Turbine Design

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 19

13-5.

Turbine Design
The main functional goal of the turbine design is to convert as much wind energy as possible into mechanical energy of the
rotating drive shaft, over the operating range of wind speeds of the turbine. Rotational speed of the shaft changes with
changing wind speed, so the power electronics of the generator is used to match electrical current leaving the turbine with the
correct frequency, phase, and voltage demanded by the electrical grid. The operating range consists of several regimes (which
are illustrated graphically by the device in Fig. 13-14):

Below the cut-in speed of the wind (3 m/s in the example turbine), there is insufficient power in the wind to turn the blades,
and the device does not operate.

In the range of wind speeds from the cut-in speed upward, turbine output increases rapidly toward its maximum value.

Since power increases with cube of wind speed, at some wind speed above the cut-in speed (9.5 m/s in the example 1.5-
MW turbine), it becomes difficult to extract the maximum available power from the wind, due to the large force exerted on
the blade. From this point upward, as wind speed continues to increase, the fraction of the available energy extracted by
the turbine begins to decline. Output continues to grow with increasing wind speed, but at a slower rate. This limiting of
output is done for engineering reasons, as cost of fabricating a turbine that would be sufficiently strong to convert all of
the wind force into rotational energy in the shaft would be prohibitive. Also, in many locations wind speed rarely reaches
the very highest values, so it would be of little commercial value to be able to capture a large fraction of the available
energy. Reducing load on the turbine blades at high wind speeds can be accomplished in one of two ways:

In fixed-pitch devices, where the position of the blade in the hub is fixed in one position, the wind passing over the
turbine blade enters a stall regime, where the stalling of the blade slows its rotation and reduces the power extracted
from the airstream. (The pitch of the blade is the angular orientation of the blade around its own axis, relative to the
plane of rotation of the entire turbine. See Sec. 13-5-3 for a more detailed description.) These devices are also known
as "stall-regulated."

In variable-pitch devices, the blades can rotate in the hub so that they extract less power from the wind as it passes.
While smaller devices currently available in the market include a mix of fixed- and variable-pitch devices, utility-scale
devices currently being installed are all variable pitch.

At some wind speed that is usually above that of maximum extraction efficiency, the device reaches itsrated wind speed,
or wind speed where the device first puts out its maximum output (13 m/s in the 1.5-MW example). Above the rated wind
speed, output may either hold approximately constant or decline with increasing wind speed, with pitch-regulated turbines
typically able to achieve the former, and stall-regulated turbines subject to the latter.

In order to protect from damage at the highest wind speeds, many devices have a cut-out speed of the wind (21 m/s in the
example), above which the turbine drive shaft rotation is stopped by some combination of changing the yaw or pitch so
that the blades are no longer catching wind, or through dynamic braking of the generator. Thereafter, a mechanical brake
is applied to prevent the device from rotating.

The device must also be capable of a controlled stop in case of failures of the electric grid, other emergencies, or other
routine events such as scheduled maintenance. Rotation of the nacelle around the vertical axis to face the wind is
accomplished either by the force of the wind itself in smaller devices, or by active mechanical rotation in utility-scale devices.

13-5-1. Theoretical Limits on Turbine Performance


One important observation about kinetic energy in wind is that it is not physically possible to design a device that extracts all
energy from the wind, for if one did, the physical mass of air being moved by the wind would stop in the catchment area

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
downstream from the device, preventing upstream air from reaching the device and acting on it. (This observation is
analogous to the Carnot limit on heat engines, which states that it is not possible to achieve 100% efficiency in a practical heat
engine because the exhaust gases are always expelled to the cold reservoir at a temperature above 0 K.) It is of interest to
quantify this limit on energy that can be extracted from the air in a theoretical model of an optimal wind device, to be used as a
benchmark against which to compare empirically measured turbine performance in actual devices.

13-5-1-1. Theoretical Evaluation of Translating Devices


This concept can be explored by first considering the case of a simple translating wind-catching device, which moves in the
same direction as the wind. Examples of such devices include a sailing vessel "running" with the wind (i.e., being pushed from
behind), an iceboat being pushed across an icy surface, or vertical-axis grain-grinding devices used in antiquity. In the case of
the sailboat or iceboat, the force acting to overcome friction is a function of the drag coefficient of the device against the
surface over which it is translating, and is a function of the difference between the wind speed and the speed at which the
device is traveling, for example, per unit of area of the surface struck by the wind:

F = [0.5ρ(U − v)2]CD

(13-13)

where F is the force in N/m2, v is velocity at which the device is traveling in m/s, and CD is the dimensionless drag coefficient
of the device against the supporting surface.

The power P extracted from the wind per unit area, expressed in W/m2, can then be written as

P = [0.5ρ(U − v)2]CDv

(13-14)

Since the power available in the wind is equal to 0.5 ρU3, we can now write the power coefficient CP as the ratio between the
power extracted by the device and the power available in the wind:

[0.5ρ(U − v)2]CDv [(U − v)2CDv]


CP = =
0.5ρU 3 U3

(13-15)

Using differential calculus, it is possible to show that the maximum value of CP that can be achieved by such a device is
4/27CD or 0.148CD (the solution is left as an exercise at the end of the chapter).

This low value for the translating device is not a fair measure of the value of a sailing device used for transportation, since the
passenger is concerned with travel between points and is not concerned with how much power the device is extracting from
the wind. For devices intended to do stationary work, such as a mechanical windmill or wind turbine, however, it is clear that
the translating device will not be effective for converting wind kinetic energy into mechanical or electrical energy. We
therefore turn our attention to rotating devices.

13-5-1-2. Rotating Devices and the Betz Limit

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
In this section, we will again solve for the maximum power extraction possible, relative to the power available in free wind with
speed U, this time for a device that rotates in a plane perpendicular to the direction of the wind. An "actuator disk" model of the
turbine is used for the purpose of theoretical modeling. The actuator disk can be thought of as a rotating disk with an infinite
number of blades that converts the translation of air in the wind into rotation of the disk. An actuator disk is therefore like a
Carnot heat engine, in that it is a device that cannot be built in practice but is useful for setting the upper limits on efficiency.

As shown in Fig. 13-18, airflow through the actuator disk starts with speed upstream from the disk equal to the free wind
speed, that is, U1 = U. As the kinetic energy is removed from the wind at the disk, the speed is reduced to U2 = U3, that is, the
airflow speed is the same just upstream and just downstream of the disk. Finally, the wind speed is further reduced to the "far
wake" speed, U4.

Figure 13-18 Diagram of airstream flow through the actuator disk.

For the actuator disk, the following assumptions apply

1. Incompressible fluid flow

2. No frictional drag on the device

3. Uniform thrust over the entire disk area

4. Nonrotating wake

5. Steady-state operation, that is, constant wind speed velocity at points along the flow through the disk, constant disk
rotational speed

The ratio of reduction in airflow speed at turbine to free wind speed (U1− U2)/U1 is called the axial induction factor a. The
reduction in airflow speed from the free stream speed U1 to the far wake speed U4 is the result of the wind exerting thrust on
the actuator disk, that is,

T = M air(U1 − U4) = ρAU2(U1 − U4)

(13-16)

Here Mair is the mass of air flowing through the disk of area A in units of kg/s.

Another fundamental starting point for analysis of the actuator disk is Bernoulli's law, which states that the sum of energy
measured in terms of pressure, velocity, and elevation is conserved between any two points along a fluid stream:

ρV12 ρV22
p1 + + ρgz1 = p2 + + ρgz2
2 2

(13-17)

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
where p is pressure, V is airflow velocity, g is the gravitational constant, and z is height.

It can be shown using the relationship between thrust and airstream velocity, and the relationship between velocity and
pressure (Bernoulli's law), that the reduction in wind speed relative to U in the far wake is always twice the amount at the plane
of the disk (this is left as an end-of-chapter exercise). Thus by rearranging terms the following relationships hold

U1 − U4 U1 − U2
= 2( )
U1 U1
a = 1 − U2/U1
U2 = U1(1 − a)
U4 = U1(1 − 2a)

(13-18)

In the remainder of this section, we will use Bernoulli's law and the function of the actuator disk to calculate the value of a at
which the maximum power is extracted from the wind, and the resulting maximum value. Assuming that change in height is
negligible, we can write equations for the airflow on the upside and downside of the disc, as follows:

p1 + 1/2 ρU12 = p2 + 1/2 ρU22


p3 + 1/2 ρU32 = p4 + 1/2 ρU42

(13-19)

The pressure difference across the disk can then be rewritten as

p2 − p3 = p1 + 1/2 ρU12 − 1/2 ρU22 + 1/2 ρU32 − p4 − 1/2 ρU42


= p1 − p4 + 1/2 ρ(U12 − U42)
= 1/
2
ρ(U12 − U42)

(13-20)

since U2 = U3 and the pressure in the far wake and the free wind are both equal to the ambient pressure. The result of Eq. (13-
20) can be used to write an equation for the thrust T in units of N acting on the disk in terms of U1 and U4, as follows:

T = A2(p2 − p3)
= 0.5ρA2(U12 − U42)

(13-21)

where A2 is the area of the disk. Next, since power P is the product of thrust and wind velocity at the disk, we can rewrite the
power in the device as follows:

P = (T ) U2 = 1/2 ρA2(U12 − U42) U2 = 1/2 ρA2U2(U1 + U4)(U1 − U4)

(13-22)

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
Substituting for U2 and U4 in terms of U1 and a, and replacing U1 with the free stream velocity U gives:

P = 1/2 ρA2U(1 − a)[U + U(1 − 2a)][U − U(1 − 2a)] = 1/2 ρA2U 34a (1 − a)2

(13-23)

The power coefficient CP of the disk is then

0.5ρA2U 34a(1 − a)2


CP = = 4a(1 − a)2 = 4a3 − 8a2 + 4a
0.5ρA2U 3

(13-24)

To find the maximum value of Eq. (13-24), we differentiate with respect to a and set equal to zero:

dCp
= 12a2 − 16a + 4
da
12a2 − 16a + 4 = 0

This expression simplifies to

3a2 − 4a + 1 = 0

The roots of the equation are then a = 1/3 and a = 1. The maximum value of CP = 0.593 occurs when a = 1/3. Thus the
maximum theoretically possible power extraction of 59.3% occurs when the wind velocity has been reduced by 1/3 at the disk
and hence by 2/3 at the far wake. This result is attributed to Betz and is given the name Betz limit.

The actuator disk model can be made more accurate by adding realism to the underlying assumptions, which has the effect of
lowering the maximum energy theoretically available from the wind below the value of 59.3% derived above. For example, in a
more realistic model of turbine rotation, the wake from the turbine rotates in the opposite direction from the rotation of the
rotor. Since the energy transferred to the rotating wake is not available for transfer to the turbine, wake rotation will tend to
reduce the maximum value of CP.

13-5-1-3. Maximum Thrust and Maximum Efficiency


The maximum value of thrust as a function of axial induction is derived in a similar way. The value of the thrust can be
rewritten in terms of a as

T = 2ρA2aU 2(1 − a)

(13-25)

The maximum value occurs when a = 0.5, so that wind speed in the far wake is reduced to zero. Theefficiency ηdisk of the
actuator disk is defined as the power extracted divided by the power available in the airstream moving at the turbine. The
power available at the turbine is thus a function of the mass flow rate ρA 2U2, rather than the free wind mass flow rate ρA 2U.
The value of efficiency is

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
ηdisk = [ 1/2 ρA2(U12 − U42)U2]/ 1/2 ρA2U12U2 = 4a (1 − a)

(13-26)

Thus the value of ηdisk , like the thrust, reaches a maximum value at a = 0.5, when the far wake wind speed is 0. The maximum
value of ηdisk is 1. At a = 0.5, CP = 0.5, while at a = 1/3, the values of T/Tmax and ηdisk are both 0.888.

13-5-2. Tip Speed Ratio, Induced Radial Wind Speed, and


Optimal Turbine Rotation Speed
In this section, we move from discussion of the actuator disk with infinite number of blades to a practical wind turbine with
finite number of airfoil-shaped blades (typically two or three). The turbine responds to wind by rotating at an angular velocity
Ω, measured in rad/s, that is determined by the airfoil design, load on the turbine's generator, and other factors.[1]

13-5-2-1. Tip Speed Ratio, Advance Ratio, and Effect of Changing Tip
Speed Ratio on Power Extracted by Blade
The ratio of the translational velocity of the tip of the turbine blade to the wind speed is called the tip speed ratio (TSR), and is
calculated as follows:

λ = Ω R/U

(13-27)

where λ is the TSR, R is the radius of the swept area of the turbine, and U is the free wind speed. A related measure to λ is the
advance ratio, denoted J, which is the inverse of the tip speed ratio, or J = 1/λ. The effect of the force of the wind on a cross
section of the airfoil of the turbine blade varies with λ. This effect is examined in Fig. 13-19.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
Figure 13-19 Winds and forces acting on a cross section of a turbine blade, showing angle of attack
α between midline of cross section and relative wind velocity Vr. Note that in this example the
blade is rotating up the page.

The rotation of the rotor about its axis induces a wind speed v in the opposite direction of the direction of rotation. Therefore,
the blade encounters a relative wind velocity Vr, which is the vector sum of U and v. The angle of attack, α, is the angle between
Vr and the center line of the airfoil cross section.

A real-world airfoil has nonzero lift coefficient CL and drag coefficient CD; the ratio of CL to CD is called the lift-to-drag ratio. (In
some calculations, the inverse drag-to-lift ratio CD/CL is used instead.) A high lift-to-drag ratio is desirable for efficiently
converting energy in the free wind into motion of the blade. Using the geometric relationship between the directions of lift,
drag, and the direction of motion of the blade, power output from the blade can be written in terms of CL, CD, and the ratio of
induced to free wind speed v/U as follows:

1/2
P = 0.5ρU 3A (v/U)[CL − CD(v/U)][1 + (v/U)2]

(13-28)

where A is the cross-sectional area of the blade facing the direction of the wind. The power coefficientCP of the blade is the
ratio of the power delivered to the rotation of the blade cross section, divided by the power available in the free wind passing
through the projected area of the cross section.[2] Solving for CP, the ratio simplifies to

1/2
CP = (v/U)[CL − CD(v/U)][1 + (v/U)2]

(13-29)

In Fig. 13-20, CP/CP-max is plotted as a function of the ratio of v/U to (v/U) max, where (v/U) max is the value where CP is reduced
to zero due to the effect of the drag term in Eq. (13-29). As v/U increases from 0, at first CL dominates the value of Eq. (13-29),
so that CP increases. Beyond a maximum point for CP at v/U = 2/3(v/U) max, however, the effect of drag dominates and CP
decreases at higher values of v/U.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
Figure 13-20 Relative value of power coefficient CP/CP-max as a function of ratio of v/U to (v/U)max ,
using Eq. (13-29).

Note that the power coefficient calculated in Eq. (13-29) is a comparison of the force acting locally on a cross section of the
blade, and not the force acting on the entire swept area of the turbine. Therefore, values greater than 1 are possible, because
induced values of v much greater than U are possible. The entire device will not, however, violate the Betz limit, and the overall
value of CP for an entire device with nonzero CD will always be less than 0.593.

13-5-2-2. Rotor Power Coefficient


The rotor power coefficient is the power coefficient of a physical rotor with finite number of blades, and is denoted CP,r as
opposed to CP defined for the actuator disk in Eq. (13-24). The rotor power coefficient is illustrated in this section using a
hypothetical turbine with 10 m radius, CL = 1.0, and CD = 0.1. CP,r is the ratio of output P from the turbine to power in the wind,
and is a function of λ:

P (λ)
CP ,r(λ) =
0.5ρU 3Aturbine

(13-30)

where Aturbine is the swept area of the turbine. Let us assume a constant wind speed and vary the turbine rotation speed and
hence λ. Here we will plot a curve for CP,r(λ) based on a sample of empirically observed data points for P(λ). For wind speed U
= 7 m/s, the following performance data points are given in Table 13-4. The four points can be used to characterize the curve
of CP,r versus λ as shown in Fig. 13-21.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
Table 13-4 Turbine Performance as a Function of λ for Representative Turbine with 10-m Blade Radius and Fixed Wind Speed
U = 7.0 m/s

Point Angular velocity (rad/s) TSR Power (kW) CP,r

1 2.8 4.0 4.0 0.064

2 4.2 6.0 11.5 0.185

3 5.6 8.0 20.0 0.323

4 7.0 10.0 26.4 0.426

5* 7.9 11.3 27.9 0.450

6 8.4 12.0 27.4 0.442

7 9.8 14.0 19.6 0.317

Note: *Maximum rotor power coefficient of 0.450 achieved at Ω = 7.9 rad/s, TSR = 11.3.

Figure 13-21 Rotor power coefficient CP,r as function of tip speed ratio λ.

Since λ is proportional to v, as λ increases and U remains fixed, the ratio v/U will also increase. Therefore, in keeping with Eq.
(13-29), as rotation speed increases from a minimum of 2.8 rad/s at point 1, the value ofCP,r at first increases due to the
increased lift exerted by the wind, then reaches a maximum at point 3 (7.9 rad/s), then declines as the force of drag comes to
dominate.

13-5-2-3. Relationship between Ideal Turbine Operation and Tip Speed


Ratio
As discussed above, for a given fixed wind speed, there is a TSR value that maximizes output, on either side of which output
declines as TSR moves away from this ideal range. Thus for each combination of wind speed and output in Fig. 13-14, there is

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
a corresponding single TSR value since there is a single turbine rotational speed associated with that point on the power
curve. Looking at the entire range of wind speeds to which the turbine is exposed, the range can be divided into two, based on
the relationship between wind speed and TSR, as follows:

1. Maximum efficiency range: As wind speed increases from the cut-in speed, ideal turbine operation seeks to keep TSR in an
ideal range to maximize output. For a utility-size turbine, TSR values between λ = 6 and λ = 8 are typical. Note that due to
continuous slight variations in exact wind speed and direction, it is not possible to maintain a precise value for TSR;
however, as long as the value is kept within a narrow range, the effect on conversion efficiency is acceptable.

2. Maximum output range: As wind speed continues to increase, at some point before the device reaches the rated wind
speed, the goal for ideal turbine operation shifts from maximizing efficiency to maintaining the rated output of the device
without excessive physical stress on the various components, at the expense of lowering TSR. Thus wind speed continues
to increase but the tip speed (product of swept area radius and angular velocity) increases only modestly compared to the
highest values achieved in the maximum efficiency range, and approaches an asymptotic value.

To illustrate the two regimes, we can revisit the representative 1.5-MW turbine in Fig. 13-14, with a radius of 34.5 m. The
maximum efficiency range occurs between wind speeds of approximately 4 and 10 m/s, which for ideal λ values of 6 or 8
gives tip speeds of between 24/32 and 60/80 m/s. Thereafter, the regime shifts to the maximum output range, where output
stabilizes at the rated 1.5 MW and efficiency declines eventually to values below 10%. Although tip speed at and above the
rated wind speed cannot be deduced from Fig. 13-14, it remains in the range of 80–100 m/s, with TSR falling correspondingly.
For example, at the cut-out speed of 21 m/s, if λ = 8 were maintained, tip speed would reach 176 m/s, or nearly 400 mph.
Because it is not practical to engineer a turbine blade for such high translational speeds, a lower TSR is maintained: at a tip
speed of 100 m/s and wind speed of 21 m/s, λ = 4.8 and Ω = 2.9 s–1 (28 RPM).

13-5-3. Analysis of Turbine Blade Design


Up to this point we have considered the rudimentary design of a wind extraction system and set some bounds on its optimal
performance. In order to design the actual blades of a turbine, it is necessary to model them explicitly and consider how their
shape will affect the ability to extract energy from the wind. For this purpose, we introduce an approach called strip theory, in
which the blade is divided into infinitesimal elements called strips, and the overall performance of the blade is taken to be the
sum of strip performance. Thereafter, we consider an approximate solution to the blade design problem in which the blade is
divided into a finite number of equally sized segments.

The application of strip theory depends on the following assumptions:

1. The performance of the individual strip in a "stream tube" of air (i.e., annular ring through which air is passing uniformly)
can be analyzed independently from any other stream tube.

2. Spanwise flow along the length of the turbine blade is negligible, so that the two-dimensional cross section of the blade
appearing in the strip can be used as a basis for analysis without loss of accuracy.

3. Uniformity of flow conditions around the circumference of the stream tube of air and uniformity of blade geometry within
the strip or element.

The reader should understand the limitations of these assumptions. For example, in practice, flow conditions do vary for a
turbine from top to bottom of its swept area due to changing flow conditions that vary with height from the ground, especially
for large devices where the span of the swept area may be 50 m or more. Also, the use of strip theory introduces two new
aspects of turbine behavior that were ignored previously. First, the possibility of wake rotation, which was ignored in the
derivation of the Betz limit, is considered, in the form of an induced tangential velocity v. Second, the blade can be thought of
as a cantilever beam extending out from the turbine hub, so that under the force of the wind, the blade will deflect in a
downwind direction from its plane of rotation.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
In order to evaluate a turbine blade design, we need to know lift and drag coefficients for the blade. These coefficients are
evaluated empirically and published for standard airfoil shapes. For example, starting in the 1930s, the National Advisory
Council on Aeronautics (NACA) in the United States carried out testing to develop "families" of airfoil shapes denoted by the
term NACA followed by a number (e.g., NACA 0012, NACA 63(2)-215, and so on). Other organizations besides NACA have also
developed standard airfoil shapes. Regardless of the source, choice of a standard shape saves time for the engineer since he
or she does not need to carry out empirical or theoretical work to derive drag and lift coefficients for an original airfoil shape.

13-5-3-1. Continuous Solution to Blade Element Design Problem Using


Strip Theory
The purpose of this section is to develop a methodology for calculating available power based on airflow and turbine blade
shape. The diagram in Fig. 13-22 introduces several new variables that are required for the methodology. The figure shows a
strip of width dr at a location r along the length of the blade from 0 to R, deflecting by a coning angle θ in response to free wind
U. In this example, the wind is traveling perpendicular to the plane of rotation of the blade, that is, along the x-axis of the
system. In practice, there will often be a yaw error ΔΨ, not shown in Fig. 13-22, which is introduced due to the turbine not
remaining perfectly aligned with the wind at all times. To take account of the coning angle, the distance along the axis of the
blade from the strip to the axis of rotation is given the value s such that r = s cos θ.

Figure 13-22 Variables used in analysis of blade element in annular ring.

Figure 13-23 shows the relationship between the axial and radial winds, Vn and v. The relative wind Vr seen by the blade is the
vector sum of Vn and v. The angle between Vr and the centerline of the cross section of the blade is called the angle of attack α.
The angle between the plane of rotation and the centerline is called the blade pitch β. The sum of α and β is the wind angle φ,
that is,

ϕ = α+β

(13-31)

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
Figure 13-23 Relationship between angle of attack α, pitch β, and wind angle φ.

We now introduce the radial induction factor a′, which is the amount by which v exceeds the angular velocity at the point of
analysis r along the blade, that is, rΩ. The axial and radial winds can be written in terms of free wind speedU, rotation speed Ω,
r, and a′, as follows:

Vn = U(1 − a) cos θ

(13-32)

v = rΩ (1 + a′) = s cos θ Ω (1 + a′)

(13-33)

Having thus introduced all necessary variables for strip theory, we develop a method for estimating the value of the local
thrust coefficient Ct for the strip that is being considered. This analysis assumes that the local value of axial induction a is
known or can be fixed in advance. The case where a is not known is considered hereafter.

To begin with, the thrust coefficient is the thrustdT divided by the available force in the streamtube, that is,

dT
Ct =
0.5(ρ)(U 2)(2πrdr)

(13-34)

where 2πrdr is the cross-sectional area of the stream tube.

The magnitude of dT is then the force acting on an area the width of the chord of the airfoil c, acting in the lift and drag
directions, multiplied by the number of blades B, that is,

dT = 0.5 ρVr2Bc(CL cos ϕ + CD sin ϕ) dr

(13-35)

Ct then simplifies to

C = ( /2 )( / )(V /U)2(C cos ϕ + C sin ϕ)


© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
Ct = (B/2π)(c/r)(Vr/U)2(CL cos ϕ + CD sin ϕ)

(13-36)

From the definition of relative wind, Vr can be rewritten as

Vr = Vn/ sin ϕ = U(1 − a) cos θ / sin ϕ

(13-37)

Substituting for Vr in Eq. (13-36) gives

Ct = (B2π)(c/r)(1 − a)2(cos2θ /sin2ϕ)(CL cos ϕ + CD sin ϕ)

(13-38)

In order to evaluate φ, it is first necessary to evaluate the induced radial velocity a′. From the angular relationship between
axial and lateral wind components, two equations for φ can be written in terms of a and a′, as follows:

tan ϕ = U(1 − a) cos θ /[r Ω (1 + a′)]

(13-39a)

tan ϕ = a′r Ω /(a cos θ U)

(13-39b)

Define x to be the local speed ratio, or x = r Ω/U. Note that the tip speed ratio introduced above is in fact a special case of the
local speed ratio where r = R. Substituting x for r Ω/U and setting Eqs. (13-39a) and (13-39b) equal to each other gives

a (1 − a)cos2θ = a′(1 + a′)x2

(13-40)

Now we can evaluate a′ in terms of a, and φ from a and a′. We therefore have sufficient information to evaluate local Ct. The
overall thrust coefficient for the turbine can be evaluated by integrating over the span of the blade from the root (point where
blade joins the turbine hub) to the tip, that is,


T = 0 dT

(13-41)

In practice, Ct is adjusted by a tip-loss factor in regions close to the outer end of the blade. The tip loss factor takes into
account the losses incurred by the interaction of the airstream near the tip of the blade with air outside the swept area of the
device. This factor has the strongest effect at points along the radius at within 3% of the length of the blade from the tip.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
Quantitative evaluation of tip loss is beyond the scope of this chapter, and the reader is referred to other works on wind energy
engineering in the references.

Example 13-5 applies Eqs. (13-39) to (13-41) to the calculation of Ct at a point along a blade.

Example

Example 13-5

A turbine has three blades of 10 m length, and rotates at 6 rad/s in a 7 m/s wind. At a point 7.5 m from the axis of rotation
a = 0.333, CL = 1, CD = 0.1, and chord ratio = 0.125. Assume that coning is negligible. CalculateCt at that point.

Solution First calculate the local speed ratio, x = r Ω/U = 7.5(6)/7 = 6.43. Since there is no coning,θ = 0. Use Eq. (13-40)
to solve for a′:

(0.333)(1 − 0.333)cos2(0) = a′(1 + a′)(6.43)2

Substituting known values reduces this equation to a′2 + a′ – 0.00537, for which the only feasible solution is a′ = 0.00535.
Substituting into Eq. (13-39a) gives

tan ϕ = a′r Ω/a cos θ = (0.00535)(7.5)(6)/(0.333 ⋅ 7) = 0.1032

Therefore, φ = 5.89°. Using φ in Eq. (13-37) gives the following value for Ct:

Ct = (B2π)(c/r)(1 − a)2(cos2θ /sin2ϕ)(CL cos ϕ + CD sin ϕ)


= (3 × 2π)(0.125/7.5)(1 − 0.333)2[1/sin2(5.89)][(1) cos(5.89) + 0.1 sin(5.89)] = 0.338

Ultimately, it is of interest to calculate the overall power coefficient CP for the entire blade. By integrating across the range of
possible values for x from 0 to λ, it is possible to calculate CP using the following:


CP = (8/λ2) 0 x3a′(1 − a)[1 − (CD/CL) cot ϕ] dx

(13-42)

In some instances, CL and CD may be known as a function of α, based on the chosen airfoil, but both a and a′ may be unknown.
For such a situation, an iterative procedure based on the following equation can be used:

CL = 4 sin ϕ(cos ϕ − x sin ϕ)/[σ′(sin ϕ + x cos ϕ)]

(13-43)

Here σ′ is the local solidity of the blade, defined as σ' = Bc/2πr. The procedure involves first guessing at values of a and a′, then
calculating φ from Eq. (13-39a) or (13-39b), then calculating angle of attack from α = φ − β, then calculating CL, then

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
calculating CL and CD using α, then using CL and CD to calculate new values of a and a′. This process is repeated until the
incremental change in a and a′ is within some predetermined tolerance.

13-5-3-2. Approximate Value of Maximum Power Coefficient


When carrying out an initial assessment of a turbine design, it is sometimes desirable to quickly calculate the value of CP,max.,
as shown in Example 13-6. The following empirical equation has been derived for this purpose:

λB0.67 1.92λ2B
CP ,max = 0.593 [ − CD/CL]
1.48 + (B0.67 − 0.04)λ + 0.0025λ2 1 + 2λB

(13-44)

Here CD/CL is the drag to lift ratio at the design angle of attack α of the turbine.

Example

Example 13-6

Refer to the hypothetical 1.5-MW turbine in Fig. 13-14. Assume CD/CL = 0.01 and that the turbine has 3 blades. (a) If the
turbine rotates at 1 rad/s in a 6 m/s wind, what is the predicted value of CP,max at that wind speed? (b) Produce a graph of
CP,max as a function of tip speed ratio λ up to λ = 15 for both CD/CL = 0.01 and CD/CL = 0, and state the maximum value of
each curve. (c) Compare the maximum value in part (b) to the value observed in Fig. 13-14.

Solution

a. Based on the turbine radius of 34.5 m, the tip speed ratio is

Ωr 1(34.5)
λ= = = 5.75
U 6

Substituting other known values into Eq. (13-43) to solve for CP,max gives

λB0.67 1.92λ2B
CP ,max = 0.593 [ − CD/CL]
1.48 + (B0.67 − 0.04)λ + 0.0025λ2 1 + 2λB

= 0.593 [
5.75(3)0.67 1.92(5.75)23
− (0.01)] = 0.5019
1.48 + (30.67 − 0.04)5.75 + 0.0025(5.75)2 1 + 2(5.75)3

b. Repeating the calculation in part (a) yields the graph shown in Fig. 13-24. For the case of CD/CL = 0.01, the curve
reaches a maximum value of CP,max = 0.5050 at λ = 8 (to the nearest whole number value of λ), and declines thereafter
for increasing λ. For the case of CD/CL = 0, since there is no drag coefficient term, asλ increases, the value of CP,max
approaches the Betz limit without bound and there is no fixed maximum value.

c. The turbine represented in Fig. 13-14 has a maximum value of CP = 0.4893 at U = 9.5 m/s, so the curve for CD/CL = 0.01
is a reasonable predictor of the upper bound on CP, although the actual turbine is slightly less efficient since the
approximation in Eq. (13-41) does not fully capture real-world turbine behavior.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
Figure 13-24 CP,max as a function of tip speed ratio for CD/CL = 0.00 or CD/CL = 0.01.

13-5-3-3. Approximate Solution to Blade Design Problem by Blade


Element Model Analysis
As an alternative to integrating thrust over the length of the turbine blade as proposed in Eq. (13-41), it is possible to perform
a discrete approximation of the thrust or power extracted by the turbine under given wind and blade conditions by dividing the
blade into a limited number of sections and analyzing the performance at the midpoint of each section. This approach is
known as blade element model (BEM) analysis. Another approach known as computational fluid dynamics (CFD) models are
used for detailed modeling of utility-scale turbines since the high cost per turbine justifies the additional time and expense of
building the models. CFD models are beyond the scope of this chapter, but BEMs provide a stepping stone toward
understanding the more complex development of CFD models.

The first step is to evaluate the lift coefficient as a function wind angle, local speed ratio, and solidity, using Eq. (13-45), and
then the characteristic equation for lift as a function of α, based on the airfoil shape. The lift coefficient equation for the airfoil
will typically be some function of α, for example, a linear function for a representative airfoil might be CL = 0.46352 +
0.11628α. The correct value of α is one that satisfies both equations for the section under study, that is, the lift coefficient
predicted by the airfoil design equals the lift predicted from airflow conditions.

For each segment of the blade, we need data on solidity σ′, local speed ratio x, and pitch angle β. In some cases, the twist θT
for the blade at the location r is given, rather than β, where θT is defined as

θ T = β − β0

(13-45)

where β 0 is the pitch angle at the tip of the blade. Note that β 0 may be positive or negative, and that typically its absolute value
is small (|β 0| < 3°). Also, x can be calculated from the ratio of location r to length of blade R

x = (r/R)λ

(13-46)

Calculation of β, σ′, and x makes it possible to calculate α for the segment. Once α, φ, CL, and CD are known, it is possible to
calculate the local contribution to power coefficient CP,i using the following:

8Δλ
( )
© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
8Δλ
CP ,i = ( ) sin2ϕ i (cos ϕ i − x sin ϕ i)(sin ϕ i + x cos ϕ i)[1 − (Cd/Cl) cot ϕ i]x2
λ2

(13-47)

Here Δλ is the change in x from one segment to the next, that is, Δλ = xi – xi − 1, which is a constant value since the width of the
segments is uniform. The overall power coefficient for the blade is the sum of all the local values of CP,i for each segment i for
each segment 1, 2, … , N. The total power produced by the blades is then the power in the wind multiplied by the overall power
coefficient, that is, Pprod = CP × Pavailable. Example 13-7 illustrates the calculation of the local contribution from a segment to
CP; the calculation for an entire blade is left as an exercise at the end of the chapter.

Example

Example 13-7

A two-bladed wind turbine is designed with the following characteristics. The blades are 13 m long, and the pitch at the tip
is –1.970°. The turbine has a design TSR of λ = 8. Lift coefficient is a function ofα as follows (note, α is in degrees): For α <
21°: CL = 0.42625 + 0.11628α − 0.00063973α2 − 8.712 × 10−5α3 – 4.2576 × 10−6α4. For α > 21: CL = 0.95. For the entire
blade, the drag coefficient is CD = 0.011954 + 0.00019972α + 0.00010332α2. For the midpoint of section 6 (r/R = 0.55), find
the following for operation at a tip speed ratio of 8: (a) angle of attack α; (b) angle of relative wind θ; (c) CL and CD; (d) the
local contribution to CP. Ignore the effects of tip losses. The chord at that point is 0.72 m and the twist is 3.400°.

Solution First calculate factors needed for calculating α: r = 0.55(13) = 7.15 m; σ′ = Bc/2πr = 2(0.72)/2π (7.15) = 0.0321; x
= rλ/R = 0.55(8) = 4.4; β = θT + β 0 = 3.400 – 1.970 = 1.430. We can now solve forα using the lift coefficient equation for
the airfoil design and Eq. (13-43). The exact solution using a computer solver is α = 7.844°, which gives CL = 1.241 from
both equations. In other words, given the values of r, σ′, x, and β, and the derived value of φ = α + β = 7.844 + 1.430 =
9.274°, the following hold:

CL = 4 sin ϕ(cos ϕ − x sin ϕ)/[σ′(sin ϕ + x cos ϕ)] = 1.241


CL = 0.4265 + 0.11628α − 0.00063973α2 − 8.712 × 10−5α3 − 4.2576 × 10−6α4 = 1.241

Table 13-5 shows values of CL as a function of α around the solution.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
Table 13-5 Comparison of Approaches to Calculating Lift Coefficient CL as a Function of Angle of Attack α

Lift Coefficient CL

Angle of Attack α (deg.) From Eq. ( 13-43) From Airfoil Design

7.70 1.2716 1.2289

7.75 1.2610 1.2331

7.80 1.2503 1.2372

7.85 1.2394 1.2413

7.90 1.2283 1.2454

7.95 1.2170 1.2495

8.00 1.2055 1.2535

Thus answer to part (a) is α = 7.844°. Part (b), φ = α + β = 7.844 + 1.43 = 9.274. Part (c), CL = 1.241, CD = 0.011954 +
0.00019972α + 0.00010332α2 = 0.01988. Part (d), for the contribution to CP,i, we first calculate Δλ = xi – xi − 1 = 4.4 – 3.6 =
0.8. Then solve for CP,i using Δλ, λ, φ, x, CL, and CD in Eq. (13-47), which gives CP,i = 0.0567.

13-5-4. Steps in Turbine Design Process


In the previous section, we investigated the analysis of the performance of a given blade design. It is also necessary to be
able to start with a desired level of performance from a turbine and design a blade that meets the requirements.

The selection of a blade design can be seen in the larger context of designing the overall wind turbine system. This design
process can be summarized in the following steps:

1. Determine the radius of the rotor based on the desired power at typical wind speeds.

2. Choose a desired tip speed ratio based on function, that is, for electric generationλ > 4 is typical, for mechanical
applications such as water pumping a lower value of λ is chosen.

3. Choose a desired number of blades.

4. Select an airfoil, and obtain all necessary data (e.g., CL and CD data).

5. Create an ideal shape for the blade, based on the chosen airfoil and dividing into 10 to 20 elements. Note that optimal
airfoil shapes are typically too difficult to fabricate exactly, so that some modification will be necessary.

6. Create an approximation of the ideal shape that provides ease of fabrication without greatly sacrificing performance.

7. Quantitatively evaluate blade performance using strip theory.

8. Modify shape of blade, based on results of step 7, so as to move blade design closer to desired performance.

9. Repeat steps 7 and 8, trading off blade performance and ease of production, until a satisfactory design is obtained.

A complete example of this procedure from start to finish is beyond the scope of this chapter. A partial example is presented
as a problem at the end of the chapter. As a starting point for the design procedure outlined in steps 1 to 9, Table 13-6
provides parameters for two representative ideal blade designs. The data given in the table provide the information necessary

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
for steps 2, 3, 5, and 6 in the turbine design process. The two representative ideal blades shown in the table have constant CL =
1.00 and CD = 0.

Table 13-6 Representative Three-Blade and Two-Blade Designs, in Relative Dimensions

Rotor 1 Rotor 2

λ = 6, B = 3 λ = 10, B = 2

r/R φ c/R φ c/R

0.15 32.0 0.191 22.5 0.143

0.25 22.5 0.159 14.5 0.100

0.35 17.0 0.128 10.6 0.075

0.45 13.5 0.105 8.4 0.060

0.55 11.2 0.088 6.9 0.050

0.65 9.6 0.076 5.8 0.042

0.75 8.4 0.067 5.1 0.037

0.85 7.4 0.059 4.5 0.033

0.95 6.6 0.053 4.0 0.029

Overall solidity 0.088 0.036

Source: After Manwell et al. (2002), p. 124. John Wiley & Sons Limited. Reproduced with permission.

Note that values are given here independent of the exact radius of the turbine, and the overall solidity is the percent of the
swept area that is covered by the area of the blades. For example, if rotor 1 has a blade length of 10 m, then the total swept
area is ~314 m2, and of this area, a viewer facing the turbine at rest would see 8.8% or 27.6 m2 of blade, and the rest open
space. A typical application of rotor 1 might be in a medium to large turbine (10 kW to 2.5 MW), whereas rotor 2 would
typically be applied to a small turbine, with a rated capacity of 2 kW or less. The turbine shapes above are ideal in the sense
that they are designed without taking into account drag on the blade (CD = 0); they assume a constant lift coefficient of CL =
1.0.

[1] Conversion: 2π radians = 1 revolution = 360°.


[2] Note that C P in Eq. (13-29) is different from the power coefficient introduced in Eq. (13-24).

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.

You might also like