2/19/2013

Energy Output
Andrew Kusiak Intelligent Systems Laboratory 2139 Seamans Center The University of Iowa Iowa City, Iowa 52242 - 1527 [email protected] Tel: 319-335-5934 Fax: 319-335-5669 http://www.icaen.uiowa.edu/~ankusiak

Outline
The wind variation Distribution plotting The average bottle fallacy Mean power of the wind Betz' law Power density Power curves The power coefficient Calculator guide The power calculator Annual energy output
The University of Iowa Intelligent Systems Laboratory

The University of Iowa

Intelligent Systems Laboratory

Characterizing Wind Variability

Characterization of the wind speed variability is of importance to the wind industry Turbine designers use it to optimize the design of the turbines, e.g., by minimizing the energy generation cost Wind farm designers use it to locate wind farms and select turbines Investors use it to estimate the income from electricity generation

Wind Variations: Weibull Distribution

The General Pattern of Wind Speed Variation

Weibull distribution describes the wind variation for a typical site The site has the mean wind speed of 7 m/s The shape parameter of the function is k = 2 The data has been collected over a year period

The University of Iowa

Intelligent Systems Laboratory

The University of Iowa

Intelligent Systems Laboratory

2/19/2013

Weibull Distribution

Weibull Distribution
Scale Shape

Median Mode

where: > 0 is the scale parameter k > 0 is the shape parameter of the distribution

Mean

For k = 3.4, the Weibull distribution appears similar to the normal distribution For k = 1, the Weibull distribution becomes the exponential distribution

where the gamma function is

The University of Iowa

http://en.wikipedia.org

Intelligent Systems Laboratory

The University of Iowa

http://en.wikipedia.org

Intelligent Systems Laboratory

Cumulative Distribution Function

PDF and CDF

The University of Iowa

Intelligent Systems Laboratory

The University of Iowa

E. Hau (2006)

Intelligent Systems Laboratory

2/19/2013

CDF: Mean vs Median Wind Speed

Wind Speed Histogram

90 80 70 Number of observations 60 50 40 30 20 10 0 0.400 1.615

2.830 4.045

5.260 6.475

7.690 8.905

10.120 11.335

12.550 13.765

14.980 16.195

Wind speed (m /s)

The University of Iowa

E. Hau (2006)

Intelligent Systems Laboratory

The University of Iowa

Intelligent Systems Laboratory

Description of Wind Speed

The area under the pdf curve is always exactly 1 (the probability that the wind is blowing at some wind speed including 0 is 100 %) Median = 6.6 m/s (Half of the blue area is to the left) This means that 50% of the time the wind speed is less than 6.6 m/s, the other 50% of the time it is greater than 6.6 m/s Note: The median 6.6 m/s is not equal the mean 7m/s due to asymmetry of the pdf
The University of Iowa Intelligent Systems Laboratory

Description of Wind Speed

Mode = 5.5 m/s (The most common wind speed) The statistical distribution of wind speeds depends on location climate conditions, the landscape, and its surface The Weibull distribution may thus vary in its shape, determined by the pdf parameters
The University of Iowa Intelligent Systems Laboratory

2/19/2013

The Average Bottle Fallacy

The average energy content of the wind at a turbine site can not be determined from the average wind speed, rather the Weibull distribution is needed How large (in volume) is the average bottle, one is 0.24m and the other is 0.76m tall and both are of same shape? Though one is only only 3.17 taller than the other, its volume is actually 3.173 = 32 times larger than the small bottle (bottle V = cube of its size)

Betz' Law
The Ideal Braking of the Wind The more kinetic energy a wind turbine extracts from the wind, the more the wind will be slowed down (as it leaves the left side of the turbine in the tunnel) An attempt to extract all the energy from the wind, would reduce the speed to zero, i.e., the air could not leave the turbine In that case we would not extract any energy at all, as the new air would obviously be prevented from entering the rotor of the turbine Passing air without speed change would lead to zero extracted energy
The University of Iowa Intelligent Systems Laboratory

The University of Iowa

Intelligent Systems Laboratory

Betz' Law
The optimal point is: An ideal wind turbine slows down the wind by 2/3 of its original speed (v2 = 1/3v1)

Power Density Function

P 0 .5 A v 3
The Power of the Wind The area under the blue curve shows theoretical power that can be extracted (Betz' law says, 16/27 of the total power in the wind) The total area under the red curve represents the electrical power a certain wind turbine could produce Turbines power curve determines the actual power produced

Betz' law (Year 1919) says that one can only convert not more than 16/27 (or 59%) of the kinetic energy in the wind to mechanical energy using a wind turbine. (Albert Betz, German Physicist)
Modern rotors achieve values of the coefficient of performance Cp = 0.4 - 0.5, which is 70% to 80% of the theoretically possible value of Cpmax = 0.59
The University of Iowa Intelligent Systems Laboratory The University of Iowa

Intelligent Systems Laboratory

2/19/2013

Power coefficient Cp()

More Detailed Power Density Function

Mechanical Power: Summary 1

P 1 AC p ( )V 3 2
0.4

r V

P A V Cp r

Mechanical power produced by the rotor Air density Rotor swept area Wind speed Tip speed ratio Power coefficient Rotor radius Rotor speed

0.35 0.3 0.25 0.2 0.15 0.1 0.05

E = Erot x Eger xEgen x E p-conv

where: E = Wind Turbine System Efficiency
E. Hau (2006) The University of Iowa

0 0 2 4 6 8 10

Tip speed ratio

= ERotor x EGearbox x EGenerator x EPowerConverter

Intelligent Systems Laboratory The University of Iowa Intelligent Systems Laboratory

Mechanical Power: Summary 2

Rotor power P (W) 800
600

Summary 3
It is important to notice is that the bulk of wind energy is extracted at wind speeds above the average wind speed This is due to the fact that that the energy content of high wind speeds is much higher than energy content of low wind speeds Basically, the non-linear (cube) relationship between the power and wind speed

400

200

0 10 8 -200 0 6 20 4 40 60 2 80 100 0

P 0 .5 A v 3

Rotor speed (rad/s)

Wind speed V (m/s)

The University of Iowa

Intelligent Systems Laboratory

The University of Iowa

Intelligent Systems Laboratory

2/19/2013

The Power Curve of a Wind Turbine

The power curve of a wind turbine plots large the electrical power output at different wind speeds The cut-in wind speed: Turbines are designed to start running at wind speeds, e.g., 3 to 5 m/s The cut-out wind speed: Turbines are programmed to stop at high wind speeds, e.g., above 25 m/s to avoid damaging the turbine
The University of Iowa Intelligent Systems Laboratory

Power Curve

A: Insufficient power to overcome friction and initial torque B: Turbine operates to maximize efficiency C: Fixed (rated) power operation
Rotor power P (W) 800
600

400

200

1 P AC p ( )V 3 2 r V
The University of Iowa

0 10 8 -200 0 6 20 4 40 60 80 2 100 0

Rotor speed (rad/s)

Wind speed V (m/s)

Intelligent Systems Laboratory

Plotting Power Curves

If the wind speed is not fluctuating too rapidly, then one may use the wind speed measurements from the anemometer and read the electrical power output from the wind turbine and plot the two values in the form of the power curve In practice the wind speed always fluctuates, one cannot measure exactly the column of wind passing and therefore one will see a swarm of points dispersed around Furthermore, the actual anemometer error can be 3% , thus leading to 9% (or even 10%) error relected in the power curve

Actual Power Curve

?
The University of Iowa

P 0 .5 A v 3
Intelligent Systems Laboratory The University of Iowa Intelligent Systems Laboratory

2/19/2013

Actual Power Curve

Validating Power Curves

Power curves are constructed using measurements in areas with low turbulence intensity, and with the wind coming directly towards the front of the turbine Local turbulence and complex terrain (e.g., turbines placed on a rugged slope) may mean that wind gusts hit the rotor from varying directions It may therefore be difficult to reproduce the power curve exactly at any given location

The University of Iowa

Intelligent Systems Laboratory

The University of Iowa

Intelligent Systems Laboratory

The Power Coefficient

Power coefficient = Turbine efficiency The power coefficient indicates how efficiently a turbine converts the wind energy into electricity
E = Erot x Eger xEgen x E p-conv

The Power Coefficient

Power coefficient = Turbine efficiency Example: A turbine was designed for the max efficiency of 44% at a wind speed of about 9 m/s The efficiency is not as important at low wind speeds as there is not much energy to harvest At wind speeds above the cut out speed the turbine wastes the excess energy Efficiency therefore matters most in the regions of wind speeds where the most energy is to be found

Power coefficient = The electrical power output The wind energy input (From the power
equation)

Illustrative power efficiency curve

P 0. 5 A v 3

The efficiency varies with the wind speed

The University of Iowa Intelligent Systems Laboratory The University of Iowa Intelligent Systems Laboratory

2/19/2013

Annual Energy Output of a Wind Turbine

The relationship between the average wind speed and the annual energy output from a wind turbine Standard atmosphere with air density of 1.225 kg/m3 For each of the Weibull shape parameters k = 1.5, 2.0, and 2.5, the annual energy output is computed at different average wind speeds at the turbine hub height
The University of Iowa Intelligent Systems Laboratory

The Capacity Factor

The actual annual energy output Turbine capacity factor = The theoretical maximum output Theoretical maximum output = if the turbine was running at its maximum rated power for 8766 hours of the year (=24h*365.25days)

Example: For a 600 kW turbine producing 1.5 million kWh per year, its capacity factor is = 1,500,000/(365.25*24*600) = 1,500,000/5,259,600 = 0.285 = 28.5% The capacity factor may theoretically vary from 0 to 100%, but in practice it usually is 20% to 70%
The University of Iowa Intelligent Systems Laboratory

Turbine/Wind Plant Availability Factor

The time turbine/wind plant is available Availability factor = The total time (in a year) Different specific definitions of the availability factor Has improved in recent years Equipment reliability Warranty terms

Comparison
The electrical power output [KW] Turbine power coefficient = The wind energy input [KW]
Aerodynamic break

The actual annual energy output [kWh] Turbine capacity factor = The theoretical maximum output [kWh] The time turbine/wind plant is available Availability factor = The total time (in a year)

The University of Iowa

Intelligent Systems Laboratory

The University of Iowa

Intelligent Systems Laboratory

2/19/2013

The Capacity Factor Paradox

Although in general a large capacity factor is preferred, it may not always be an economic advantage This is often confusing to those in conventional or nuclear technology In a very windy location, for instance, it may be an advantage to use a larger generator for a giver rotor diameter This would tend to lower the capacity factor (using less of the capacity of a relatively larger generator), but it may imply a substantially larger annual production Non-linear vs quasi-liner capacity relationship: Capacity factor (Wind speed) power produced vs Production capacity (Machine speed) production
The University of Iowa Intelligent Systems Laboratory

The Capacity Factor Paradox

Considering a lower capacity factor for a relatively larger generator, depends both on wind conditions, and on the price of different turbine models Another way of looking at the capacity factor paradox is to say that to a certain extent a choice can be made between a relatively stable power output (close to the design limit of the generator) with a high capacity factor - or a high energy output (fluctuating) with a low capacity factor

The University of Iowa

Intelligent Systems Laboratory

Acknowledgement
The material included in the presentation comes largely from the Danish Wind Industry Association

The University of Iowa

Intelligent Systems Laboratory