Amplitude modulation for different weather conditions Conny Larsson Uppsala University, Sweden

EWEA Technology Workshop: Wind Turbine Sound 2014

Economical support from Swedish Energy Agency 2010-2014 Results from two Wind Turbine sites: Forest, flat ground; Forest, heterogeneous terrain Special analyses of AM sound presented here.

Measurements

Rythmic sound heard at distance

•  Changes of the sound level occur due to the weather but does the weather also change the character of the sound at an immission point far away?

•  Oerlemans demonstrated that mainly the downward movement of the turbine blades, as perceived on the ground, produces the swish sound.

•  According to van den Berg, the synchronizing of sound between different WTs during propagation in a stable atmosphere could explain the thumping noise heard at greater distances from WTs.

Measurement sites

Photo: Hans Blomberg

Photo: Svevind

Dragaliden

Ryningsnäs

Dragaliden, 1 km

Ryningsnäs, 400 m

AM study, 1 year at Ryningsnäs and Dragaliden

•  Measure 8 times/s (Fast) •  Divide into 15 s long periods and analyze. •  Observations at the sites and studies of different

amplitude modulation spectra •  Our definition: AM if the AMS-value (periodicities

between 0.6 – 1.0 Hz) is ≥ 0.4

Lundmark, Measurement of swish noise, a new method, Fourth International Meeting of Wind Turbine Noise, INCE/Europe, Rome, Italy (2011).

C. Larsson, and O. Öhlund, Amplitude modulation of sound from wind turbines under various meteorological conditions, J. Acoust. Soc. Am. 135(1), 2014, 67-73

Amplitude modulation

Sound level recording from the Dragaliden site from 20 July 2011 at different clock hours. The measurements are taken 3 km from the closest WT. To the right is the amplitude modulation spectra with AMS value for a given 15 s period.

Day

Night

15:42:15 15:42:30

28

30

32

34

Time

Inst

anta

neou

sso

und

leve

l (dB

A)

0 1 2 30

0.2

0.4

0.6

Frequency (Hz)

AMS

valu

e (d

B)

00:01:30 00:01:45

28

30

32

34

Time

Inst

anta

neou

s s

ound

leve

l (dB

A)

0 1 2 30

0.2

0.4

0.6

Frequency (Hz)

AMS

valu

e (d

B)

Amplitude modulation

Sound level recording from the Dragaliden site from 20 July 2011 at different clock hours. The measurements are taken 3 km from the closest WT. To the right is the amplitude modulation spectra with AMS value for a given 15 s period.

Day

Night

15:42:15 15:42:30

28

30

32

34

Time

Inst

anta

neou

sso

und

leve

l (dB

A)

0 1 2 30

0.2

0.4

0.6

Frequency (Hz)

AMS

valu

e (d

B)

00:01:30 00:01:45

28

30

32

34

Time

Inst

anta

neou

s s

ound

leve

l (dB

A)

0 1 2 30

0.2

0.4

0.6

Frequency (Hz)

AMS

valu

e (d

B)

AMS < 0.4

AMS ≥ 0.4

AM AM

No AM No AM

AM heard at distance.

00:27:15 00:28:0440

45

50In

stan

tane

ous

sou

nd le

vel (

dBA)

00:27:15 00:28:0426

28

30

32

34

36

Inst

anta

neou

s s

ound

leve

l (dB

A)

100 101 102 103 104 105−20

−15

−10

−5

0

5

10

15

20

25

30

Soun

d pr

essu

re le

vel (

dBA)

Frequency (Hz)

Emission (upper left) and immission (down left) sound level measurements and immission sound spectrum for the given time frame at Dragaliden site. Distance 2-4 km from the WT.

100 Hz

1000 Hz

Ryningsnäs, Småland Nordex 2500 LS (2,5 MW) hub height 80 and100 m

Ryningsnäs, 400 m

0 100 200 3000

20

40

60

Wind direction (o)

Per

cent

AM

of t

otal

tim

e in

spe

cific

sec

tor (

%)

a)

0 10 20 30 400

20

40

60

Turbulence intensity100m (%)

Per

cent

AM

of t

otal

tim

e in

spe

cific

bin

(%)

b)

−0.1 −0.05 0 0.05 0.10

20

40

60

Sound speed gradient (s−1)calculated between 120−0 m

Per

cent

AM

of t

otal

tim

e in

spe

cific

bin

(%)

c)

−50 0 500

20

40

60

Solar elevation angle (o)

Per

cent

AM

of t

otal

tim

e in

spe

cific

bin

(%)

d)

Percentage of AM in a specific bin for the Ryningsnäs site.

Dragaliden, Enercon-E82 (2 MW) WTs hub height 108 och 138 m

Dragaliden, Enercon-E82 (2 MW) WTs hub height 108 och 138 m

0 100 200 3000

20

40

60

Wind direction (o)

Per

cent

AM

of t

otal

tim

e in

spe

cific

sec

tor (

%)

a)

0 10 20 30 400

20

40

60

Turbulence intensity100m (%)

Per

cent

AM

of t

otal

tim

e in

spe

cific

bin

(%)

b)

−0.1 0 0.10

20

40

60

Sound speed gradient (s−1)calculated between 120−0 m

Per

cent

AM

of t

otal

tim

e in

spe

cific

bin

(%)

c)

−40 −20 0 20 400

20

40

60

Solar elevation angle (o)

Per

cent

AM

of t

otal

tim

e in

spe

cific

bin

(%)

d)

Dragaliden, 19 - 21 July 2011, 2-4 km distance

00:00 00:000

1

2

AM fa

ctor

im

mis

sion

poi

nt (d

B)

a)

00:00 00:0020

40

60

I

nsta

ntan

eous

so

und

leve

l im

mis

sion

poi

nt (d

BA)

b)

00:00 00:0020

30

40

Cal

cula

ted

imm

issi

on s

ound

leve

l (dB

A)

c)

00:00 00:0012141618202224

Tem

pera

ture

(o C)

d)

T3mT100m

00:00 00:00

200

250

WD 11

1m (o )

e)

00:00 00:000

10

20

TI11

9m (%

) f)

00:00 00:000

1

2

AM fa

ctor

e

mis

sion

poi

nt (d

B)

g)

00:00 00:005

10

Win

d sp

eed

(m/s

)

h)

u51mu98m

Ampl i tude modu la t ion (percentage of selected measurements) Evening-night-morning Ryningsnäs (400 m) 33 % Dragaliden (1 km) 19 %

Results – AM during 1 year

AM more common during specific meteorological conditions

Conclusions

•  The weather governs the immission sound level even for high elevated sources (WT).

•  The weather governs the character of sound from WTs – Amplitude modulation.

•  Amplitude modulation 20 - 30 % of time. •  Most frequent at evening-night-morning at stable

stratification.

19