INTERNATIONAL JOURNAL OF ENERGY RESEARCH Int. J. Energy Res.

2006; 30:553566 Published online 13 January 2006 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/er.1163

Thermodynamic analysis of wind energy

Ahmet Duran S , ahin1,z,}, Ibrahim Dincer2,n,y and Marc A. Rosen2,}
1

Aeronautic and Astronautics Faculty, Meteorology Department, Istanbul Technical University, 34469 Istanbul, Turkey 2 Faculty of Engineering and Applied Science, University of Ontario Institute of Technology (UOIT), Oshawa, Ontario, Canada L1H 7K4

SUMMARY Wind energy is assessed thermodynamically, from resource and technology perspectives. The thermodynamic characteristics of wind are considered. Wind speed is aected by air temperature and pressure and has an eect on wind turbine performance, based on wind chill eect and Bernoullis equation. The wind chill eect leads to temperature dierences that suggest enthalpy and entropy components must be considered in a thermodynamic analysis. The wind pressure eect based on Bernoullis equation aects the entropy of wind. These components have not previously been considered in evaluations of wind turbine eciency for electricity generation. A new eciency formula for wind energy systems is described, which provides important information about the system. It is seen that average dierences between energy and exergy eciencies are approximately 40% at low wind speeds and up to approximately 55% at high wind speeds. Copyright # 2005 John Wiley & Sons, Ltd.
KEY WORDS:

eciency; energy; exergy; model; wind speed; wind turbine

1. INTRODUCTION Life is directly aected by energy and its consumption, and energy research is therefore important. Wind energy is among the worlds most signicant and rapidly developing renewable energy sources. Recent technological developments, concerns over fossil fuel demands and the corresponding environmental eects and the continuous increase in the consumption of conventional energy resources have reduced wind energy costs to economically acceptable levels

Correspondence to: I. Dincer, Faculty of Engineering and Applied Science, University of Ontario Institute of Technology (UOIT), Oshawa, Ontario, Canada L1H 7K4. y E-mail: ibrahim.dincer@uoit.ca z E-mail: sahind@itu.edu.tr } E-mail: marc.rosen@uoit.ca } On leave as a postdoctoral research fellow at UOIT. Contract/grant sponsor: Istanbul Technical University, Turkey Contract/grant sponsor: University of Ontario Institute of Technology Contract/grant sponsor: Natural Sciences and Engineering Research Council of Canada

Copyright # 2005 John Wiley & Sons, Ltd.

Received 2 March 2005 Revised 17 June 2005 Accepted 27 June 2005

554

A. D. S , AHIN, I. DINCER AND M. A. ROSEN

in many locations. Wind energy farms, consequently, are being considered as an alternative energy source in many jurisdictions. Most new and renewable energy sources, such as wind, solar, hydraulic and wave energy, are related to meteorological variables. If the meteorological characteristics of these renewable energy sources are not well known and understood, there can be important gaps in knowledge related to energy investments. The concept of wind power meteorology has been discussed by Petersen et al. (1998) and Sahin (2004). As a meteorological variable, wind energy refers to the energy content of wind. In electricity generation wind takes the same role as water does for hydraulic generation, and based on this importance wind variables should be analysed. Wind velocity deviation and changeability depend on time and location. Understanding such characteristics is the subject of wind velocity modelling. Determining the atmospheric boundary layer and modelling it is a special consideration in wind power research. Much research has been carried out on these subjects. For instance, Petersen et al. (1998) considered wind power meteorology and sought the relationship between meteorology and wind power. During the preparation of the Denmark Wind Atlas detailed research was performed on wind energy, as a meteorological energy source (Freris, 1981; Petersen et al., 1981). Meteorological variables such as temperature, pressure and moisture play important roles in the occurrence of wind. Generally, in wind engineering, moisture changeability is negligible and air is assumed to be dry. Wind as a meteorological variable can be described as a motion of air masses on a large scale with potential and kinetic energies. Pressure forces lead to kinetic energy (Freris, 1981, 1990; Sen, 1997). In wind engineering applications horizontal winds are important because they cover great areas. The dynamic behaviour of the atmosphere generates spatio-temporal variations in such parameters as pressure, temperature, density and moisture. These parameters can be described by expressions based on continuity principles, the rst law of thermodynamics, Newtons law and the state law of gases. Mass, energy and momentum conservation equations for air in three dimensions yield balance equations for the atmosphere. Wind occurs due to dierent cooling and heating phenomena within the lower atmosphere and over the earths surface. Meteorological systems move from one place to another by generating dierent wind velocities. With the growing signicance of environmental problems, clean energy generation has become increasingly important. Wind energy is clean, but it does not persist continually for long periods of time at a given location. Fossil fuels often must supplement wind energy systems. Many scientic studies have addressed this problem with wind energy (e.g. Cherry, 1980; Justus, 1978; Troen and Petersen, 1989; Sahin, 2002). During the last decade, wind energy applications have developed and been extended to industrial use in some European countries including Germany, Denmark and Spain. Successes in wind energy generation have encouraged other countries to consider wind energy as a component of their electricity generation systems. The clean, renewable and in some instances economic features of wind energy have drawn attention from political and business circles and individuals. Development in wind turbine technology has also led to an increased usage. Wind turbine rotor eciency increased from 35 to 40% during the early 1980s, and to 48% by the mid-1990s. Moreover, the technical availability of such systems has increased to 98% (Karne and Jrgensen, 1995; Gipe, 1995; Neij, 1999; Salle et al., 1990). Koreneos et al. (2003) applied exergy analysis to renewable energy sources including wind power. But in this paper only the electricity generation of wind turbines is taken into account
Copyright # 2005 John Wiley & Sons, Ltd. Int. J. Energy Res. 2006; 30:553566

THERMODYNAMIC ANALYSIS OF WIND ENERGY

555

and the exergy eciency of wind turbines for wind speeds above 9 m s1 is treated as zero. Jia et al. (2004) carried out an exergy analysis of wind energy. They considered wind power for air compression systems operating over specied pressure dierences, and estimated the system exergy eciency. Dincer and Rosen (2005) investigated thermodynamic aspects of renewables for sustainable development. They explain relations between exergy and sustainable development. Wind speed thermodynamic characteristics are given by Go et al. (1999), with the intent of using the cooling capacity of wind as a renewable energy source (i.e. using the wind chill eect for a heat pump system). Although turbine technology for wind energy is advancing rapidly, there is a need to assess accurately the behaviour of wind scientically. Some of the thermodynamic characteristics of wind energy are not yet clearly understood. The capacity factor of a wind turbine sometimes is described as the eciency of a wind energy turbine. But there are diculties associated with this denition. The eciency of a wind turbine can be considered as the ratio of the electricity generated to the wind potential within the area swept by the wind turbine. In this denition only the kinetic energy component of wind is considered. Other components and properties of wind, such as temperature dierences and pressure eects, are neglected. The purpose of this paper is to develop an improved approach for the thermodynamic analysis of wind energy using energy and exergy. Such a formulation can provide a physical basis for understanding, rening and predicting the variations in wind energy calculations. In line with this objective, which includes providing an improved wind energy eciency denition based on exergy analysis, the paper contains three main parts. First, wind energy and its energy components are discussed. Second, exergy analysis is applied to wind, and the exergy is formulated of wind energy and its components. Last, energy and exergy eciencies are compared and shown to depend on the area considered.

2. ENERGY AND EXERGY ANALYSES OF WIND ENERGY ASPECTS People sense whether air is warm or cool based not only on air temperature, but also on wind speed and humidity. During cold weather, faster wind makes the air feel colder, because it removes heat from our bodies faster. Wind chill is a measure of this eect, and is the hypothetical air temperature in calm conditions (air speed, V 0) that would cause the same heat ux from the skin as occurs for the actual air speed and temperature. The heat transfer for an air ow over a surface is slightly modied in some versions of the wind chill expression (Stull, 2000). The present wind chill expression is based on the approaches of Osczevski (2000) and Zecher (1999), and was presented at the Joint Action Group for Temperature Indices (JAG/TI) (2001) meeting held in Toronto. The JAG/TI expression makes use of advances in science, technology and computer modelling to provide a more accurate, understandable and useful formula for calculating the dangers from winter winds and freezing temperatures. In addition, clinical trials have been conducted and the results have been used to verify and improve the accuracy of the expression, which is given as Twindch 35:74 0:6215Tair 35:75V 0:16 0:4274Tair V 0:16 where the wind chill temperature Twindch is in 8F and wind speed V is in mph.
Copyright # 2005 John Wiley & Sons, Ltd. Int. J. Energy Res. 2006; 30:553566

556

A. D. S , AHIN, I. DINCER AND M. A. ROSEN

Another wind speed factor is wind pressure. When the wind approaches an obstacle, the air ows around it. However, one of the streamlines that hits the obstacle decelerates from the upstream velocity of vs to a nal velocity of zero (or to some lower velocity). The pressure (dynamic pressure) at this stagnation point is higher than the free stream pressure (static pressure) well away from the obstacle. The dynamic pressure can be calculated from Bernoullis equation. For ow at constant altitude, the only two terms that change in Bernoullis equation are kinetic energy and pressure.

3. ENERGY ANALYSIS Wind energy E is the kinetic energy of a ow of air of mass m at a speed V. The mass m is dicult to measure and can be expressed in terms of volume V through its density r m=V: The volume can be expressed as V AL where A is the cross-sectional area perpendicular to the ow and L is the horizontal distance. Physically, L Vt and wind energy can be expressed as
3 E1 2 rAtV

Betz (1946) applied simple momentum theory to the windmill established by Froude (1889) for a ship propeller. In that work, the retardation of wind passing through a windmill occurs in two stages: before and after its passage through the windmill rotor. Provided that a mass m is air passing through the rotor per unit time, the rate of momentum change is m(V1V2) which is equal to the resulting thrust. Here, V1 and V2 represent upwind and downwind speeds at a considerable distance from the rotor. The power absorbed P can be expressed as % 3 P mV1 V2 V On the other hand, the rate of kinetic energy change in wind can be expressed as
2 2 Ek 1 2 mV1 V2

The expressions in Equations (3) and (4) should be equal, so the retardation of the wind, % ; before the rotor is equal to the retardation, V % V2 ; behind it, assuming that the V1 V direction of wind velocity through the rotor is axial and that the velocity is uniform over the area A. Finally, the power extracted by the rotor is % % V1 V2 V 5 P rAV Furthermore,   V V2 2 % 2 V1 V2 rA 1 P rAV V1 V2 2 and Pr
3 AV1 1 a1 a2 4

where a

V2 V1

Dierentiation shows that the power P is maximum when a 1 3; i.e. when the nal wind velocity V2 is equal to one third of the upwind velocity V1. Hence, the maximum power that can be 3 8 3 extracted is rAV1 27; as compared with rAV1 =2 in the wind originally, i.e. an ideal windmill 16 could extract 27 (or 0.593) of the power in the wind (Golding, 1955).
Copyright # 2005 John Wiley & Sons, Ltd. Int. J. Energy Res. 2006; 30:553566

THERMODYNAMIC ANALYSIS OF WIND ENERGY

557

4. EXERGY ANALYSIS Exergy analysis is a technique that uses the conservation of mass and conservation of energy principles together with the second law of thermodynamics for the analysis, design and improvement of energy and other systems (Dincer and Rosen, 2005, 2004). Exergy is dened as the maximum amount of work which can be produced by a system or a ow of matter or energy as it comes to equilibrium with a reference environment. Unlike energy, exergy is not subject to a conservation law (except for ideal, or reversible, processes). Rather exergy is consumed or destroyed, due to irreversibilities in any real process. The exergy consumption during a process is proportional to the entropy created due to irreversibilities associated with the process. Exergy is a measure of the quality of energy which, in any real process, is not conserved but rather is in part destroyed or lost. For exergy analysis, the characteristics of a reference environment must be specied. This is commonly done by specifying the temperature, pressure and chemical composition of the reference environment. The results of exergy analyses, consequently, are relative to the specied reference environment, which in most applications is modelled after the actual local environment. The exergy of a system is zero when it is in equilibrium with the reference environment. This tie between exergy and the environment has implications regarding environmental impact that are discussed elsewhere (Dincer and Rosen, 2005). In this paper, real atmospheric conditions are taken for reference environment. Energy and exergy balances for a ow of matter through a system can be expressed as X X X h ke pein min h ke peex mex Qr W 0 8
in ex r

X
in

exin min

X
ex

exex mex

X
r

ExQ ExW I 0

where min and mex denote mass input across port in and mass exiting across port ex, respectively; Qr denotes the amount of heat transfer into the system across region r on the system boundary; ExQ is the exergy transfer associated with Qr; W is the work (including shaft work, electricity, etc.) transferred out of the system; ExW is the exergy transfer associated with W; I is the system exergy consumption; and h, ke, pe, and ex denote specic values of enthalpy, kinetic energy, potential energy and exergy, respectively. Note that the exergy consumption I is greater than zero for an irreversible process and equal to zero for a reversible process. Since min=mex=0, for a closed system, Equations (8) and (9) simplify in that case to X Qr W 0 10
r

X
r

ExQ ExW I 0

11

Exergy of a owing stream of matter. Consider a ow of matter at temperature T, pressure P, chemical composition mj of species j, mass m, specic enthalpy h, specic entropy s, and mass fraction xj of species j. The environment is considered as in an equilibrium state with intensive properties at T0, P0 and mj00. The environment is taken to be large enough that its intensive properties are negligibly aected by any interactions with the system. With the above
Copyright # 2005 John Wiley & Sons, Ltd. Int. J. Energy Res. 2006; 30:553566

558

A. D. S , AHIN, I. DINCER AND M. A. ROSEN

considerations, the specic exergy of a ow of matter can be expressed as " # X mj0 mj 00 xj ex ke pe h h0 T0 s s0

j

12

Note that the above equation can be separated into physical and chemical components. If ke 0 and pe 0; the physical exergy h h0 T0 s s0 is the maximum P available work from a ow as it is brought to the environmental state. The chemical exergy j mj0 mj00 xj is the maximum available work extracted from the ow as it is brought from the environmental state to the dead state. Exergy of work. The exergy associated with work is ExW W 13

The exergy of wind energy can be estimated with the work exergy expression, because there are no heat and chemical components. Exergy consumption. The amount of exergy consumed due to irreversibilities during a process is I T0 Sgen 14

5. ENERGY AND EXERGY EFFICIENCIES The expressions for energy (Z) and exergy (c) eciencies for the principal types of processes considered in the present study are based on the following denitions: Z = (energy in products/total energy input), and c = (exergy in products/total exergy input). Here, exergy eciencies can often be written as a function of the corresponding energy eciencies. Note that the exergy eciency frequently gives a ner understanding of performance than the energy eciency. The exergy eciency weights energy ows by accounting for each in terms of usefulness. The exergy eciency stresses that both external losses and internal irreversibilities need to be dealt with to improve eciency. In many cases, the internal irreversibilities are more signicant and more dicult to deal with than external losses. Work production. Electric work production processes produce shaft work W and the eciencies for shaft work production from electricity through a wind energy system are Zm ; e W =W e cm;e E W =E We W =We Zm;e 15 16

Electricity generation. The eciencies for electricity generation through a wind energy system are Ze; f We =mf Hf ce; f E We =mf gf Hf Ze; f
Copyright # 2005 John Wiley & Sons, Ltd.

17 18
Int. J. Energy Res. 2006; 30:553566

THERMODYNAMIC ANALYSIS OF WIND ENERGY

559

Therefore, the exergy eciencies for the electricity generation process can be taken as equivalent to the corresponding energy eciencies. Kinetic energy production. The eciencies for the fossil fuel- and wind-driven kinetic energy production processes, which produces a change in kinetic energy Dke in a stream of matter ms, are as follows (Dincer and Rosen, 2005): Zke; f ms Dkes =mf Hf cke; f ms Dkes =mf gf Hf Zke; f 19 20

For a wind turbine, kinetic energy is dominant and there is no potential energy change or chemical component. The general specic exergy equation is then given as " # X mj0 mj 00 xj ex ke pe h h0 T0 s s0 21
j

In this paper, the input and output variables for the system are described in Figure 1. Output wind speed is estimated using the continuity equation. The total electricity generated is a related to the decrease in wind potential. Subtracting the generated power from the total potential gives the wind turbine back-side wind potential (Figure 1): s 2Epotential Egenerated V2 3 22 rAt In addition, the total kinetic energy dierence gives the generated electricity which can be written as DKe Egenerated 23

V2 Tat Pat PV2

V1 Tat Pat PV1

Figure 1. Wind turbine and representative wind energy input and output variables.
Copyright # 2005 John Wiley & Sons, Ltd. Int. J. Energy Res. 2006; 30:553566

560

A. D. S , AHIN, I. DINCER AND M. A. ROSEN

Air mass ow with time depends on density and wind speed, and can be shown as m rAV 24 Here, the exergy of a matter ow is dened as the maximum work that can be acquired when the air ows from state (T2, P2) to the ambient state (T1, P1). The equations of specic enthalpy, entropy and exergy at states 1 and 2 are given below. DH mC p T2 T1
1 1

25

where DH is the change in enthalpy (J kg ); h is the specic enthalpy (J kg ), m is mass ow rate of air, which depends on time, T1 is the wind chill temperature at the input to the wind turbine; and T2 is the wind chill temperature at the exit of the wind turbine. The total entropy of system and entropy dierence can be written as DS DSsystem DSsurround       T2 P2 Qloss Cp ln R ln T1 P1 Tat r Pi Pat V 2 2 and Qloss mC p Tat Taverage 29 Here, DS is the specic entropy (J kg1 K), Tat is the atmospheric temperature, P2 is the pressure at the exit of the wind turbine for a wind speed V2 and P1 is the pressure at the inlet of the wind turbine for a wind speed V1, Qloss represents heat losses of wind turbine and Taverage is the mean value of input and output wind chill temperatures. Thus, the total exergy for wind energy can be expressed by summation of Equations (23)(29) as       T2 P2 Qloss Ex Egenerated mC R ln 30 p T2 T1 mT at Cp ln T1 P1 Tat As explained above the rst term on the right side of this equation is the generated electricity. The second and third parts are enthalpy and entropy contributions, respectively. 26

DS mT at where

27

28

6. CASE STUDY In order to evaluate and assess wind energy potential, a database is considered of hourly wind speed and direction measurements taken between May 2001 and May 2002 at seven stations in the northern part of Istanbul, which is located between 40.978E longitude and 29.088N latitude. For this research, values from only one station are considered. This area comes under the inuence of the mild Mediterranean climate during the summer, and consequently experiences dry and hot spells for about four to ve months, with comparatively little rainfall. During the wintertime, this region comes under the inuence of high-pressure systems from Siberia and the Balkan Peninsula and low-pressure systems from Iceland. Hence, northeasterly or westerly winds inuence the study area, which also has high rainfall amounts in addition to snow every
Copyright # 2005 John Wiley & Sons, Ltd. Int. J. Energy Res. 2006; 30:553566

THERMODYNAMIC ANALYSIS OF WIND ENERGY

561

year in winter. Air masses originating over the Black Sea also reach the study area (For details, see Sahin, 2002).

7. RESULTS AND DISCUSSION In this paper, measured generated power data from a group in Denmark are used to obtain a power curve. Pedersen et al. (1992) recommend wind turbine power curve measurements be used to determine the wind turbine required in relation to technical requirements and for approval and certication of wind turbines in Denmark. In this study, output electricity power data for a 100 kW wind turbine with a rotor diameter at 18 m and hub height 30 m are given. The data power curve of this wind turbine is shown in Figure 2(a). The power curve is seen to exhibit two main types of behaviour, depending on wind speed. At low wind speeds, the power increases with wind speed until the rated power wind speed is reached. A second degree polynomial curve t can be obtained using a least-squares minimization technique. A curve is tted between the cut-in and rated power wind speeds and its coecient of determination (R2) is estimated as 0.99. At high wind speeds (above 16 m s1), the power generation levels o and then tends to decrease from the rated power with increasing wind speed. The cut-out wind speed of this turbine is 20.3 m s1. In the rated wind speed region, a third degree polynomial curve is tted and its R2 value is calculated as 0.78. The tted curves for electricity power generation, based on measured data, are illustrated in Figure 2(b). The exergy analysis of wind energy shows that there are signicant dierences between energy and exergy analysis results. According to one classical wind energy eciency analysis technique, which examines capacity factor, the resultant wind energy eciency is overestimated relative to what it really is. The capacity factor normally refers to the percentage of nominal power that the wind turbine generates. The given test turbine capacity factor is also compared with modelled desired area calculations. It is seen that, as for the power curves in Figure 2, there is a close relation between capacity factors. Energy and exergy eciencies are therefore calculated with Equations (17) and (18), respectively. The dierences between exergy and energy eciencies are shown in Figure 3. Below the cut-in wind speed (3.8 m s1) and over the cut-out wind speed

Generated Electricity (kW)

Generated Electricity (kW) 0 5 10 15 Wind Speed (m/s) 20 25 (b)

120 100 80 60 40 20 0

120 100 80 60 40 20 0 0 5 10 15 20 Wind Speed (m/s) 25 30

(a)

Figure 2. Test wind turbine power curve: (a) showing electricity generated as a function of wind speed; and (b) showing regression curves for electricity generated as a function of wind speed.
Copyright # 2005 John Wiley & Sons, Ltd. Int. J. Energy Res. 2006; 30:553566

562
0.6 0.5 0.4 0.3 0.2 0.1 0

A. D. S , AHIN, I. DINCER AND M. A. ROSEN

Capacity Factor Energy Efficiency Exergy Efficiency 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Time (h)

Figure 3. Variation of capacity factor, and energy and exergy eciencies, using a sample set of wind data during the day.

60 Exergy Efficiency 24per. Mov. Avg. 50 Exergy Efficiency (%) 40 30 20 10 0 1 802 1603 2404 3205 4006 4807 5608 6409 7210 8011 Hours (Beginning May 2001- May 2002)

Figure 4. Mean daily exergy eciencies.

(20.3 m s1) electricity generation is zero. As a result, energy and exergy eciencies also are zero in those ranges. Since wind speed exhibits high variability during the day, with greater uctuations than all other meteorological parameters, the uctuations in energy and exergy eciency values are high. All exergy eciencies are calculated by using Equation (18) for a selected point, and given in Figure 4 as 24 degrees moving average values. In this gure, 24 moving average values are used to show that the daily changes depend on seasonal variability and to see the periodicity of the
Copyright # 2005 John Wiley & Sons, Ltd. Int. J. Energy Res. 2006; 30:553566

THERMODYNAMIC ANALYSIS OF WIND ENERGY

563

exergy eciencies. Moving average is a statistical method for smoothing highly uctuating variables. In this study, 24 moving average is considered to illustrate daily variability. The beginning of the data is May 2001, and shows that during spring and summer an approximately constant variability is observed, but in winter the uctuations increase. In other words, during high wind speeds and cold weather the eciencies are more variable. In addition, exergy eciencies of wind energy are low in autumn. After autumn, high wind speeds occur since the region comes under the inuence of high pressure from Siberia and the Balkan Peninsula and low pressure from Iceland. Figures 3 and 4 show the variations and large uctuations of the eciencies. These gures are more useful for meteorological interpretation than engineering application. For power generation application, electricity generation, which depends on the power curves, is estimated. Then, the enthalpy and entropy parts of Equation (30) are calculated, and the energy and exergy eciencies are evaluated with Equations (17) and (18). For each eciency calculation, 8637 data values are employed. Then, regression analysis is applied to wind speeds between the cut-in and cut-out levels, and energy and exergy eciencies are calculated (Figure 5). The lowest eciencies are observed at the cut-in and cut-out wind speeds. As seen in Figure 6, energy and exergy eciencies exhibit important dierences at every wind speed. For this reason, we suggest that exergy eciencies be used in energy assessments instead of energy eciencies. Such an approach yields more realistic results and provides more information about wind energy systems. In Figure 5, it is seen that the exergy eciency curve is smoother than energy eciency curve. In other words, deviations for energy eciencies are higher than for exergy eciencies. To illustrate these variations, mean standard deviations of these eciencies are calculated for each wind speed interval (Figure 6). It is observed that at lower wind speeds standard deviations for energy eciencies are higher than those for exergy eciencies. Above wind speeds of 9 m s1, same standard deviations are observed.

0.6

Energy and Exergy Efficiency

0.5

0.4 Energy Efficiency Exergy Efficiency 0.3

0.2

0.1

0 0 5 10 15 Wind Speed (m/s) 20 25

Figure 5. Variation of energy and exergy eciencies as a function of wind speed.

Copyright # 2005 John Wiley & Sons, Ltd. Int. J. Energy Res. 2006; 30:553566

564
0.08 0.07 Standard Deviation 0.06 0.05 0.04 0.03 0.02 0.01 0 3

A. D. S , AHIN, I. DINCER AND M. A. ROSEN

Standard Dev. Energy Eff. Standard Dev. Exergy Eff.

11 13 15 Wind Speed (m/s)

17

19

21

Figure 6. Standard deviation of exergy and energy eciencies with wind speed.

60 50 Efficiencies (%) 40 30 20 10 0 3 4 5 6 7 8

40 30 20 10 0 9 10 11 12 13 14 15 16 17 18 19 Wind Speed (m/s)

Figure 7. Mean exergy and energy eciencies, and percent dierences between these values as a function of wind speed.

In Figure 7, mean energy and exergy eciencies are presented as a function of wind speed. This gure emphasizes the dierences between the eciencies, and shows the over-estimation provided by energy eciencies. The relative dierences between energy eciency and exergy eciency, where exergy eciency is taken as the base value, are given in the same gure. There, it is seen that the relative dierence is lowest at a wind speed of about 7 m s1, and increases at lower and higher wind speeds. These relative dierences imply that the exergy eciency approach should be applied to wind energy systems for better understanding.
Copyright # 2005 John Wiley & Sons, Ltd. Int. J. Energy Res. 2006; 30:553566

Relative Difference (%)

Mean Capacity Factor Mean Energy Efficiency Mean Exergy Efficiency Relative Difference

60 50

THERMODYNAMIC ANALYSIS OF WIND ENERGY

565

8. CONCLUSIONS A new exergy formulation for wind energy is developed and described. The present formulation is more realistic and also accounts for the thermodynamic quantities enthalpy and entropy. Using this formulation, dierences are illustrated between energy and exergy eciencies as a function of wind speed. The dierences between the eciencies can be signicant. It is thus suggested that exergy eciency should be used for wind energy evaluations and assessments, so as to allow for more realistic modelling, evaluation and planning for wind energy systems. NOMENCLATURE A E ex ExQ ExW I h H ke L M min mex Qr P1 P2 P pe S T1 T2 Tair V V1 V2 V W =control area =ux of moving wind energy =specic exergy =exergy transfer associated with Qr =exergy transfer associated with W =system exergy consumption =specic enthalpy =enthalpy =specic kinetic energy =horizontal distance =air mass =mass input =mass exiting =heat transfer into system across region r on system boundary =pressure at inlet to wind turbine for a wind speed V1 =pressure at exit from wind turbine for a wind speed V2 =power absorbed by wind turbine =specic potential energy =entropy =wind chill temperature at inlet to wind turbine =wind chill temperature at exit from wind turbine =air temperature at a height of 10 m =wind speed =upwind speed from wind turbine at a considerable distance from the rotor =downwind speed from wind turbine at a considerable distance from the rotor =volume =work (including shaft work, electricity, etc.) transferred out of system

Greek letters mj Z c r =chemical composition of species j =energy eciency =exergy eciency =density
Int. J. Energy Res. 2006; 30:553566

Copyright # 2005 John Wiley & Sons, Ltd.

566

A. D. S , AHIN, I. DINCER AND M. A. ROSEN

ACKNOWLEDGEMENTS

The authors acknowledge the support provided by Istanbul Technical University in Turkey and University of Ontario Institute of Technology and the Natural Sciences and Engineering Research Council of Canada in Canada.

REFERENCES Betz A. 1946. Windenergie und ihre Ausnutzung durch Windmu ttingen: Vandenhoek and Ruprecht, Go ttingen. hlen. Go Cherry NJ. 1980. Wind energy resources methodology. Journal of Industrial Aerodynamic 5:247280. Dincer I, Rosen MA. 2004. Exergy as a driver for achieving sustainability. International Journal of Green Energy 1(1):119. Dincer I, Rosen MA. 2005. Thermodynamic aspects of renewables and sustainable development. Renewable and Sustainable Energy Reviews 5(9):169189. Freris LL. 1981. Meteorological aspects of the utilization of wind as an energy source. WMO-No: 575, Switzerland. Freris LL. 1990. Wind Energy Conversion Systems. Prentice-Hall International Ltd.: U.K. Froude RE. 1889. On the part played in propulsion by dierences of uid pressure. Transactions of the Institute of Naval Architects 30:390. Gipe P. 1995. Wind Energy Comes of Age. Wiley: New York. Go LH, Hasert UF, Go PL. 1999. A new source of renewable energy: the coldness of the wind. International Thermal Science 38:916924. Golding EW. 1955. The Generation of Electricity by Wind Power. E.&F. N. Spon Limited: London. JAG/TI (Joint Action Group for Temperature Indices) 2001. New wind chill equation. JAG/TI Meeting, Toronto, Canada, 2 August. Jia GZ, Wang XY, Wu GM. 2004. Investigation on wind energy-compressed air power system. Journal of Zhejiang University Science 5(3):290295. Justus CG. 1978. Wind and Wind System Performance. Franklin Institute Press: Philadelphia. Karne P, Jrgensen U. 1995. Samfundsmssig vrdi af vindkraft: Dansk vindmlleindustris internationale position og udviklingsbetingelser [Socio-economic assessment of wind power: the Danish wind turbine manufacturers position and condition for development]. AKF Forlaget: Copenhagen, Denmark. Koroneos C, Spachos N, Moussiopoulos N. 2003. Exergy analysis of renewable energy sources. Renewable Energy 28:295310. Neij L. 1999. Cost dynamics of wind power. Energy 24:375389. Osczevski RJ. 2000. Windward cooling: an overlooked factor in the calculation of wind chill. Bulletin of the American Meteorological Society 81(12):29752978. Petersen EL, Mortensen NG, Landberg L, Hojstrup J, Frank HP. 1998. Wind power meteorology}part I: climate and turbulence. Wind Energy 1:2545. Pedersen TF, Petersen SM, Paulsen US, Fabian O, Pedersen BM, Velk P, Brink M, Gjerding J, Frandsen S, Olesen J, Budtz L, Nielsen MA, Stiesdal H, Petersen K, Danwin PL, Danwin LJ, Friis P. 1992. Recommendation for wind turbine power curve measurements to be used for type approval of wind turbines in relation to technical requirements for type approval and certication of wind turbines in Denmark. Danish Energy Agency, September. Petersen EL, Troen I, Frandsen S, Hedegaard K. 1981. Wind atlas for Denmark. A rational method for wind energy sitting. Riso-R-428, Riso National Laboratory, Roskilde. S , ahin AD. 2002. Spatio-temporal modelling winds of Turkey. Ph.D. Dissertation, ITU, Institute of Science and Technology, Istanbul (in Turkish). S , ahin AD. 2004. Progress and recent trends in wind energy. Progress in Energy and Combustion Science 30:501543. Salle SA, Reardon D, Leithead WE, Grimble MJ. 1990. Review of wind turbine control. International Journal of Control 52(6):12951310. S , en Z. 1997. Statistical investigation of wind energy reliability and its application. Renewable Energy 10:7179. Stull R. 2000. Meteorology for Scientists and Engineers (2nd edn). Brooks/Cole Thomson Learning: Belmont, California. Troen I, Petersen EL. 1989. European Wind Atlas. Commission of the European Communities. Riso National Laboratory: Roskilde, Denmark. Zecher JBM. 1999. A new approach to an accurate wind chill factor. Bulletin of the American Meteorological Society 80(9):18931899.

Copyright # 2005 John Wiley & Sons, Ltd.

Int. J. Energy Res. 2006; 30:553566