirs of hot or cold liquids. Some thermal solar power systems already generate molten salt that can generate electricity when there is no sunlight. Other systems use stored hot or cold liquids to provide cooling or heating. For example, Austin’s municipally owned power company uses excess electricity generated by nighttime wind power to create an ice-water slurry that is then pumped under city streets to provide district air conditioning of much of downtown. During periods of low demand, the excess electricity could also be used to generate liquid hydrocarbon fuels that store energy for later use. Similarly, if electricity prices during such periods were low, it might be economical for energy-intensive processes such as aluminum smelting or water desalination to run only during times when electricity supply exceeds demand.
Local variations in the supply of wind and solar power can be reduced by transporting energy between regions of the U.S. Also, solar power peaks during the summer, whereas wind power in most of the U.S. is lowest during the summer. Thus, combining wind and solar can reduce the seasonal fluctuations associated with each power source separately.
As noted, the technical challenges in transmitting renewable energy over long distances are not large, but long-distance transport differs from local energy delivery in significant ways. For local electric-power delivery, the flow of electrical current alternates with time, so this type of electricity is called AC (alternating current) power. Those changes with time result in a leakage of power from the electric-power lines. Over short distances, that leakage is not very significant, but over long distances over land the leakage becomes important. The leakage is even more rapid if the electricity travels in power lines that run underwater.
Fortunately, if the direction of the current does not change with time (DC current), the loss for long-distance transport is greatly reduced. In addition, increasing the voltage at which the electricity flows lowers the fraction of the electrical power that leaks from the power lines. As a result, the United States already has high-voltage DC power lines that transport electricity over long distances; however, converting to 100 percent renewable energy would require many more lines, and these power lines are not particularly attractive. If many detached houses generated most of the energy that they use, then requirements for energy transport would be greatly reduced.
In sum, science and technology are not preventing us from achieving a 100 percent U.S. renewable-energy economy. All of the barriers are political, economic, and social. These will need to be solved if the renewable-energy future is to be realized.