Guidelines for Design of Wind Turbines − DNV/Risø

in which U10,15 = 15 m/s is a reference wind N

speed, IT,15 is the characteristic value of the
turbulence intensity at 15 m/s, and a is a
I T , total = m (1 − N ⋅ pw ) I T m + pw ∑I
i =1
T ,w
m
( si )

slope parameter. IT,15 = 0.18 and a = 2 are to

be used in the category for higher turbulence pw = 0.06
characteristics, while IT,15 = 0.16 and a = 3
are to be used in the category for lower si = xi / D
turbulence characteristics. The expression
for the characteristic value σU,c is based on a
1
definition of the characteristic value as the IT , w = 2
+ IT 2
mean value of σU plus one standard (1.5 + 0.3 ⋅ si ⋅ v )
deviation of σU.
N number of closest neighbouring wind
3.1.3 Turbulence intensity turbines
m Wöhler curve exponent corresponding
The turbulence intensity IT is defined as the
to the material of the considered
ratio between the standard deviation σU of
structural component
the wind speed, and the 10-minute mean
v free flow mean wind speed at hub
wind speed U10, i.e. IT = σU/U10. height
pw probability of wake condition
Note that the presence of a wind turbine will xi distance to the i'th wind turbine
influence the wind flow locally, and that the D rotor diameter
turbulence in the wake behind the turbine IT free flow turbulence intensity
will be different from that in front of the IT,w maximum turbulence intensity at hub
turbine. This phenomenon of a wind turbine height in the centre of the wake
influenced turbulence is known as a wake
effect. Typically, the presence of the wind The number of closest neighbouring wind
turbine will lead to increased turbulence turbines N can be chosen as follows:
intensity in the wake. Wake effects need to
be considered for wind turbines installed 2 wind turbines: N=1
behind other turbines with a distance of less 1 row: N=2
than 20 rotor diameters. This is of particular 2 rows: N=5
interest wherever wind farms with many in a farm with more than 2 rows: N=8
turbines in several rows are to be installed.

The following method, Frandsen 2001, can

be used to take wake effects into account.
By this method, the free flow turbulence
intensity IT is modified by the wake
turbulence intensity IT,w to give the total
turbulence intensity IT,total. In the evaluation
of the wake effect, a uniform distribution of
the wind direction is assumed. The formulas
can be adjusted if the distribution of wind
direction is not uniform. Reference is made Figure 3-3. Example of determination of neighbouring
to Frandsen, 2001. wind turbines.

36 3 – External Conditions
 Guidelines for Design of Wind Turbines – DNV/Risø

Figure 6-27. Relation between wind speed, yaw error and yaw moment.

Figure 6-27 shows a computer simulation of
the yaw moment as a function of the yaw
error and wind speed for a three-bladed
wind turbine. It appears that the yaw moment
oscillates with frequencies equal to the rotor
frequency and three times the rotor
frequency. It also appears that the sign of the
yaw moment changes when the yaw
direction changes, and – not surprisingly –
the yaw moment increases when the yaw
error and the wind speed increase.
6.10.2 Yaw drive
The yaw drive is the system of components
used to cause the yaw motion. A large yaw
speed will produce gyro effects that will
result in large loads on the wind turbine. The
yaw speed must therefore be small enough
for gyro effects to become negligible.
Reference is made to Section 4.2.1.
A passive yaw system will normally require
some sort of damping arrangement to reduce
the yaw speed.
For large turbines with active yaw systems,
the yaw speed is usually lower than 1°/sec,
which is small enough for the gyro effects to
be ignored. To achieve such a low yaw
speed, the yaw motors needs to be connected
through a gearbox.
Such drives can be delivered as standard
equipment from manufacturers of
electrically operated motors.
The yaw drives must have sufficient power
to overcome the largest mean yaw moment

6 – Nacelle 161
 Guidelines for Design of Wind Turbines − DNV/Risø

Large blade deflections are typically H

encountered in extreme wind situations, but
also in the operating condition when the yaw
Fz (h) = FzT + ρt ∫ A( z) dz ⋅ g
h
angle is large or when the terrain is sloping.
Atypical wind profiles may also give rise to M z (h) = M zT
large blade deflections. Special attention
should therefore be given to wind turbine
designs that are sensitive to special load F y ( h) = F yT
+ F w (h)
cases such as negative wind shear in
complex terrain. M x (h) = M xT + F yT ⋅ ( H − h) + M w (h)
+ FzT ⋅ (δ ( H )-δ (h))
The current Danish practice uses validated
stiffness data in the aeroelastic calculations
MzT
for analysis of deflections. Validated
stiffness data are blade stiffness data, which
comply with the experimental data from FzT
static tests of the blades. This implies that
FyT MxT
one may have to produce two sets of
aeroelastic calculations. The first is a load wind
calculation, in which the model is tuned to
the correct natural frequencies and damping
properties of the structural system. The
second is a special deflection model where
the deflections are tuned to be in agreement H
with the static experiments. Ideally, the two
models would be the same, but in practice,
this is not always the case.

7.4 Tubular towers

7.4.1 Loads and responses h
For the purpose of calculating section loads
in the tower, the tower can be viewed as a
cantilever beam as shown in Figure 7-4.
External loads, denoted by index T in this Figure 7-4. Cantilever beam model of a tubular tower
figure, are applied at the tower top flange, subject to loading at the level of the hub.
which is located at a height H above the
tower base. Note that this height may Fy thrust from wind load
deviate somewhat from the hub height. Mx bending moment from wind load
Fz gravity force
Section loads in the tower at height h can be Mz torsional moment
calculated from the loads applied at the top ρt density of tower including
of the tower: appurtenances
A(z) cross-sectional area as a function of
height z
δ deflection of tower due to thrust
from wind

7 – Tower 173
 Guidelines for Design of Wind Turbines − DNV/Risø

scenarios leads to the most critical bearing
capacity for the actual foundation.
e Aeff
Reference is made to Figure 8-2. The
effective area representation that leads to the
poorest or most critical result for the bearing
capacity of the foundation is the effective
area representation to be chosen.

For a circular foundation area with radius R,

an elliptical effective foundation area Aeff
LC1 can be defined as

leff
 e 
Aeff = 2 R 2 arccos( ) − e R 2 − e2 
 R 

beff with major axes

Aeff
be = 2(R − e )
e
and

2
 b 
le = 2 R 1 − 1 − e 
 2R 
LC2
leff

Aeff

beff
R
Figure 8-2. Quadratic footing with two approaches to
how to make up the effective foundation area. LC
leff
le

Scenario 1 corresponds to load eccentricity

with respect to one of the two symmetry
axes of the foundation. By this scenario, the
following effective dimensions are used:
beff

beff = b − 2 ⋅ e , l eff = b
Scenario 2 corresponds to load eccentricity
with respect to both symmetry axes of the
foundation. By this scenario, the following
effective dimensions are used:
be
Figure 8-3. Circular and octangular footings with
effective foundation area marked out.
The effective foundation area Aeff can now
be represented by a rectangle with the
following dimensions

beff = leff = b − e 2

8 – Foundations 191
 Guidelines for Design of Wind Turbines − DNV/Risø

ES is the value of E at a depth z equal to the
pile diameter D, and EP is the Young’s
modulus of the pile material.
The formulas given for foundation
stiffnesses in the tables in this section can be
used to calculate spring stiffnesses to
support the tower in aeroelastic wind turbine
analyses. As a rule of thumb, the natural
frequency of the tower will be reduced by
0% to 5%, when the assumption of a rigid
foundation (fixed-ended tower) is replaced
by a realistic finite foundation stiffness.
Under special conditions this error may
however be up to 20%.

Table 8-5. Circular footing on stratum over bedrock or on stratum over half-space.
On stratum over bedrock On stratum over half-space

Mode of motion Foundation stiffness Foundation stiffness

R
1 + 1.28
4GR R 4G1 R H ; 1≤H/R≤5
Vertical KV = (1 + 1.28 ) KV =
1 −ν H 1 −ν 1 R G1
1 + 1.28
H G2
R
1+
8GR R 8G1 R 2 H ; 1≤H/R≤4
Horizontal KH = (1 + 0.5 ) KH =
2 −ν H 2 −ν 1 1 + R G1
2 H G2
R
1+
8GR 3 R 8G1 R 3 6 H ; 0.75≤ H/R ≤2
Rocking KR = (1 + ) KR =
3(1 − ν ) 6H 3(1 − ν 1 ) R G1
1+
6H G2
16GR 3
Torsion KT = Not given
3

204 8 – Foundations
 Guidelines for Design of Wind Turbines − DNV/Risø

Table 8-6. Circular footing embedded in stratum over bedrock

Mode of motion Foundation stiffness

4GR R D D D/H
Vertical KV = (1 + 1.28 )(1 + )(1 + (0.85 − 0.28 ) )
1 −ν H 2R R 1− D / H
8GR R 2D 5 D
Horizontal KH = (1 + )(1 + )(1 + )
2 −ν 2H 3 R 4H
8GR 3 R D D
Rocking KR = (1 + )(1 + 2 )(1 + 0.7 )
3(1 − ν ) 6H R H
16GR 3 8D
Torsion KT = (1 + )
3 3R

Table 8-7. Flexible pile.

Standardised springs at pile head
Soil profile KH K K H ,R
Horizontal Rocking 3 R Coupled 2
DE S D ES D ES
Linear increase E 
0.35
E 
0.80
E 
0.60

with depth E = 0.6 P 

 0.14 P 
 − 0.17 P 

ESz/D  ES   ES   ES 
Increase with
0.28 0.77 0.53
square-root of E  E  E 
depth E = 0.8 P 
 0.15 P 
 − 0.24 P 

 ES   ES   ES 
ES z / D
0.21 0.75 0.50
Homogeneous E  E  E 
1.08 P 
 0.16 P 
 − 0.22 P 

E = ES
 ES   ES   ES 

8 – Foundations 205
 Guidelines for Design of Wind Turbines – DNV/Risø

E.1.2 Alloy steel

The values given in Table E-4 apply to
smooth specimens with diameter d [mm] or
thickness t [mm].

Table E-4. Properties of selected alloyed steel.

Grade d ≤ 16 16< d ≤ 40 40 < d ≤ 100 100 < d ≤ 160 160 < d ≤ 250
t≤8 8 < t ≤ 20 20 < t ≤ 60 60 < t ≤ 100 100 < t ≤ 160
fy fu fy fu fy fu fy fu fy fu
1000 - 900 - 800 - 750 - 700 -
34CrMo4 800 650 550 500 450
1200 1100 950 900 850
1100 - 1000 - 900 - 800 - 750 -
42CrMo4 900 750 650 550 500
1300 1200 1100 950 900
1100 - 1000- 900 - 850 - 800 -
50CrMo4 900 780 700 650 550
1300 1200 1100 1000 950
1100 - 1000 - 900 - 800 - 750 -
36CrNiMo4 900 800 700 600 550
1300 1200 1100 950 900
1200 - 1100 - 1000 - 900 - 800 -
34CrNiMo6 1000 900 800 700 600
1400 1300 1200 1100 950
1250 - 1250 - 1100 - 1000 - 900 -
30CrNiMo8 1050 1050 900 800 700
1450 1450 1300 1200 1100
fy yield strength [MPa]
fu tensile strength [MPa]

E.2 Cast iron

Table E-6. Mechanical Mat lamina, Weaved
The values given in Table E-5 are valid for a properties for GRP. 30 weight roving
wall thickness below 50 mm. For other % glass lamina, 50
dimensions the values must be reduced in weight %
glass 1
accordance with DIN 1693 and proper Values to be used in
reduction factors for surface roughness and strength calculation
notches must be included. Tensile strength 80 MPa 170 MPa
Compressive strength 100 MPa 100 MPa
Table E-5. Properties of selected cast iron (minimum Bending strength 150 MPa 170 MPa
values). Shear strength
Grade Tensile Yield Elongation interlaminar 10 MPa 10 MPa
strength strength at rupture in plane 20 MPa 20 MPa
fy [MPa] fu [MPa] δ [%] Flexural modulus 4 000 MPa 8 000 MPa
GGG-35.3 350 220 22 Values to used in de-
GGG-40 400 250 15 formation calculation
GGG-40.3 400 250 18 Modulus of elasticity 6 000 MPa 12 000 MPa
GGG-50 500 320 7 Shear modulus 2 200 MPa -
GGG-60 600 380 3 Poissons ratio 0.35 -
1. Applies to balanced web, loaded in the main
direction.

E.3 Fibre Reinforced Plastics

These values are to be reduced in
E.3.1 Glass fibre reinforced plastics accordance with DS 456 depending on
temperature, load duration, cyclic load and
Table E-6 shows the guiding values given in
fabrication method.
the Danish design code for the mechanical
properties of glass fibre reinforced plastics.

E – Materials Properties 269