Keywords

Wind Energy; Potential

4.1 Introduction

Fossil fuels—coal, oil, and natural gas—currently account for close to 80% of total global primary consumption of energy and for the bulk of emissions of the key greenhouse gas CO₂. The nations of the world at the 21st meeting of the Conference of the Parties (COP 21) to the UN Framework Convention on Climate Change (UNFCC) in Paris in December 2015 committed to restrict future greenhouse gas emissions to ensure that the consequent increase in global average surface temperature should be limited to 2°C or less referenced with respect to conditions that applied in the preindustrial era. Meeting this objective will require no less than a sea change in the manner in which the world sources future energy [1]. Physical prospects for growth in hydro and biomass are limited. Nuclear is expensive. Geothermal could play a role though major investments in relevant research and development will be required to realistically evaluate its potential. The best options, given current understanding, involve combinations of wind and solar. This chapter offers an assessment of the overall potential for wind.

The energy absorbed by the Earth from the Sun over the course of a year totals approximately 3.9×10²⁴ J (3.7 million quads where 1 quad=10¹⁵ BTU and 1 BTU=1055.06 J). To place this number in context, the total energy consumed globally by humans amounted to 5.66×10²⁰ J (536.9 quads) in 2014, a little more than 1 part in 10 000 of the total supplied by the Sun. The solar energy absorbed by the Earth is realized primarily initially as heat, manifest in the atmosphere as internal energy (IE) complemented by a source of potential energy supplied by evaporation of water (LE). A fraction of IE (~40%) is converted to potential energy (PE), energy the air includes by virtue of its elevation relative to the surface. The fraction of the energy of the atmosphere manifest in kinetic form (KE)–wind—is extremely small relative to either IE or PE. It accounts for as little as 763.8×10¹⁸ J (724 quads), 0.06% of the total energy content of the atmosphere. It is replaced, however, on average every 6.9 days implying a global annual source of about 4.04×10²² J (38 300 quads), approximately 73 times the total global demand for commercial energy, 466 times global consumption of energy in the form of electricity.

There are two important sources for KE in the atmosphere: one results from work supplied by the force of gravity acting to change the elevation of specific air masses; the second relates to work performed by the force associated with spatial gradients of pressure driving air across isobars, causing it to move from regions of high to regions of low pressure. Dissipation by friction in the near surface environment, at altitudes below about 1 km, is responsible for approximately 50% of the net global sink for KE with the balance contributed by viscous dissipation of small-scale turbulent elements at higher altitudes [2]. Huang and McElroy [3] using a temperature–pressure–wind record based on reanalysis of meteorological data covering the period January 1979 to December 2010, concluded that work supplied by gravitational and pressure forces was responsible for a globally averaged net source of KE equal to 2.46 W m⁻² over this time interval, slightly more in the southern hemisphere, less in the north (2.49 W m⁻² as compared to 2.44 W m⁻²) [3]. As indicated earlier, this source would be sufficient to supply a quantity of kinetic energy significantly greater than the energy implicated in current demand for electricity or even in the demand for energy in all forms. Only a fraction of this global supply could be harnessed of course under realistic circumstances to produce electricity.

The electricity generated from an individual turbine is determined ultimately by the kinetic energy intercepted by the blades of the turbine. This depends in turn on the area swept out by the blades, on the density of the air intercepted by the blades, and on the cube of the wind speed (a factor proportional to the square of the wind speed defining the kinetic energy contained in a given volume of air, an additional factor to specify the rate at which this energy may be delivered to the turbine). In general, the greater the elevation of the rotor and the greater the diameter of the blades, the greater is the potential yield of electricity. Wind speeds increase typically as a function of elevation accounting for the advantage of the first of these considerations. The area intercepted by the blades of the turbine varies in proportion to the square of the diameter of the rotor accounting for the advantage of the second. In practice, there is a range of wind speeds over which a typical turbine may be expected to operate economically. At low speeds, frictional losses would be sufficient to offset any potential production of electricity. If wind speeds are too high, operation of the turbine could be hazardous and the blades are typically feathered to avoid damage. The quantity of electricity produced as a function of wind speed for a particular turbine design is defined in terms of what is referred to as the power curve. Power curves for the two representative turbine designs considered, the GE 2.5 and 3.6 MW models, are displayed in Fig. 4.1.

The yield of electricity anticipated from a particular wind farm depends on a number of factors, including the quality of the wind resource, the design of the turbines included in the facility and their spacing. The choice of spacing reflects a trade-off involving considerations of costs for individual turbines, costs for development of the site, and costs for laying power cables, in addition to costs anticipated for routine operation and maintenance of the facility. Turbines must be spaced to minimize interference in airflow due to interactions among individual turbines. This requires a compromise between the objective of maximizing the power generated per turbine and the competing incentive to maximize the number of turbines sited per unit area.

Two sources of data have been employed in the literature to evaluate the global potential for generation of electricity from wind. One is based on surface and/or sounding measurements of winds. The second makes use of what is referred to as assimilated meteorological data, with wind speeds derived from retrospective analysis of global meteorological data using a state-of-the-art weather/climate model incorporating inputs from a wide variety of observational sources [5] including not only surface and sounding measurements but also results from a diverse suite of measurements and observations taken from a combination of aircraft, balloons, ships, buoys, dropsondes and satellites—in short the gamut of all of the observational data employed to provide the world with the best possible meteorological forecasts, enhanced by application of these data in a retrospective analysis. The former approach was adopted by Archer and Jacobsen in their pioneering early study of potential global wind resources [6]. The latter was favored by Lu et al. and provides the basis for the bulk of the results discussed below [4].

The study by Archer and Jacobsen used data for 2000 incorporating inputs from 7753 individual surface meteorological stations complemented by results from 446 stations for which vertical soundings were available [6]. They restricted their attention to power that could be generated using a network of 1.5 MW turbines tapping wind resources from regions with annually averaged wind speeds in excess of 6.9 m s⁻¹ (wind class 3 or better) at an elevation of 80 m. The meteorological stations employed in their analysis were concentrated to a significant extent in the United States, Europe, and Southeastern Asia. As a consequence, results inferred for other regions are subject to considerable uncertainty. To estimate the wind potential at a particular location, Archer and Jacobson used six empirical functions to develop a best least square fit to wind profiles observed at individual neighboring stations for which information was available from soundings [6]. They used an inverse square approach to average data from the five closest meteorological stations to select input for the wind profile adopted to calculate the power potential at a particular sample location. They argued that the estimates for wind power derived using this approach should be conservative on the low side for two reasons: first, the potential for bias introduced by use of the least square methodology should trend in that direction; and, second, the fact that the meteorological stations employed in the analysis were not selected optimally to capture most favorable wind conditions. They concluded that 20% of the total available global wind power potential could be tapped to provide as much as 123 PW h of electricity annually, seven times total global consumption, comparable to consumption of energy globally in all forms [6].

We outline in Section 4.2 the procedures adopted by Lu et al. [4] in their study of global wind potential. Results are presented and discussed in Section 4.3. Section 4.4 addresses limitations in our ability to respond definitively to the challenges posed in the title to this chapter: to define the ultimate global potential for the generation of electric power using wind.

4.2 Methodology

The Lu et al. study took advantage of a simulation of global wind fields provided by Version 5 of the Goddard Earth Observing System Data Assimilation System (GEOS-5 DAS) [4]. The GEOS-5 analysis employs a terrain-following coordinate system defined by 72 vertical layers extending from the surface to a pressure level of 0.01 hPa (an altitude of approximately 78.2 km) [5]. Pressure levels are selected to resolve features of the atmosphere including both troposphere and stratosphere. Individual volume elements are defined in terms of their horizontal boundaries (latitude and longitude) and by the pressures at their top and bottom. The horizontal resolution of the simulation is 2/3-degree longitude by 1/2-degree latitude (equivalent to approximately 67 km by 50 km at mid-latitudes). The model provides three-dimensional pressure fields at both layer centers and at layer edges, in addition to wind speeds (meridional and zonal) and temperatures at the midpoint of individual layers with a time resolution of 6 hours. The three lowest layers are centered at altitudes of approximately 71, 201, and 332 m. The 6 hour data for the three lowest layers were employed in the Lu et al. [4] analysis using an interpolation scheme described as follows to estimate temperatures, pressures, and wind speeds at 100 m, the hub height for the 2.5 and 3.6 MW turbines considered as representative and illustrative for purposes of the present discussion.

Knowing the values of pressures at the lower and upper edges of individual layers, together with temperatures and pressures at the midpoints of the layers, Lu et al. [4] calculated altitudes corresponding to the midpoints of the layers using an iterative application of the barometric law assuming a linear variation of temperature between the midpoints of the individual layers. The barometric law was applied also to calculate the pressure at 100 m. Wind speeds and temperatures at 100 m were computed using a cubic spline fit to the corresponding data at the midpoints of the three lowest layers.

The power curves reported by the General Electric Company for the turbine models considered here assume an air density of 1.225 kg m⁻³ under conditions corresponding to an air temperature of 15°C at a pressure of 1 atm [7]. To account for the differences in air density at the rotor elevations as compared to this standard, wind speeds in the published power/wind speed curves (Fig. 4.1) were adjusted according to the Eq. (4.1).

$V_{\mathrm{corrected}}={\left(\frac{P}{1.225·R·T}\right)}^{1/3}{V}_{\mathrm{original}}$ (4.1)

(4.1)

where P and T identify air pressures and temperatures at the hub height and R denotes the gas constant, 287.05 Nm (kg K)⁻¹ for dry air.

In estimating the global potential for wind-generated electricity, it will be important to exclude locations for which it would be either impractical or uneconomic to install turbines. To this end, Lu et al. [4] elected to eliminate areas classified as forested, areas occupied by permanent snow or ice, and areas identified as either developed or urban. They were guided in their selection of locations for turbine deployment by data from the Moderate-Resolution Imaging Spectroradiometer (MODIS) instruments included in the payloads of NASA’s Terra and Aqua satellites.

MODIS provides a record of the spatial distribution of different types of land cover observed over the Earth in 2001. Following a classification introduced by the International Geosphere-Biosphere Programme (IGBP), the record identifies 17 categories of land cover including 11 different classes of natural vegetation, 3 classes of developed areas, and 3 classes identifying areas occupied by permanent snow or ice (notably Greenland and Antarctica). It singles out also regions classified as barren, areas with at most a sparse coverage of vegetation, and regions covered by water. It pinpoints regions identified as either urban or heavily developed. The horizontal resolution of the record is approximately 1 km by 1 km. Lu et al. [4] used the MODIS record to exclude from their analysis areas classified as forested, areas occupied by permanent snow or ice, areas covered by water and areas identified as either developed or urban. Environments identified by Lu et al. as inappropriate for wind resource development are indicated in Fig. 4.2.

Topographic relief data for both land and ocean areas were derived using the Global Digital Elevation Model (GTOPO30) of the Earth Resources Observation and Science (EROS) Data Center of the US Geological Survey (USGS). The spatial resolution of this data source for offshore environments (bottom topography) as applied here is approximately 1 km by 1 km [8]. A number of factors conspire to limit the development of offshore wind farms. Aesthetic considerations, e.g., have restricted development of wind resources in the near shore environment in the United States (the Cape Wind controversy as a case in point [9]) although objections to near shore installations in Europe appear to have been less influential. There is a need further to accommodate requirements for shipping, fishing, and for wildlife reserves, and to minimize potential interference with radio and radar installations. To account for these limitations, Musial and colleagues, in studies of the offshore wind power potential for the contiguous United States, chose to exclude placement of wind farms within 9.3 km (5 M where M refers to nautical mile) of shore and to limit development to 33% of the area between 9.3and 37 km (5 and 20 M) offshore, while expanding potential development to 67% of the area between 37 and 92.6 km (20 and 50 M) [9–12].

The expense of installing wind turbines offshore generally increases as a function of water depth and as a function of distance from shore. General Electric recommends that its turbines should be installed using state-of-the-art monopole structures fixed to the seabed for water depths less than 20 m. Water jacket tripod or quadrapod structures can support towers in waters up to 50 m [13,14]. For greater depths, it is necessary to resort to floating structures using technology developed by the oil and gas extraction industry [15]. Experience with the use of floating structures in the wind power business is relatively limited to date but its development is not expected to pose insuperable problems in the future, although it will certainly be more expensive.

Lu et al. [4], followed by Dvorak et al. [14], considered three possible regimes for offshore development of wind power defined by water depths of 0–20, 20–50, and 50–200 m. Somewhat arbitrarily, they limited potential deployment of wind farms to distances within 92.6 km (50 M) of the nearest shoreline, assuming that 100% of the area occupied by these waters could be available for development. A schematic summary of the approach they adopted in calculating wind power potential both onshore and offshore is presented in Fig. 4.3.

A further consideration in estimating the total global potential for wind-generated electricity concerns the criteria adopted to determine the spacing of turbines in particular wind farms. Restricting downstream interturbine wake power loss to less than 20% requires a downwind spacing of more than seven turbine rotor diameters with cross-wind spacing of at least four diameters [6,16]. Applying this constraint to the 2.5 MW GE turbines [17] (rotor diameter 100 m, radius 50 m) selected as representative for onshore wind deployment implies an interturbine spacing of 1 per 0.28 km². Given the much higher expense for development of offshore wind farms, Lu et al. [4] elected to impose a greater relative spacing in this case in order to limit interturbine interference to less than 10%. This translates to a requirement for an interturbine spacing of 5×10 rotor diameters. Assuming deployment of GE 3.6 MW turbine [18] (rotor diameter 111 m, radius 55.5 m), positioning of turbines in this case should be restricted to 1 every 0.62 km². The results presented in what follows reflect these assumptions.

4.3 Results

4.3.1 Global Perspective

We restrict attention in what follows to locations in which wind conditions are projected to allow for the turbine choices considered here to function on an annual basis with capacity factors (CFs) of no less than 20%.

Results on a country-by-country basis are summarized in Fig. 4.4A and B for onshore and offshore environments, respectively. Placement of the turbines onshore and offshore was restricted as discussed earlier. Table 4.1 presents a summary of results for the 10 countries identified as the largest national emitters of CO₂ [19,20]. The data included here refers to national reporting of CO₂ emissions of 2012 and electricity consumption for these countries in 2011. Wind power potential for the world as a whole and for the contiguous United States is summarized in Table 4.2.

Table 4.1

Onshore and Offshore Wind Potential for the 10 Countries Identified as the Largest National Emitters of CO₂ [4]

No.	Country	CO2 Emission/(106 Metric Tonnes)	Electricity Consumption/(TW h)	Potential Wind Energy/(TW h)
				Onshore	Offshore	Total
1	China	8547.7	4207.7	39 000	4600	44 000
2	United States	5270.4	3882.6	74 000	14 000	89 000
3	India	1830.9	757.9	2900	1100	4000
4	Russia	1781.7	869.3	120 000	23 000	140 000
5	Japan	1259.1	983.1	570	2700	3200
6	Germany	788.3	537.9	3200	940	4100
7	South Korea	657.1	472.2	130	990	1100
8	Iran	603.6	185.8	5600	–	5600
9	Saudi Arabia	582.7	211.6	3000	–	3000
10	Canada	499.1	551.6	78 000	21 000	99 000

Note: CO₂ emission for 2012 and electricity consumption for 2011.

Source: Data from Boden TA, Andres RJ, Marland G. Preliminary 2011 and 2012 global & national estimates. In: Fossil-fuel CO2 emissions. Oak Ridge, TN: Carbon Dioxide Information Analysis Center; 2013. p. 4 [19] and US EIA. International energy outlook. Washington, DC: U.S. Energy Information Administration; 2013. p. 312 [20].

Table 4.2

Annual Wind Energy Potential for Installations Onshore and Offshore for the World as a Whole and for the Contiguous United States

Areas		Worldwide		Contiguous United States
		No CF Limitation	20% CF Limitation	No CF Limitation	20% CF Limitation
Energy (onshore areas)/(PW h)		1100	690	84	62
Energy (offshore areas)/(PW h)	0–20 m	47	42	1.9	1.2
	20–50 m	46	40	2.6	2.1
	50–200 m	87	75	2.4	2.2
Energy total/(PW h)		1300	840	91	68

Analysis assumes loss of 20% and 10% of potential power for onshore and offshore, respectively due to interturbine interference [4].
Note: All data assume offshore location distance within 92.6 km (50 M) of the nearest shoreline.

If the top 10 CO₂ emitting countries were ordered in terms of wind power potential, Russia would rank number one, followed by Canada with the United States in third position. There is an important difference to be emphasized, however, between wind power potential in the abstract and the fraction of the resource that is likely to be developed when subjected to realistic economic constraints. Much of the potential for wind power in Russia and Canada is located at large distances from population centers. Given the inevitably greater expense of establishing wind farms in remote locations and potential public opposition to such initiatives, it would appear unlikely that these resources will be developed in the near term. Despite these limitations, it is clear that wind power could make a significant contribution to the demand for electricity for the majority of the countries listed in Table 4.1, in particular for the four largest CO₂ emitters—China, the United States, India and Russia. It should be noted, however, the resource for Japan is largely confined to the offshore area, 82% of the national total. To fully exploit these global resources will require, inevitably, significant investment in transmission systems capable of delivering this power to regions of high load demand. Results for the contiguous United States and for China will be discussed in more detail in the following sections.

The electricity that could be generated potentially on a global basis using wind, displayed as a function of an assumed CF cutoff on installed turbines, is presented in Fig. 4.5A and B for onshore and offshore environments, respectively. The results in Fig. 4.5A suggest that total current global consumption of electricity could be supplied by wind, while restricting installation of land-based turbines to regions characterized by most favorable wind conditions, regions where the turbines might be expected to function with CFs greater than 53%. If the cutoff CF was lowered to 36%, the energy content of electricity generated using wind with land-based turbines globally would be equivalent to total current global consumption of energy in all forms. Cutoff CFs needed to accommodate similar objectives using offshore resources would need to be reduced as indicated in Fig. 4.5B. To place these considerations in context, we would note that CFs realized by turbines installed recently in the United States (in 2004 and 2005) have averaged close to 36% [21].

4.3.2 US Perspective

An estimate of the electricity that could be generated for the contiguous United States on a monthly basis (subject to the siting and capacity limitations noted earlier) is illustrated for both onshore and offshore environments in Fig. 4.6. Results presented here were computed using wind data for 2006. Not surprisingly, the wind power potential for both environments is greatest in Winter, peaking in January, lowest in Summer, with a minimum in August. Onshore potential for January, according to the results presented in Fig. 4.6, exceeds that for August by a factor of 2.5: the corresponding ratio computed for offshore locations is slightly larger, 2.9.

Fig. 4.6 includes also monthly data for consumption of electricity in the United States during 2006. Demand for electricity exhibits a bimodal variation over the course of a year with peaks in Summer and Winter, minima in Spring and Fall. Demand is greatest in Summer during the air-conditioning season. Summer demand exceeds the minimum in Spring/Fall demand typically between 25% and 35% on a US national basis depending on whether Summers are unusually warm or relatively mild. The correlation between the monthly averages of wind power production and electricity consumption is negative. Very large wind power penetration can produce excess electricity during large parts of the year. This situation could allow options for the conversion of electricity to other energy forms. Plug-in electric vehicles, e.g., could take advantage of short-term excesses in electricity system, while energy rich chemical species such as H₂ could provide a means for longer term storage.

Potential wind-generated electricity available from onshore facilities on an annually averaged state-by-state basis is presented in Fig. 4.7A. Note the high concentration of the resource in the central plains region extending northward from Texas to the Dakotas, westward to Montana and Wyoming, and eastward to Minnesota and Iowa. The resource in this region, as illustrated in Fig. 4.7B, is significantly greater than current local demand. Important exploitation of this resource will require, however, a significant extension of the existing power transmission grid. Expansion and upgrading of the grid will be required in any event to meet anticipated future growth in electricity demand. It will be important in planning for this expansion to recognize from the outset the need to accommodate contributions of power from regions rich in potential renewable resources, not only wind but also solar. The additional costs need not, however, be prohibitive [21]. ERCOT, the operator responsible for the bulk of electricity transmission in Texas, estimates the extra cost to transmit up to 4.6 GW of wind-generated electricity at about $180 per kilowatt, approximately 10% of the capital cost for installation of the wind power-generating equipment [22].

An important issue relating to the integration of electricity derived from wind into a grid incorporating contributions from a variety of sources relates to the challenge of matching supply with load demand incorporating a contribution from supply that is intrinsically variable both in time and space and subject to prediction errors. This challenge can be mitigated to some extent if the variations of wind sources contributing to an integrated transmission grid from different regions are largely uncorrelated. An anomalously high contribution from one region can be compensated in this case by an anomalously low contribution from another. To investigate the significance of this potential compensation, Lu et al. [4] examined the covariance of wind resources from three specific regions, one in Montana, the second in Minnesota, the third in Texas, as indicated in Fig. 4.8. Analysis of 6 hour averaged potential wind-generated supplies of electricity from the three regions over the four seasons, Winter, Spring, Summer, and Fall, yielded the results summarized in Table 4.3. Contributions from the three regions are essentially uncorrelated during the Winter months (October–March) with r values of less than 0.07. Correlation coefficients (r values), however, are relatively high in Summer (July–September) with values ranging from 0.28 (Montana vs Texas) to 0.37 (Montana vs Minnesota) with intermediate values in Spring. The analysis suggests that wind power could make a relatively reliable contribution to anticipated base load demand in Winter. It may be more difficult to incorporate wind power resources into projections of base load demand for other seasons, particularly for Summer.

Table 4.3

Correlations of Wind Power Potential Between Selected Regions of Montana (MT), Minnesota (MN), and Texas (TX) in Different Seasons for 2006 [4]

Correlation Coefficient (r)	January–March	April–June	July–September	October–December
MN–MT	0.027	0.11	0.37	−0.15
MN–TX	0.069	0.29	0.29	−0.060
MT–TX	0.065	0.26	0.28	−0.0024

4.3.3 China Perspective

McElroy et al. [23] applied the assimilated meteorology data resource as described earlier to assess also the potential for wind-generated electricity in China. The approach they followed in this case was generally similar to that adopted for the global and US applications described by Lu et al. [4]. In particular, they elected to exclude as possible sites for turbine deployment forested areas, areas occupied by permanent snow or ice, areas covered by water, and areas identified as either developed or urban. They excluded in the Chinese application also land areas with slopes larger than 20% [21]. Recognizing that turbine sizes installed in China are generally smaller than those favored in the United States, they chose to focus their study on deployment of a suite of 1.5 MW turbines, selecting for this purpose the GE 1.5 MW xle design [24]. The hub height for this model is at 80 m; the rotor diameter measures 82.5 m. The spacing between turbines in representative wind farms was selected similar to that adopted for wind farms installed in Inner Mongolia, 9 rotor diameters in the downwind direction, 5 rotor diameters in the direction perpendicular to the prevailing wind (9D×5D), slightly larger than the spacing of 7D×4D, adopted by Lu et al. [4]. Overall power loss due to turbine–turbine interactions with the spacing assumed in the China application is estimated at about 10% [25].

The spatial distribution of CFs evaluated for deployment of the 1.5 MW turbines considered here is illustrated in Fig. 4.9. CF defines the fraction of the rated power potential of a turbine that is actually realized over the course of a year given expected variations in wind speed. CF values for wind farms deployed in Inner Mongolia, as illustrated, e.g., in Fig. 4.9, are estimated to reach values as high as 40% indicating that 1.5 MW turbines installed in this region could potentially provide as much as 5.26 GW h of electricity over the course of a year. Wind conditions are notably favorable, and CF values are consequently large, over extensive regions of northern China (Inner Mongolia, Heilongjiang, Jilin, and Liaoning) and in parts of the west (Tibet, Xinjiang, Qinghai, and Gansu). Wind farms deployed recently in the United States have achieved operational CFs as high as 48%, with an average of close to 35% [26]. By way of comparison, CFs for wind farms installed in China have been significantly lower than for the United States, close to 23% on average [27]. The relatively low operational performance for wind farms in China is attributed to a combination of factors: lower quality of the largely domestically produced turbines deployed in China as compared with turbines available on the international market; bottlenecks introduced by limitations imposed by the existing Chinese electricity grid; and suboptimal siting of wind farms due to inadequate prior screening of potentially available wind resources [28].

Electricity that could be generated from wind irrespective of price, restricted however to installations capable of operating with CFs greater than 20%, is illustrated for the existing seven electric grid areas of China in Fig. 4.10. The figure includes also results expressed as ratios with respect to the current production of electricity in these grid regions. The data displayed here suggests that a suite of 1.5 MW turbines deployed in onshore regions with favorable wind resources could provide potentially for as much as 24.7 PW h of electricity annually, more than seven times current national Chinese consumption.

Demand for electricity in China is spread more evenly throughout the year as compared to the United States where demand peaks in Summer. The pattern for China reflects the fact that the largest fraction of electricity in the country is used by industry (70%) as compared to only 29% for industry in the United States where demand for electricity is spread more evenly among residential, commercial, and industrial usage. Incorporating base load sources of electricity from coal-fired power plants poses relatively minor problems for grid managers charged with matching supplies of electricity with demand. Adjusting to an important, intrinsically variable, supply such as that from wind will require a more complex, and consequently more costly, grid management protocol.

Operators of an electric utility face a formidable and continuing challenge to ensure that production of electricity is targeted in real time to meet projected demand. In a typical power system, nuclear and coal-fired systems provide sources of what is referred to as base load power. That is to say, the assumption is that these systems will operate essentially continuously, with minimal opportunity to respond to either increases or decreases in demand. Typically, gas-fired systems, which can be turned on or off rapidly, provide the flexibility needed to react to changes in demand. Accommodating an input of power from an intrinsically variable source such as wind poses a particular problem for the orderly operation of a complex electric utility network.

Integration of wind energy into China’s coal-heavy electricity system presents significant challenges owing to wind’s variability and the grid’s system-wide inflexibilities. As indicated in a recent study [28], China has greater capacity for wind installation compared to the United States (145.1 vs 75.0 GW) but generates less electricity from wind (186.3 vs 190.9 TW h). A study by Davidson et al. suggested a potential production of 2.6 PW h per year by 2030 [29]. Although this represents 26% of total projected electricity demand, it is only 10% of the total estimated physical potential of wind resources in the country. Increasing the operational flexibility of China’s coal fleet would allow wind to deliver nearly three-quarters of China’s target of producing 20% of primary energy from nonfossil sources by 2030.

4.4 Concluding Remarks

The discussion to this point has sought to estimate the quantity of electricity that could be generated by selective placement of state-of-the-art wind turbines in regions judged suitable for their deployment. As indicated, the wind data employed in this analysis were derived from retrospective analysis of past meteorological conditions. In this sense, the present analysis may be interpreted as identifying the electricity that could have been produced from turbines installed at some point in the past when wind conditions may have been similar, and remained similar, to those identified in the database adopted here. The past is of course at best an imperfect prologue for the future. But, in planning for the future it may be the best option at our disposal.

Reservations that should be noted in addressing the charge indicated in the title of this chapter—to define the global potential for wind-generated electricity—include the following. Placement of a concentration of wind turbines at a particular location could have the potential to alter local and potentially even regional wind conditions. An extensive deployment of wind farms could have an impact on the budget of atmospheric kinetic energy leading to a potentially consequential change in the circulation of the global atmosphere. And responding to increasing concentrations of greenhouse gases, climate and wind conditions in the future may differ significantly from conditions that prevailed in the past. Quantitative projections for future wind power potential will be subject therefore to a level of inevitable unavoidable uncertainty.

The impact of wind farms on local meteorological conditions has been explored in a number of recent studies. Zhou et al. [30]. used satellite data covering the period 2003–11 to analyze the response of regional surface temperature to the development of a wind farm in Texas. They found evidence for a significant upward trend in surface temperature, by as much as 0.72°C per decade, particularly at night and especially in the immediate neighborhood of the wind farms. Roy and Traiteur [31] found a similar pattern in their study of the response of temperatures to the development of a wind farm in San Gorgonio, California. They reported evidence for a statistically significant increase in temperature, by about 1°C, at an elevation of 5 m downwind of the wind farm at night. The increase persisted through the early morning, followed by modest cooling during the day. They suggested that the impacts of wind farms on local weather could be minimized by modifying the design of rotor systems, or by siting wind farms in regions defined by high levels of natural turbulence. They further identified the Midwest and Great Plains regions of the United States as ideal for the placement of low-impact wind farms.

If the entire demand for electricity in the United States were to be accommodated by wind, the accompanying sink for kinetic energy would amount to approximately 6% of the sink contributed naturally by surface friction over the entire contiguous US land area, 11% for the sink identified with the area indicated in the foregoing as most favorable for wind farm development. The impact on the circulation of the atmosphere of potentially major commitments to wind power was explored in a number of recent studies, notably by Kirk-Davidoff and Keith [32] and Keith et al. [33]. They concluded that exploitation of wind resources at high levels of penetration might be expected to lead to significant changes in the circulation of the atmosphere, even in regions remote from the location of the involved turbines. They argued that the budget for the global inventory of atmospheric kinetic energy is regulated primarily by processes on the input side of the ledger rather than by the sink. An increase in friction resulting from the operation of large numbers of power-generating turbines could be compensated in this case, they argued, by a decrease in the dissipation of momentum by friction elsewhere. The global average surface temperature, they concluded, would not be expected to change significantly in the face of a large investment in wind-generated electricity. Temperatures at high latitude could decrease to a modest extent in response to an anticipated decrease in the efficiency of meridional heat transport. The impact could be viewed as positive in this case, an offset to some extent for the amplified warming projected to arise for this environment in response to the human-induced increase in the concentration of greenhouse gases.

The impact on the circulation of the atmosphere of a large-scale investment in wind farms was investigated also by Miller et al. [34] and by Marvel et al. [35]. Employing a simple parameterization approach to simulate the influence of turbine operations as a sink for atmospheric momentum, Miller et al. [34] concluded that turbines distributed uniformly over the Earth’s surface could harvest kinetic energy sustainably at a rate ranging up to as much as 400 TW. If the turbines were deployed at an elevation of 100 m, the yield could be as great as 1800 TW. Using an alternate approach to parameterize the sink for momentum associated with the exploitation of wind resources, Jacobson and Archer [36] concluded that as the number of wind turbines increased over a large geographic region, extraction of power should first increase linearly, converging eventually to a limit, estimated as in excess of 250 TW for turbines sited at 100 m, rising to 380 TW for turbines deployed at an altitude of 10 km.

There is a notable discrepancy between these various estimates of wind potential. Adams and Keith [37] addressed the issue using a mesoscale model. They concluded that the generation of power from wind should be limited to an average of about 1 W m⁻² for facilities distributed over an area of approximately 100 km². They argued further that the results obtained using a mesoscale model should provide a useful guide to what might be expected on the basis of a more complete global model. The assertion, however, remains to be demonstrated.

Modern wind turbines are designed to operate effectively for life cycles ranging up to 25 years or even longer. Predictions of wind power for the next 25 years, including the need to anticipate the impact of intrinsic variability, will pose a challenge for prospective investors. Global and regional climate models have difficulty in accounting for historical trends in wind regimes. There is little reason to believe that they will be more successful in predicting the future. Pryor et al. [38], on the basis of existing research, argued that the changes in mean wind speeds and energy density anticipated for the future are unlikely to exceed the year-to-year variability ($\pm$15%) observed most recently over much of Europe and North America. Surface winds have declined in intensity in China, the Netherlands, the Czech Republic, the United States, and Australia over the past few decades [39–42]. The precise cause of this decline is uncertain. Vautard et al. [43] analyzed the extent and potential cause for the changes in surface wind speeds observed over northern mid-latitudes between 1979 and 2008, using data from 822 surface weather stations. They indicated that surface wind speeds declined by 5%–15% over almost all continental areas at northern mid-latitudes with the decrease greatest at higher wind speeds. In contrast, upper-air winds inferred from sea-level pressure gradients, and winds derived from weather reanalyses, exhibit no such trend. It has been suggested that an increase in surface roughness resulting from increases in biomass and related changes in land cover over Eurasia could account for as much as 25%–65% of the decrease in surface winds observed over this region.

Huang and McElroy [3], using assimilated meteorological data for the period January 1979–December 2010, investigated the origin of wind energy from both mechanical and thermodynamic perspectives. Their results indicate an upward trend in kinetic energy production over the past 32 years, suggesting that wind energy resources might increase in a warming climate. They highlighted further the fact that the total kinetic energy stock of the atmosphere displays significant interannual variability, responding notably to the changing phases of the El Niño–Southern Oscillation (ENSO) cycle. The potential for wind as a source of electricity at any particular location may be expected thus to vary not only on the long term but also interannually in response to natural fluctuations in the circulation of the atmosphere.

The overall conclusion from this chapter is that wind resources on a global scale could accommodate a large fraction of present and anticipated future demand for electricity. Concentration of facilities in specific regions might be expected to contribute to a change in prevailing local meteorological conditions. This change is unlikely, however, to be sufficiently disruptive as to offset the advantages that could be realized from the concentration in the first place. Generation of electricity by capturing kinetic energy from the wind may be considered as an additional contribution to the surface friction that serves as the natural offset to the atmosphere’s global production of kinetic energy. At high levels of penetration, wind facilities could have an appreciable influence on the budget of this important quantity: climate might be expected to adjust accordingly. Given foreseeable near term expansion of wind systems, however, this is unlikely to pose a serious problem. The most important limitation for future growth is likely to involve rather the challenge of responding to the intrinsic variability of the input from wind, compounded by the fact that this source may not be matched ideally to patterns of power demand.

Acknowledgments

The work was supported by the State Environmental Protection Key Laboratory of Sources and Control of Air Pollution Complex, Collaborative Innovation Centre for Regional Environmental Quality, the National Key R&D Program “Formation mechanism and control technology of air pollution” (2016YFC0208900), and the Volvo Group in a research project of the Research Center for Green Economy and Sustainable Development, Tsinghua University. It was also supported by the Harvard Climate Change Solutions Fund and the Harvard Global Institute.