Resources and Technologies - ece.citadel.eduece.citadel.edu/peeples/ELEC 427/Wind Energy.pdf · Source: F Vanek, L Albright, and L Angenent. (2012) Energy Systems Engineering: Evaluation

Wind Energy – Chapter 13Resources and Technologies

Energy Systems Engineering

Source: F Vanek, L Albright, and L Angenent. (2012) Energy Systems Engineering: Evaluation and Implementation, 2nd Ed., McGraw-Hill.

This slide may be distributed as long as this attribution is maintained with the slide.

Further Readings

David Spera, Ed., “Wind Turbine Technology: Fundamental Concepts of Wind Turbine Engineering”, ASME Press, New York, 1994.

Manwell, J et al (2002) “Wind Energy Explained.” Wiley, West Sussex, 2003.



Learning Objectives Understanding wind data

Estimating available wind using statistics

Overview of turbine function

Actuator disc model

Blade element theory

Economics of wind energy

Other design issues and consideration of local impacts

Future of turbine design

Integrating wind and other intermittent sources into the grid



Key Concepts Energy source

Wind = solar energy converted to kinetic energy

Terminology

Distinction between “mill” and “turbine”

Two basic types:

Horizontal Axis Wind Turbine (HAWT)

Vertical Axis Wind Turbine (VAWT)

Modern turbine depends on

Aerospace technology

Modern materials engineering

Sophisticated fluid mechanics

Precise electronic controls



Early Wind Generators

An early example:

Brush Electric Generator, Cleveland, OH, 1880s, 12kW

Early adaptation of wind-powered mechanical pump to generating electricity

First attempt at > 1 MW turbine

Smith-Putnam Turbine, Vermont, 1940s, 1.25 MW

Failed prematurely, not repeated

Development of modern utility-scale turbine

California, Denmark in 1970s and 1980s

Experimentation with vertical axis turbines, but eventually settled on horizontal axis design



Madison Wind Farm, Madison, NYSource: www.photosfromonhigh.com

http://www.photosfromonhigh.com/



Figure 13-1. Growth in annual and cumulative installed

capacity of utility-scale wind turbines in U.S., 1980-2014

Source: American Wind Energy Association



Figure 13-2. Total installed capacity, 2005 and 2010

Source: Global Wind Energy Consortium

Total = 59.1 GW Total = 195.0 GW



Figure 13-5. Main parts of a utility-scale wind turbine

Not to scale

Fluid dynamics

Material advances

•Lighter, stronger and more efficient

assemblies

•Increased swept areas

•Slower blade speeds

•Taller towers

Blades rotate along the ‘roll’ axis

Nacelle rotates along the ‘yaw’ axis

Most utility scale towers have 3 blades

•Trade off between energy extraction

efficiency, cost and weight.



Operating Requirements for HAWTs

Start, stop, and control output during operation

Assisted startup: use turbine as motor

Change pitch during operation to modulate power

Stopping function: “loss of load” emergency

Winds above design speed

Can also shed wind at high wind speeds

“Solidity” of blades: area of blades relative to total swept area



Vertical Axis Wind Turbine (VAWT)

No need for ‘yaw’ rotation Responds to wind from any directions

No utility scale VAWTs

Some 500 to 20KW units

Study continues



So, Where do we site these things?Brazos Wind Ranch in Texas

Where the wind is!

Madison Wind Farm, Madison, NY

Smoky Hills Wind Farm in Kansas





















We need to …

Understand wind data

Estimate available wind using statistics



Data Gathering for Wind Sites

For most sites, 1 year record predicts mean wind speed to

within accuracy of 10%

Importance of data from all seasons of year

E.g., summer versus winter usually different

Estimation of interannual variability nearly as important as

estimating mean annual wind

E.g., data gathering at Fenner wind farm, NY state

Average 7.7 m/s @ 65 meters, or 17.3 MPH

Gathered data for several years prior to installation at site

Resulting output from site in kWh of electricity varies + 5% from

predicted average from year to year



Table 13-2. Classification of wind resource by wind speed

range in m/s and mph at hub height of turbine



Wind Statistics

Two methods

Sample Average speed from limited number of readings

Statistical wind maps estimate local wind from nearby readings adjusted for terrain and prevailing winds.

Detailed Histograms (bin analyses) of continuous wind

readings at a site…8760 hours per year!

Year to year variations are generally <10%.



Hypothetical Year

About 40% of year is good to outstanding plus range (blue)

If we are OK up to 25 mph, then we have about 50% useful wind

We should expect about 220,000 kWh/year from our 50 kW turbine

Bin Min Max Hr/Year % Avg m/s

1 0 0 80 0.9 0

2 0 1 204 2.3 0.5

3 1 2 496 5.7 1.5

4 2 3 806 9.2 2.5

5 3 4 1211 13.8 3.5

6 4 5 1254 14.3 4.5

7 5 6 1246 14.2 5.5

8 6 7 1027 11.7 6.5

9 7 8 709 8.1 7.5

10 8 9 549 6.3 8.5

11 9 10 443 5.1 9.5

12 10 11 328 3.7 10.5

13 11 12 221 2.5 11.5

14 12 13 124 1.4 12.5

15 13 14 60 0.7 13.5

16 14No limit 2 0.02no record

8760

0

200

400

600

800

1000

1200

1400

1 2 3 4 5 6 7 8 9 10111213141516

Hr/Year

Hr/Year



No Detail Data, what then?

Rayleigh Distribution (special case of Wiebull Distribution) can be applied if average wind speed is known.

Calculates approximate bins based on Uavg, and the shape of the Rayleigh statistical curve

Accuracy depends on site conditions



Comparison of observed and Rayleigh estimated probabilities of

wind speeds in a given bin for wind speeds up to 14 m/s

Average wind speed 5.57 m/s

0

200

400

600

800

1000

1200

1400

1 3 5 7 9 11 13 15



Statistical Wind Mapping

Uses statistical approach to predict wind at sites

without exhaustive measuring

Data inputs required

Prevailing winds

Elevation

Terrain

Calibrate & verify accuracy of mapping by comparing

to sites with known, measured wind data

Is the wind map’s prediction for these sites sufficiently

accurate?

Companies like AWS Truewind specialize in wind

mapping.



Wind Data Cautions

A Raleigh approximation might fit one set of

measured data very well, but may not accurately

predict another.

Small differences between predicted and actual are

important because available power varies with the

cube of wind speed!

Power per unit swept area, P = 0.5ρU3

ρ = air density, on the order of 1 kg/m3

Varies with elevation and climate

Example Pavg = 0.5(1.15)(5.57)3 = 99 W/m3

P14 = 0.5(1.15)(14)3 = 1578 W/m3

Big Difference



Figure 13-9. Comparison of observed and Rayleigh estimated

probabilities of wind speeds in a given bin for wind speeds up to 14 m/s



Observed vs. Estimated Power

Observed Raleigh for Uavg = 5.57 m/s Annual Power Output

Bin Min Max Bin % Min Cum Bin % Observed Estimated

1 0 0 0.9 0.00% 0.00% 0.0 0.0

2 0 1 2.3 0.00% 2.50% 0.0 0.0

3 1 2 5.7 2.50% 7.13% 1.0 1.2

4 2 3 9.2 9.63% 10.74% 7.2 8.5

5 3 4 13.8 20.37% 12.93% 29.8 27.9

6 4 5 14.3 33.31% 13.59% 65.6 62.4

7 5 6 14.2 46.89% 12.91% 119.0 108.2

8 6 7 11.7 59.80% 11.27% 161.8 155.9

9 7 8 8.1 71.07% 9.14% 172.1 194.2

10 8 9 6.3 80.21% 6.92% 194.9 214.1

11 9 10 5.1 87.13% 4.91% 220.2 212.2

12 10 11 3.7 92.05% 3.28% 215.7 191.2

13 11 12 2.5 95.33% 2.06% 191.5 158.0

14 12 13 1.4 97.39% 1.22% 137.7 120.5

15 13 14 0.7 98.61% 0.69% 86.8 85.1

16 14 No limit 0.02 99.30%

1603.5 1539.3

From Table 13-1

2)/)(4/(1 avgbin

binwindspeed eP

35.0 UP

• Very good agreement in this case

• But that is what we expected from the good curve fit

Another Approach

Multiply the average by (6/π)1/3 (the limit

of the average of the cube over the cube of

the averages)

(5.57)(2.66134) = 6.91m/s

P = 0.5(1.15)(6.91)3 = 18.97 W/m2

Annual output = 18.97)(8760) = 1662 kWh/m2

Using the straight average wind speed of 5.57 m/s

would have resulted in a low estimate of only 870

kWh/m2. This highlights the importance of the

higher wind speeds because of the ‘cubed effect’.



Relative value of wind speed and power as a function of height above

the ground, indexed to value at 30 m. = 1.00.

396 W

8.86 m/s

Wind speed increase rapidly at first, but then less rapidly with height. We use a

shear coefficient to estimate the relative values of wind versus height. Power

can then be calculated. See example 13-3 on pp 445 for more detail.



Figure 13-11. Seasonal distribution of wind speed in m/s by month of

year at proposed Enfield, New York, wind farm site (height =58m)



Figure 13-12. Hourly distribution of wind speed for example 365 days of

year at proposed Enfield, New York, wind farm site (height = 58m)



Figure 13-13. Wind rose from 1-year wind data for proposed wind

turbine site at Ithaca College, Ithaca, New York

Source: Prof. Beth Clark, Ithaca College



Turbine Operating Modes

Cut-in Speed – below which blades do not spin

Operating Range – Output increases rapidly with speed

High Operating Range – Output increases slow and at a point

load must be reduced

Fixed pitch blades are allowed to stall

Variable pitch blades reduce their pitch

Rated Wind Speed – speed of maximum output, will either hold

steady or decline with increases wind speed

Cut Out Speed – put on the brakes



Figure 13-14. Power curve for 1.5-MW turbine for wind

speeds from 0 to 21 m/s, with extraction efficiency



Table 13-3. Calculation of annual output for turbine with

power curve from Fig 13-14 and empirically measured

distribution at hypothetical site with Uavg= 8.4 m/s



Modeling Aerodynamic Behavior

of Wind Turbines

Levels of modeling:

The Actuator Disc Model assumes an idealized rotor

with infinite number of blades.

Strip Theory incorporates blade geometry.

Computational fluid dynamics allow more accurate

models.

All are Limited

Most overpredict

Models can provide an upper limit, like Carnot



Actuator Disk Analysis

Rotates in plane perpendicular to wind

Composed of infinite number of blades

Translate wind energy into rotation

Like a Carnot engine

Cannot be built in practice

Useful to estimate theoretical limits

Assumptions

Incompressible fluid flow

No drag or friction

Uniform thrust over disk surface

No wake rotation

Steady state



Introduction to Strip Theory

Divides blades into sections (strips or elements) which

can be analyzed separately.

Sometimes called Blade Element Theory.

Propellers put energy into the fluid (air or water) stream.

Wind tubines extract energy from the stream.

Blade element approximation good for HAWT analysis

Expanding wake puts shed vortices outside of rotor diameter

Unlike air or water propellers



Power Coefficient, CP

The Ratio of Power Extracted to Power Available

Translating Devices Move with the Wind

Sailboats for example

Efficiency limited by velocity relative to the wind, and

by drag coefficient

Power Coefficients on the order of 15%

Wind Turbine Blade

Angle of Attack

A function of relative velocity



Figure 13-16. Winds and forces acting on a cross section of

a turbine blade, showing angle of attack between midline

of cross section and relative wind velocity Vr

Note that the blade is rotating up the page



Figure 13-17. Relative value of power coefficient CP/CP-max as a

function of ratio of v/U to (v/U)max, using Eq. (13-29)



Tip Speed Ratio and Advance Ratio

TSR “λ”: λ = ΩR/U where

Ω is rotation speed in rad/s

R is swept area radius (length of blade) in m

U is free wind speed in m/s

Advance Ratio J: J = 1 / λ



Figure 13-18. Rotor power coefficient CPr as function of tip

speed ratio



Table 13-4. Turbine performance as a function of for representative

turbine with 10-m blade radius and fixed wind speed U = 7.0 m/s



Economics of Wind

Choice of site using wind data

Land requirements

Cost elements: Installed capital cost: turbines + balance of system: roads,

cables, substations, etc

Operation & maintenance (~2% of cap charge)

Insurance, royalties, land rentals (~ 0.5%)

Balance of system costs ~ 20% of total capital costs for on-shore systems

Other factors e.g. Volatility of natural gas price gives wind a market advantage

Proximity to transmission line

Air quality permitting not required (advantage over fossil fuels)



Land Requirements for

Large-Scale Wind Production

Suppose we replace 500 average coal plants = 1 trillion kwhr/yr (100 million US avg. homes)

Take Fenner site as standard:

$1.5 M per turbine

7 turbines per sq mi

4.4 M kwhr per turbine per year

Number of turbines required: 230,000

Compare: goal of 230,000 MW installed in Europe by 2020

Area required: 33,000 sq mi (21% of North & South Dakota combined)

Cost: approx $350 billion



Effect of technological

advance on cost

Optimal size tradeoff: Energy increases with square of radius, turbine cost increases

with cube of radius

In functional form:

Net Benefit from increasing size = a*R2 - b*R3

Coefficients “a” and “b” depend on the technology

Technological improvement reduces value of coefficient “b” >> larger turbines

Advantage of slow speed: Less noise, fewer bird kills

Only possible with large turbines (~500 kW+), since they can better absorb gearing losses



Design Issues and Adverse Local

Environmental Impacts

Placement: given upwind obstacle w/ height H

Turbine should be min 2xH above and 20xH

downwind from obstacle

Visual effect on surroundings

Use photomontage to show visual effect of turbines

Noise: must meet local zoning requirements for

not exceeding noise limits



Layout Considerations for Wind Parks

Each turbine must have enough space around the post to rotate in any direction

Turbines in a line perpendicular to prevailing wind must have 2x rotor radius space to avoid collisions

Turbines along line of prevailing wind must have 5x to 10x rotor radius to avoid negative effects of turbulence



Impact on birds

Relative quantity of bird fatalities from collisions with turbines

Some number of birds killed each year due to striking turbines

However, the number is small compared to fatalities from cars, windows, cats, etc.

Changing turbine technology has helped birds

Slower rotation speed

Compare 100 kW turbine in California in 1980s vs 1.5 MW turbine today

Change from lattice-work to solid tower

Turbine tower does not encourage nesting

Wind energy helps birds over long term since it slows climate change



Future Objectives for Wind Power

Maximize turbine blade size (150 – 200 m diameter)

Offshore turbines on floating platforms, moored to bottom

Dynamic feathering of blades in real time to respond to changes in wind, maximize output

Urban wind turbine concepts for building tops Possibly large HAWTs atop buildings

Large or small VAWTs also envisioned



Using Software to Optimize

Wind Farm Layout

Data requirements

Topographical data

Wind data

Technology characteristics of turbines

Given number of devices, terrain, wind, optimizes location to maximize output



Integrating Wind into the Grid

Conventional (“dispatchable”) supply:

Baseline

Load following

Peaking

Goal: use up capacity with lowest variable cost first

If wind energy is available, use first

Load-following plants adjust in real time to changing demand, changing availability of wind

This works up to a point

With sufficient presence of wind in the energy mix, may need additional infrastructure to offset

Resources and Technologies - ece.citadel.eduece.citadel.edu/peeples/ELEC 427/Wind Energy.pdf · Source: F Vanek, L Albright, and L Angenent. (2012) Energy Systems Engineering: Evaluation

Documents