Minnesotans For Sustainability^©

[ Home ] [ Up ] [ Feedback Please ] [ Table of Contents ] [ Search MFS ] [ MFS News ]

Sustainable Society:  A society that balances the environment, other life forms, and human interactions over an indefinite time period.

Net Energy from Wind Power

Gene Tyner, Sr.¹*
January 2002

Part 1 – Overview
    Methodology
    Table 1 – Estimated Net Energy for U.S. Economy
Part II – Net Energy Analysis
    Figure 1 - Btu's to Produce 1$GDP
A Critique of Standard Economic Reasoning
Historical Energy Consumption Information
    Figure 2 - U.S. Energy Disposition
    Figure 3 - Energy Flow Diagram
Part III – Net Energy Analysis of Wind Power
Fractions of U.S. Electricity Supply
    Table 2 – Assumptions Used in the Fractions of US Electricity Models.
    Table 3a - Lifecycle Static Net-Energy Analysis
    Table 3b – Dynamic 100-Year Net-Energy Analysis
    Figure 4 - Annual Net Energy - 1/4th of US Electricity
    Figure 4a - Annual Net Energy - All US Electricity
    Figure 4b - Cumulative Net Energy - 100 Years
    Table 4 – Summary of Cost x E/GDP Analyses – 100-yr. Scenarios
Analysis, continued
     Table 5 – Variable Used in Maximum Generation Models
     Figure 5 - Annual Net Energy - Maximum Wind Optimistic
    Table 6 – Summary of Cost x E/GDP Analyses – 100-yr. Scenarios
Summary and Conclusions
    Dilemma
    Should large Wind systems be installed as a matter of national policy?
Appendix
References

Part 1 – Overview

This paper is part of a continuing effort to refine estimates net-energy yield of wind power in the United States. Additional details and explanations are included in the appendix.

Probably, the most counted-on substitute for fossil fuels has been the generation of electricity from nuclear power (fission, breeder reactors, and fusion;) but there is uncertainty about ultimate costs, safety, the efficacy of the undertaking, the risk of serious environmental consequences, the cost of decommissioning of the systems, safe disposal of long-term radioactive waste, the associated impact on global ecosystems, and now the terrorist threat has arisen in the United States. A terrorist attack on a nuclear facility could possibly render a large area of the United States uninhabitable for a long period of time. The likelihood of a catastrophic accident increases as large numbers of nuclear plants are operated. Further, the availability of Uranium for fuel, alone, may be a limiting factor on the long-term viability of nuclear power.²

Others are suggesting other alternative forms of energy such as wind, photovoltaic, hydrogen, biomass, ad infinitum, to solve the dilemma of fossil-fuel depletion. All candidates for substitution must be subjected to careful and comprehensive net-energy scrutiny to insure that we do not pour energy “down a rat hole” and promise more than can be delivered.

According to the American Wind Energy Association (AWEA)³ “installed wind energy generating capacity totaled about 2550 MW in year (2000) and generated about 5.5 billion kWh of electricity –less than 1% of U.S. electricity generation. By contrast, the total amount of electricity that could be generated from wind in the United States has been estimated at 10,777 billion kWh annually.

The analyses in this paper are divided into two major categories and three subcategories in each of the major ones. The question is asked: How much net energy can be expected from wind systems if wind power were to become a major supplier of electricity for the United States? Following is an outline of these scenarios:
 

1.  Portions of total U.S. electricity supply (Y2000 production is used as base line) by wind power:
        a.  25% scenario.
        b.  50% scenario.
        c.  100% scenario.

2.  Maximum construction of wind machines to maximum capability claimed by AWEA (3 times that generated in Y2000.) In these scenarios I have assumed that the entire capability of the reported (AWEA) potential is constructed and operated and estimated net energy using optimistic to less optimistic conditions:

        a.  Optimistic scenario.
        b.  More Likely scenario.
        c.  Less Optimistic scenario.

Representative “static analysis” (lifecycle analysis of one plant) and dynamic analysis (analyses of total wind-powered systems over time) models will be shown. Results of all analyses will be shown in summary tables.

Methodology:

The relationship between Lifecycle Cost, Energy Consumption and Gross Domestic Product is used as a surrogate variable to estimate the amount of Energy Required to Get Energy (Input Energy) and compared Input Energy with estimated output energy to arrive at Net Energy – or as some others have termed Energy Return On Investment (EROI.)

It is a relatively simple procedure:

(1)  Estimate Lifecycle System Cost (LSC) of an energy-transformation system,
(2)  Multiply LSC times a selected ratio of annual energy consumption (E) to annual consumption of Gross Domestic Product (GDP),
(3)  Compare the result to the expected energy output from the energy system under consideration.

In the current study complete wind-powered electricity generating systems are analyzed, including construction of all plant, including transmission and distribution plant, lifecycle operation and maintenance, and, finally, disposition of residual wastes after the plant has been retired. Table 1 is exemplary of how this methodology works out for Total U.S. Energy Consumption for selected years from 1980 to 1997.

Table 1 – Estimated Net Energy for U.S. Economy

In order to estimate the total financial outlay for energy, total energy consumed (Line 1.) is multiplied times the average cost per Btu. The results are shown on Line 3. Line 3 is multiplied by line 4 to estimate the amount of energy that was required to get the total energy shown on line 1. Net Energy is found by subtracting Line 5 from Line 1 and annual Net Energy results for selected years are shown on Lines 6 and 7. Percent available (amount of energy available to the rest of the economy) is shown on Line 8. The Output/Input ratio (Line 1 divided by Line 5) is shown on Line 9.

It was striking to find that the “% Available” energy is uniformly 89-90% for the selected years with an average Output/Input ratio of approximately 9 to1. Total Energy Consumed includes all forms of energy and is expressed in BTU’s as reported by the U.S. Department of Energy.

Part II – Net Energy Analysis:

Many net-energy studies (EROI, Pay Back) only count as Input Energy the “direct” fuels closely associated with the latent processes of construction and operation of an energy-transformation systems and ignore the energy embodied in all of the in all the processes leading up to the construction, operation and delivery of energy to final users. Such procedures are sort of analogous to marginal cost analyses in economics. Consequently, many studies of nuclear power show high output/input ratios.⁴

A reason for exclusion may be the lack of data to know what all was involved in the historic processes. Attempts to do this have been classified as Net-Energy Process Analyses.

It is argued that a complete net energy analysis should attempt to account for all of the energy that was required to put an energy form in place (of end use.) It is important to have a feel for the capability of an energy-transformation entity to continuously sustain and reproduce itself from it’s own output and leave a sufficient residual for the rest of society to carry on the needs of an industrial society. In this study a zero growth, constant-dollar level of economic output is assumed with no real growth in the economy for 100 years.

It is assumed that all money payments made by the energy-transformation systems are essential to the system’s existence. It is then argued that there is a direct (approximate) relationship between monies paid and energy consumed. This relationship offers a more comprehensive departure point for estimating the Net Energy to be expected from energy-transformation systems than can be obtained in most other approaches to energy analysis (except for Odum’s eMergy Analysis.)⁵ The ratio of Annual Energy Consumption (E) and Annual Gross Domestic Product (GDP), i.e. the E/GDP Ratio, is an average of how much energy is depleted to produce one dollar of GDP – in this case $1996.

Figure 1 indicates that less energy is required in recent years than previous years to produce a dollar of GDP. Increased efficiency accounts for part of this drift but Cleveland and Kaufman attribute most of it to energy-quality changes.⁶ All analyses in this study assume 10,000 Btu’s are needed to produce a dollar of GDP.

Figure 1 - Btu's to Produce 1$GDP ($1996)

(Click image to enlarge.)

A Critique of Standard Economic Reasoning:

The reasoning of standard economists has dominated the world perspective on future energy supply, and hence energy policy, or lack thereof. On the demand side, it may be correct to reason that rising energy prices may alter human behavior to behave differently, such as conservation or change in lifestyle. On the supply side, however, standard economic theory holds that if any good is in short supply (and many standard economists regard energy as another ordinary good,) rising prices will trigger entrepreneurial ingenuity to provide an additional supply of the same good or an acceptable substitute. While this is acceptable reasoning for ordinary goods (as long as energy and resources are available,) energy is not an ordinary good. Various energy forms (petroleum, natural gas, wood, electricity etc.) are somewhat substitutable for each other, but nothing can substitute for energy, per se. Finally, ordinary goods can only be made available by depleting some form of energy. It takes energy to get energy. Standard economists appear to not worry about energy or resource depletion because of their faith in markets to solve all shortage problems. This seems more like religion than science.

The First Law of Economics should be: Goods and services cannot be made available unless energy is depleted. The existing tools, factories, infrastructure and housing stock of the current civilization, including energy transformation and distribution systems, was initially put in place with relatively cheap and abundant energy (energy with a high net-energy yield.) As rich fossil-fuel stocks are depleted, more energy is needed to get energy, consequently, relatively less will be available for production of other goods and services. Cleveland and Kaufman have estimated the “Net-Energy From the Extraction of Oil and Gas in the United States, 1954-1957.⁷

Energy-acquisitions systems also enjoy the benefits of the in-place infrastructure of society. The existing U.S. infrastructure set (waterways, railways, roadways, capital-in-place, housing stock, airways, ad infinitum) gradually deteriorates, wears out, and requires maintenance and replacement. Replacements require further energy depletion and as time passes, only sources with less net energy are likely to be available to do this.

Since infrastructure serves energy-acquisition systems, it seems reasonable to assume that net-energy yield will decline accordingly (and more gross energy will be required to produce a dollar of GDP.) Of course, as society accumulates information, we may learn to use energy and materials more efficiently, but all efforts to increase efficiency are not without an energy cost of achieving that efficiency.

Recycling, although a reasonable practice is only achieved by an input of energy and materials.

Now, a bold new energy requirement has arisen in the Age of Terrorism: The Energy Required to Maintain order in a very unstable world!

Further, the current set of energy types (coal, natural gas, petroleum, etc.) has a corresponding set of end-using technologies (automobiles, ships, trains, automobiles, aircraft, manufacturing and communications systems, space-heating systems, ad infinitum.) Adapting any new form of energy — say an all-electric system— requires a different set of end-user equipments. Much energy would be required to make such transition(s).

Despite relatively low prices of current energy supply, it is clear that world fossil-fuel supplies will ultimately be depleted to a point where it is no longer practical to mine them. At some point, more energy will be required for recovery and transformation than can be realized from them. M. King Hubbert estimated that about 80% of world petroleum supply would be consumed between 1960 and 2030; 80% of natural gas between 1934 and 1999; and 80% of coal between 2000 and 2300. In 1976, Fowler revised and extended this analysis with similar conclusions. Duncan has placed the peak of per-capita energy production as 1979 that has precipitously declined from 1979 to 1999.⁸ Colin Campbell provides an updated Hubbert-Type of discussion.⁹ A more recent work is by Kenneth S. Deffeyes, Hubbert’s Peak, The Coming Oil Shortage.

Some standard economists explicitly held that capital could be substituted for energy. While this is true for individual activities,¹⁰ capital is only made available in the overall system by a corresponding depletion of some form of energy and materials and therefore, substitution of an alternate good for energy does not really occur in a broad sense. Energy is an extra-ordinary, essential substance having unique properties for which there is no substitute.¹¹

Historical Energy Consumption Information

Figure 2 depicts Total-Energy Consumption, Energy-System Losses, and End-Use Consumption from1973 to 2000. It should be noted that End Use is approximately 75% of total consumption.

Figure 2 - U.S. Energy Disposition

Figure 3 is a duplicate of the Department of Energy Information Agency Diagram on the production and consumption of Electricity for Year 1999. Figures are Quadrillion Btu’s of Energy (Q). It should be noted that it takes about 3 Btu’s of Fossil Fuel to provide 1 Btu in the form of electricity.¹²

Figure 3 - Energy Flow Diagram

Figure 3, above, shows that 33.32 Quads of heat energy is consumed to provide 11.14 Quads (in Btu’s) of Electricity to end users. It is recognized that electricity is a “high-quality” energy form, with unique qualities for lighting, operating motors, communications systems, computers, i.e., for electricity-specific applications. However, it should be recognized that when a kilowatt-hour of electricity is needed for heat (for example, space heating) only 3413 Btu’s of heat is available from each kilowatt hour of electricity, i.e., quality may be less material. Petroleum, for example, is superior in quality to electricity for powering our current transportation systems (aircraft, automobiles, locomotives, shipping, etc.) It will take some doing to have an all-electric transportation system.

Part III – Net Energy Analysis of Wind Power

The Cost x (E/GDP) Model is as follows:

(1)  Estimate the total lifecycle monetary cost of a single system.
(2)  Multiply that cost times a selected ratio of Annual Energy Consumption to Annual Gross Domestic Product (E/GDP Ratio) to estimate the input energy needed to get energy.
(3)  Compare Estimated Output to Estimated Input.
(4)  Calculate the Expected Net Energy from constructing a large number of individual systems over time (Dynamic Net Energy Analysis.)

Among other things, the accuracy of the model is limited by the inability of market price and associated GDP accounting procedures to account for “externality problems” such as impact on and services from environmental resources, distribution of income and wealth, social justice (who benefits and who pays), monopoly effects, joint production, etc. The model also involves substantial averaging of data, necessarily, and is thereby a broad (but inclusive) method. It is argued that most other net-energy methodologies (except Howard Odum’s eMergy procedure) exclude energy costs that are essential to understanding the viability of any proposed substitute for fossil fuels.

Most net-energy analyses only estimate the net-energy available at the point of initial distribution (wellhead for oil and gas, bus-bar for electricity, etc.) and do not include the energy cost of transmission and distribution and some did not include the energy cost of plant decommissioning and disposition of waste; and as previously stated, many only include the direct and latent consumption of energy in the construction and operation of the plants. Importantly, they did not consider the embodied-energy consumption associated with labor, interest, dividends and tax costs. Further, it is important to understand that there are large upfront energy investment in the plant construction and fuel acquisition, and the associated transmission and distribution systems. Several years pass after construction begins before any net-energy profit is realized (if any.)

With respect to wind energy, (repeating) according to the American Wind Energy Association³, “installed wind energy generating capacity totaled about 2550 MW in year and generated about 5.5 billion kWh of electricity –less than 1% of U.S. electricity generation. By contrast, the total amount of electricity that could be generated from wind in the United States has been estimated at 10,777 billion kWh annually ---three times the electricity generated in the U.S. today (2000).”

Fractions of U.S. Electricity Supply:

Three scenarios are analyzed whereby it is assumed that Wind Power generates, large portions of electricity for the United States. Annual rates of construction of 750 KW wind machines were found (by iteration) that would furnish selected portions (1/4th, ½ , or all) of the U.S. Electricity Supply (using Y2000 generation as a base line.) The models calculate the expected annual net energy and cumulative net energy for a period of 100 years.

1 . One Fourth of the U.S. Electricity Supply.
2.  One Half of the U.S. Electricity Supply.
3.  All of the U.S. Electricity Supply.

Assumptions used in the above Scenarios are shown in Table 2.

Table 2 – Assumptions Used in the Fractions of US Electricity Models.
 

Table 3a is an extracted image of the spreadsheet model that performs the “static” lifecycle analysis of one plant. The displayed mode is for a 750 kilowatt power plant over a 29-year period (3 yrs to build, operate 25 years, decommission in 1 yr.)
 

Table 3a - Lifecycle Static Net-Energy Analysis – Lifecycle Analysis of One 750 KW Plant

Table 3a shows an Output/Input ratio of 4.15 to 1 (~9 to 1 in Table 1) and an average price of electricity to end users is of 8.23 cents per kilowatt hour and the per cent of energy available to end users is 76%. This compares to approximately 90% available shown in Table.

It is important for reader to understand that the same reasoning was used in both Tables 1 and 2. What is being illustrated is an estimate of the “relative merit” of an energy system of mostly fossil fuels (Table 1) compared to what is likely to be expected from a wind-powered system (Table 2.)

Table 3b is extracted from the Dynamic Net-Energy Spreadsheet Model and illustrates the model used in analyzing the construction and operation of a large system of wind machines. Based on the static analysis results shown in Table 2, by interpolation, it was found that in order for wind power to produce 1/4th of the U.S. electricity supply. 878 750-megawatt machines would need to be constructed each year. This analysis was carried out over 100 years.

Table 3b uses Table 3 static net-energy results (Column H) and distributes the respective energy expenditures over the lifecycle of all the systems. Four years of the model are shown in columns L thru O. Column H (duplicate of Table 2) is also shown on this table.
 

Table 3b – Dynamic 100-Year Net-Energy Analysis (1/4th U.S. Electricity)

Figure 4, below, shows the net-energy yield of a system of wind machines related to Table 3, above, over a 100-year cycle of construction, operation and decommissioning worn out plants. Graph shows that it is estimated that an annual construction program of 878 new 750 KW machines reaches a maximum level of 21,950 machines in year 28, and operating continuously, would yield less than 1/10th of a quad of net energy, annually.

Figure 4 - Annual Net Energy - 1/4th of US Electricity, 21,950 750MW Machines

 

Figure 4a illustrates the results of the analysis of the scenario where Wind Power supplies all of the electricity needs of the United States (as in Y2000.) Here, less than 1/10th of a quad of net energy is available from the system.

Figure 4a - Annual Net Energy - All US Electricity, 104,550 Plants

Figure 4b, shows the estimated Cumulative Net Energy of a 100-year program of construction and operation of a large system of wind machines capable of producing an equivalent amount of electricity that was produced in Y2000. It can be noted that over 100 years only 6 quads of net energy are produced. Table 1 shows that the ANNUAL available net energy used by the 1997 fossil-fuel-based economy was about 85 quads. Bear in mind that the same logic that produced Table 1 was used to provide all of the data behind the graphs on this page.

Figure 4b - Cumulative Net Energy - 100 Years, All US Electricity Production

Table 4 is a summary of results for all three scenarios, described above.

Table 4 – Summary of Cost x E/GDP Analyses – 100-yr. Scenarios
Wind Generation (Fraction of U.S. Prod in Y2000)

Analysis, cont.

Maximum Electricity from Wind – United States constructs all available wind potential (AWEA):

The American Wind Energy Association says there is enough wind blowing (on average) in the United States to provide more than 3 times the electricity generated in the Year 2000. Relying on this finding, the scenarios in this section assume that U.S. energy policy is directed to using all this wind energy (by whatever means necessary.)