The American Energy Security Crisis Solution—Space Solar Power

Section VII – Assessing Ground-Based Solar Energy and Wind for Meeting U.S. 2100 Energy Needs

With conventional and advanced fusion nuclear energy being unlikely to replace fossil fuels this century, the only other practical terrestrial options are the renewable energy sources of wind, ground solar, hydroelectricity, geothermal-electricity, biomass, and tidal/wave-generated electricity. Can they provide the equivalent of 6,500 GW of dispatchable generation capacity?

The last four options fall into the category of either being impractical, e.g., tidal/wave-generated electricity, or not being capable of significant expansion.

  • The United States has about 78 GW of installed hydroelectric generating capacity and the potential to add only about 30 GW of new generating capacity.[21]
  • The United States has about 4 GW of geothermal-electricity generation. In 1978, the U.S. Geological Survey estimated the total identified and undiscovered geothermal electrical power generation potential in the United States at 95-150 GW. Yet, over the last 30 years, very little of this potential has been developed indicating the difficulty in commercializing this potential.[22]
  • In 2005, the Departments of Energy and Agriculture evaluated the potential of land biomass as a fuel source.[23] This author estimated that the Government’s projected potential could yield about 16.4 quadrillion BTU or 2.8 billion BOE of combustible fuels—alcohol, biodiesel, etc.[24] This required the substantial use of genetically-modified crops to increase residual biomass production and the use of nearly all recoverable agriculture, farm, and forestland waste from roughly one million sq. mi. of farmland and forestland. A key point of this 2005 study, however, was that it was based on meeting the food and feed needs of the U.S. at the present time and not in 2100 when the population will likely have doubled. All of these factors indicate that any significant expansion of biomass use for energy production is unlikely.

Consequently, of these remaining terrestrial renewable energy alternatives, only ground solar and wind have the potential to be scaled up to the necessary capacity. By using the information from the earlier all-nuclear energy assessment, the practicality of building ground solar and wind farms of sufficient scale to meet the 2100 energy needs can be readily evaluated.

A. The 14 MW Nellis Air Force Base solar farm is used as a baseline for evaluating the potential of ground solar energy

{Author’s note: These calculations are updated in the 2015 paper “Becoming Spacefaring: America’s Path Forward in Space” to make use of the performance on several newer solar farms. The results, however, are very similar in terms of land area required.]

In 2007, the U.S. Air Force installed a moderately-sized ground solar photovoltaic farm at the Nellis Air Force Base outside of Las Vegas, Nevada. Nellis Air Force Base is a primary flight training facility, indicating that clear blue skies are the norm and good solar insolation (watts of sunlight/sq. ft.), should be available most days. In fact, in terms of the level of solar insolation, this is one of the best locations in the continental United States. This makes this solar farm’s performance a good baseline for evaluating the potential of ground solar energy.

The solar farm covers 140 acres (0.219 sq. mi.) and is comprised of solar photovoltaic panels mounted either on a translating stand, as seen in the bottom photograph in Fig. 17, or a standard fixed panel stand. The advantage of the translating stand is that it rotates the panels from east to west to track the movement of the sun across the sky to maximize solar-electricity output throughout the day. However, the disadvantage is the tracking system’s added cost and maintenance needs.

Figure 17 – Nellis Air Force Base 140-acre solar farmThe nameplate generation capacity of the 72,000 installed panels totals about 14 MW.[25] The monthly and annual performance of this solar farm over the years 2008-2012 is shown in Fig. 18a and 18b. The monthly output is shown in Fig.18a while the year-to-year variation in total annual output is shown in Fig.18b.

Figure 18a – Nellis Air Force Base solar farm monthly performance Figure 18b – Nellis Air Force Base solar farm yearly performance

During the first five years of operation, the 0.219 sq. mi. solar farm produced an average of 32.0 GWh/yr. of electrical energy. This equals 146.1 GWh per sq. mi. per year.

32.0 GWh ÷ 0.219 sq. mi. = 146.1 GWh/sq. mi.

To model a solar farm output using this Nellis data, the following adjustments are included:

  • Increase the net output of the solar panels by 33% to account for more efficient photovoltaic cells, mounting, and positioning within the farm.
  • Apply a 90% adjustment to account for lower average insolation values, primarily due to weather, as the area of the solar farms expands to cover most of the Southwestern United States.
  • Apply a 73.9% adjustment to account for the use of lower-cost and easier-to-maintain fixed-panel mounting rather than the translating stand used primarily at Nellis.
  • Assume 95% availability.

Applying these adjustments to the real-world Nellis data yields a model estimate of 122.8 GWh/sq. mi. for solar farms located across the American Southwest. This will be used in computing how many sq. mi. of solar farms are needed to yield the 31.25 billion BOE of gross thermal energy needed in 2100.

146.1 GWh/sq. mi. x 1.33 x 0.9 x 0.739 x 0.95 = 122.8 GWh/sq. mi.


B. To meet U.S. 2100 energy needs with ground solar energy would require about 521,000 sq. mi. of solar farms

As mentioned, a primary issue with ground solar (and wind) is the variability of the electricity produced by a solar farm, as seen in Fig. 18b. The U.S. electrical power infrastructure is tightly regulated and controlled to ensure continuous, high-quality electrical power at all times. What the end-consumer receives from the utility is referred to as “dispatched electricity.” This electricity must be continuously generated because it only takes a fraction of a second for the generated electrical power to reach the end-consumer. (Electricity is not stored in the utility’s transmission and distribution system.)

As can be easily imagined, trying to deliver high-quality dispatched electricity from a variable input source, such as ground solar or wind, is very difficult, especially as the scale of production grows. The solution used in this model is to change the solar-electricity into hydrogen, store the hydrogen, and then use hydrogen-fueled gas-turbine generators at the local utilities to generate the needed dispatched electricity. The overall efficiency of this, using the same improved technology assumptions as were included in the previous nuclear model, is 43% (See Fig. 19). This means that 1 GWh of solar-electricity from a solar farm will yield 0.43 GWh of dispatched electricity from the utility to the customer.

Figure 19 – Overall efficiency in producing dispatched electricity from a variable electrical power source

From Fig. 15, the U.S. will need 7.42 million GWh of dispatched electricity in 2100. To provide this from ground solar farms, the total area of the farms would need to be about 141,000 sq. mi.

7.42 million GWh needed in 2100
÷ (122.8 GWh/sq. mi. of solar farm x 0.43) = 140,520 sq. mi.

A slightly different analysis is used to compute how many sq. mi. of solar farms are needed to provide the 18.47 billion BOE of hydrogen fuels needed in 2100. For this simple analysis, all of the solar-electricity generated for this purpose is assumed to be converted to hydrogen fuel. As in the all-nuclear case, the conversion rate is assumed to be 2,529 kWh per BOE of hydrogen stored at 6,500 psi. Repeating the calculation from the all-nuclear analysis, this requires around 46.7 million GWh. With each sq. mi. of solar farms yielding an estimated 122.8 GWh, the area needed to produce fuel in 2100 is about 380,000 sq. mi.

18.47 billion BOE of hydrogen fuel x 2,529 kWh/BOE of hydrogen @ 6,500 psi
÷ 1000 kW/MW ÷ 1000 MW/GW = 46,710,630 GWh

46,710,630 GWh ÷ 122.8 GWh/sq. mi. = 380,380 sq. mi. of solar farm

By adding these two estimates, the total net area of advanced ground solar farms needed in 2100 is about 521,000 sq. mi. The continental United States totals about 3 million sq. mi. Nearly 18% of the U.S. lower 48 states would need to be bulldozed flat and planted with solar arrays. Additional ground would be needed for access roads, transmission and distribution systems, substations, etc.

140,520 sq. mi. for dispatched electricity
+ 380,380 sq. mi. for fuels = 520,900 sq. mi. of solar farms

An important point to recognize is that in the Southwestern United States, only a modest percentage of the ground is sufficiently flat to be used for solar farms. Hence, while the actual farms may require 520,900 sq. mi., this will be spread out over a much larger geographic area. For comparison, the entire land area of New Mexico and Arizona totals only about 236,000 sq. mi. Hence, virtually all of the flat ground in the southwestern states extending as far east as western Texas and as far north as northern Nevada would be needed for solar farms. Is this practical?

C. To meet the U.S. 2100 energy needs with wind-electricity would require 1.4 million sq. mi. of wind farms

Wind has been the fastest growing segment of the renewable energy portfolio. Wind, like ground solar, is a variable power source and must be treated in much the same way by producing hydrogen to generate both dispatched electricity and end-consumer fuel.

The Federal Government has mapped the wind energy potential across the United States. Figure 20 shows the distribution of average wind speed at 80 meters (262 ft.) above the ground. This corresponds to the hub height of a typical 1.5-MW wind turbine. The purple-red areas in the map below have the greatest potential, with average wind speeds in the range of 8.5-9.5 meters/sec (19-21 mph). Most of the continental United States, however, has poor wind power potential. This means that wind farms must necessarily be located in the central United States—the primary food growing region of the country.

Figure 20 – U.S. 80-meter wind power map

Figure 21 shows the variation in monthly output for four 1.8-MW wind turbines—7.2 MW total—located in northwestern Ohio. The “capacity factor” is the percentage of the total potential wind energy—expressed in GWh—that the wind turbine actually generates each month or year.[26] For the 12-month period of November 2003-October 2004, the average capacity factor was about 22%. To be clear, this means that over this 12-month period, the wind turbines produced only 22% of the energy that would have been produced had the turbines been generating their nameplate 7.2 MW continuously.

Figure 21 – Actual monthly capacity factor at four Ohio 1.8 MW wind turbines for Nov 2003–Oct 2004

Early wind farms were concentrated on low mountain ridges in California because the ridge accelerated the wind’s speed and, consequently, the available wind power. These wind farms positioned the turbines along the ridge because the wind direction was usually blowing in just one direction—across the ridge. Such ideal ridge locations are only a small percentage of the land area of the United States with good wind conditions.

In more typical circumstances, the wind turbines are spaced in a grid to enable the wind to be harnessed regardless of the direction the wind is blowing. Wind turbines extract power by slowing down the wind. If the turbine spacing is too close, the wind speed does not have sufficient distance to recover and the wind farm loses generation potential.

For this reason, wind turbines are assumed to be optimally spaced in a grid such that the total installed nameplate power per sq. mi. of wind farm is about 12.9 MW.[27] If a wind farm uses 1.5-MW turbines, optimally 8.6 turbines would be installed per sq. mi. If a wind farm uses the 500 ft. tall 2.5-MW turbines, optimally 5.16 would be installed per sq. mi.

Using wind power surveys, the Federal Government has projected the wind energy potential of the United States. This is shown in Fig. 22 for a range of minimum capacity factors and hub heights. From this estimate, wind farms, with 100 m (328 ft.) hub heights and covering 936,000 sq. mi. of primarily the central United States, would be capable of generating about 45 million GWh of variable wind-electricity per year.[28] Assuming 95% availability, about 46 GWh of wind-electricity is generated per sq. mi. per year.

45 million GWh ÷ 936,000 sq. mi. x 0.95= 45.7 GWh/sq. mi.

Figure 22 – Potential installed wind energy annual output

Recall that the annual energy output of the ground solar farms was estimated to be 122.8 GWh/sq. mi. This required a total of 520,900 sq. mi. of advanced solar farms to meet the U.S. 2100 energy needs. Scaling this farm area up to account for the lower output from the wind farms, the required wind farm area in 2100 would be about 1.4 million sq. mi.—substantially greater than the suitable land in the United States for commercial onshore wind farms according to Fig. 22.

520,900 sq. mi. of solar farms x 122.8 GWh/sq. mi. of solar farms
÷ 45.7 GWh/sq. mi. of wind farms = 1,399,705 sq. mi. of wind farms

Other issues associated with large-scale wind farms include distribution of royalties to benefiting vs. impacted landowners; safe setback distances from inhabited buildings and roads; bird and insect kills; farm land compaction during construction and loss of productivity; interference with pivot irrigation systems and aerial spraying; impact on aviation, especially general aviation; impact on pollination; impact on soil moisture content; impact on crop moisture conditions; and a general change in the visual (shadow flicker) and acoustic conditions of the impacted and surrounding farmland. Given the obvious increasing demand for food as the nation’s population more than doubles by 2100, any measurable impact on agricultural output will be a significant issue. As seen in Fig. 20, the heart of the wind power zone is America’s breadbasket states in the central United States.

Offshore wind farms are now being installed around the world because the average wind speed is often greater. As shown in Fig. 20, the United States has belts along its coasts and on the Great Lakes that have substantial wind power potential. The challenge in installing substantial offshore farms is that they impede ship transport, often impact the view from the shore where tourism is important, are more difficult to connect to onshore utility grids, and can require elaborate anchorage systems in deeper waters, especially where hurricanes and/or ice are possible. Consequently, the potential for added wind-electricity generation from offshore farms is likely quite modest.

D. Neither ground-solar nor wind power provide practical solutions for meeting U.S. 2100 energy needs

The net land area required to meet the U.S. 2100 energy needs of a population of 625 million consuming 50 BOE/yr. using ground solar and wind farms is, respectively, 521,000 sq. mi. and 1.4 million sq. mi. This is what is required to equal the 6,500 GW of continuous nuclear-electricity sized to provide the same 2100 energy needs. To help appreciate the impact of the needed land areas, these are illustrated in Fig. 23.

Figure 23 – Comparison of ground solar farm and wind farm new areas needed to meet U.S. 2100 energy needs

An important point to reemphasize is that these are the net land areas, not the gross impacted land areas. The actual impacted land area in each case will be greater due to local terrain; set-asides for parks, roads, existing construction, etc.; local social/political opposition; aviation flight restrictions; availability of electrical power transmission lines, etc. With this understanding, it quickly becomes apparent that neither ground solar nor wind—or a combination of these—will be capable of providing a substantial percentage of the U.S. 2100 non-fossil fuel energy sources.