November 2008Oregon TWS Meeting

The Dalles, Oregon

Wallace EricksonWestern EcoSystems Technology, Inc.

Cheyenne, Wyoming

Outline

• Fatality Impacts• Avoidance/Minimization• Collision Risk Modeling• Cumulative Impacts• Habitat

History of the Avian/Wind Turbine Issue

Result: New wind plants often heavily scrutinized

• Altamont Pass, CA– 7000+ turbines– high number of raptor fatalities discovered

• golden eagles (30-70 fatalities per year)• red-tailed hawks (300-500 fatalities per

year)• American kestrels (several hundred)• Burrowing owls (50-1000? per year)

100 kW turbine

660 KW turbine

Not exactly to scale

Illustration from APWRA of scale issues in Per turbine metric

Fatality monitoring studies covering all seasons of occupancy for a 12 month period adjusted for carcass detection and carcass removal biases

New Project since NWCC 2001 Summary

Data reported in NWCC 2001 SummaryAdditional data collected since NWCC 2001 Summary

Condon, OR Top of Iowa

Altamont, CA

Tehachapi Pass, CASan Gorgonio, CA

Montezuma Hills, CA

Algona IowaMeyersdale, PASomerset, PANWTC, CO

Searsburg, VTCombine Hills, OR

Hopkins Ridge, WA

NPPD Ainsworth, NE

High Winds, CA

Diablo Winds, CA

Maple Ridge, NY

Buffalo Gap

ALL REGIONS11%

3%

74%

1%1%

2%1%6%

1%

Doves/Pigeons

Gamebirds

Other Birds

Passerines

Rails/Coots

Raptors/Vultures

Shorebirds

Unidentified Birds

Waterbirds

Waterfowl

Proportion of fatalities at sites reporting fatalities by species, summarized for All Regions Where studies have been conducted (Pacific North West Mid-West, Rocky Mountains, and East (sources in literature cited).

National Composition

East Region

81%

2%2%

2%2%

6%3%

2%Doves/Pigeons

Gamebirds

Other Birds

Passerines

Rails/Coots

Raptors/Vultures

Shorebirds

Unidentified Birds

Waterbirds

Waterfowl

Midwest Region3% 2%

78%

2%

2%

1%1%

5% 6%

Doves/Pigeons

Gamebirds

Other Birds

Passerines

Rails/Coots

Raptors/Vultures

Shorebirds

Unidentified Birds

Waterbirds

Waterfowl

Rocky Mountain

86%

1% 1% 1%4%

1%6%

Doves/Pigeons

Gamebirds

Other Birds

Passerines

Rails/Coots

Raptors/Vultures

Shorebirds

Unidentified Birds

Waterbirds

Waterfowl

Pacific Northwest

18%

2%

69%

7%1%

1%1% 1%

Doves/Pigeons

Gamebirds

Other Birds

Passerines

Rails/Coots

Raptors/Vultures

Shorebirds

Unidentified Birds

Waterbirds

Waterfowl

Pacific Northwest

• 636 records• 73 species• Horned lark• Golden-crowned

kinglet• American kestrel

and red-tailed hawk

Species Number ofFatalities

Percent Composition

horned lark 202 31.8 golden-crowned kinglet 41 6.4 ring-necked pheasant 35 5.5 gray partridge 34 5.3 chukar 22 3.5 western meadowlark 21 3.3 American kestrel 21 3.3 unidentified passerine 21 3.3 European starling 17 2.7 dark-eyed junco 15 2.4 mourning dove 15 2.4 white-crowned sparrow 13 2.0 red-tailed hawk 12 1.9 yellow-rumped warbler 10 1.6 rock pigeon 10 1.6 unidentified bird 9 1.4 winter wren 8 1.3 short-eared owl 8 1.3 ruby-crowned kinglet 8 1.3 black-billed magpie 5 0.8 northern flicker 5 0.8 unidentified sparrow 5 0.8 house wren 5 0.8 Townsend’s warbler 5 0.8 Brewer’s sparrow 5 0.8 red-breasted nuthatch 4 0.6 American robin 4 0.6 Canada goose 4 0.6 common nighthawk 4 0.6 unidentified kinglet 4 0.6

Fatality Monitoring Objectives

• Determine whether overall avian and bat fatality rates or raptor fatality rates are low, moderate, or high relative to other projects

• Determine whether predicted mortality is a reasonable estimate

• Determine whether a wind project has a fatality problem

Bird Fatality Estimates

0

2

4

6

8

10

12

14

SG ALT SL VAN KLO NC FC1 FC2 WIS BR1 BR2 BR3 BM MO

# bi

rd fa

talit

ies/

MW

/yr

CA – old turbinesPacific NorthwestRocky MountainUpper MidwestEast

Bird Fatality Estimates

0

2

4

6

8

10

12

14

SG ALT SL VAN KLO NC FC1 FC2 WIS BR1 BR2 BR3 BM MO

# bi

rd fa

talit

ies/

MW

/yr

CA – old turbinesPacific NorthwestRocky MountainUpper MidwestEast

CA – old turbinesPacific NorthwestRocky MountainUpper MidwestEast

Other Fatality Monitoring Objectives

• Compare rates to exposure or relative abundance

• Estimate the influence of weather on fatality levels

• Estimate the effect of lighting • Determine effectiveness of

minimization or mitigation measures

• Similar methods and metrics for across site comparisons

Progression of fatality metrics

• Altamont = ~0.1/turbine/year or 1/MW/yr (Orloff 1992)

• Per turbine estimates from a few studies– Assumes fatality rates the same for 600 kw to

1.8 MW• Per Nameplate MW or per RSA estimates• Would like actual power production

2.0

3.3 3.0

5.9

1.41.0

2.8 2.92.6

1.0

3.1

1.3

11.7

2.7

1.81.4

2.52.0

2.5

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

WI B1 B2 B3 TI VA NC SL CH K1 K2 HR BM MO DW HW F1 F2 AI

# B

ird F

atal

ities

/ M

WSummary of fatalities for all Birds displayed by landscape type for 19 sites.

Wind ProjectAgriculture Ag / Grass / CRP Forest Grass / Steppe

0

0.04

0 0 0.01 0

0.05

0.09

0 0

0.110.14

00.02

0.56

0.42

0.05 0.06 0.07

0

0.1

0.2

0.3

0.4

0.5

0.6

WI B1 B2 B3 TI VA NC SL CH K1 K2 HR BM MO DW HW F1 F2 AI

# R

apto

r Fat

aliti

es /

MW

Summary of fatalities for raptors displayed by landscape type for 19 sites.

Wind ProjectAgriculture Ag / Grass / CRP Forest Grass / Steppe

Fatality Rates (#/MW/yr)Project

RaptorsAll

birdsNocturnal Migrants Source

Combine Hills, OR 0 2.6 0.27 Young et al. 2005Klondike, I OR 0 0.9 0.35 Johnson et al. 2003a

Klondike II, OR 0.11 3.1 2.11 NWC and WEST, 2007Vansycle, OR 0 1 0.32 Erickson et al. 2000Stateline, WA/OR 0.09 2.9 0.73 Erickson et al. 2004Hopkins Ridge, WA 0.14 1.2 0.46 Young et al. 2007Nine Canyon, WA 0.05 2.8 0.45 Erickson et al. 2003Wild Horse, WA 0.09 1.6 0.88 Erickson et al. 2008Bighorn I, WA 0.15 2.6 0.57 Kronner et al. 2008Mean 0.07 2.2 0.68

Raptor Use Raptor Fatality

Facilities

High Winds

Diablo Winds

Hopkins Ridge

Klondike

Klondike II

Stateline

Nine Canyon

Foote Creek Rim

Vansycle

Buffalo Ridge

Combine Hills

Range: 0 – 0.15/MW/yr

Limitations and Supporting Studies

• All raptors treated equally• Migration vs. Resident• Need data to fill in gaps• Altamont (2008)

– Monthly use correlated with monthly fatality rates for red-tail hawks

– No correlation for American kestrels

Proportion of flights below and within the rotor swept area from studies conducted in the Pacific

NorthwestPacific Northwest StudiesFlight Height Characteristics

Species 0 - 25m 26 - 125 mAmerican kestrel 82 15Ferruginous hawk 42 50Red-tailed hawk 45 45Golden eagle 15 67Northern harrier 85 10Swainson's hawk 23 67

From select studies: not a full summary of all studies

• Offshore GE 3.6 MW 104 meter rotor diameter

Slide provided by Bob Thresher, NREL

Avian Strike Probability Versus Turbine Size

15 Meter Diameter and 100 kW

93 Meter Diameter and 2.5MW

Altamont ScaleNext Generation Scale

Large vs. Small Turbines• Diablo Winds – 660 kW turbines

in the Altamont• Other Turbines in the Altamont

(40-150 kW)– All raptors – 63% fewer raptors

killed per MW/yr– American kestrels - 84% fewer– Red-tailed hawks - 34% fewer– Golden eagles – none found– Burrowing owls - 34% fewer

• Reference (Altamont Monitoring Team 2008)

Lighting and Birds• Meta-analysis

suggests no large differences in fatality rates at lit and unlit turbines

• Gehring (2008) work in Michigan supports this result

Summary of Comparison of Fatality Rates for Lit and Unlit TFatility Comparison

Project lit minus unlit rate Stat. sign.Klondike - not sign.Stateline + not sign.Nine Canyon + not sign.High Winds + not sign.Wisconsin + not sign.Vansycle + not sign.Mountain + not sign.Big Horn - not sign.Wildhorse + not sign.

eer

Siting

• Micro-Siting Examples–Updrafts–Habitat–Saddles–Nests

EAGLES

Eagle Observation

Project Area Boundary

Section Boundary

0 ½ 1 2

A

Count LocationApproximate Raptor Point

T20N

T19N

25 30

20

19 20

30

31 33

8

1

12

13

21

Arlington

4

B

C

D

E

F

A

Rim Edge

1999

EAGLES

Eagle Observation

Project Area Boundary

Section Boundary

0 ½ 1 2

A

Count LocationApproximate Raptor Point

T20N

T19N

25 30

20

19 20

30

31 33

8

1

12

13

21

Arlington

4

B

C

D

E

F

A

Rim Edge

1999

Eagles

off rim rim edge on rim0.0

0.2

0.4

0.6

0.8

1.0

Use w/i rsa Us

FOOTE CREEK RIM WYOMING

Avoid Saddles

Avoid Windward Side of Ridges

STATELINE WIND PROJECT

Collision Risk Modeling• Approach used

when:– No empirical

fatality data– Sensitive species– To try to estimate

avoidanceRotor Zone

Strike Zone

Over-flight

Fly-thruFatality Risk

Collision Model• Individual-based model to determine bird/turbine collision

probability• Model features

– Turbine design: static & dynamic characteristics– Wind farm layout: number & spatial arrangement of turbines– Bird characteristicso Body sizeo Flight altitude, speed, & directiono Avoidance behavior– Local wind conditions: speed & direction– Incorporates Tucker (1996) model for collision with blades– Accounts for collision with tower & nacelle

Collision Model

• Simulates individual flights through the wind farm

• Collision probability = proportion of flights that end in collision

Rotor Zone

Strike Zone

Over-flight

Fly-thru

Fatality Risk

Wind Park Avoidance

Model InputsBird Flight Characteristics

0 50 100 150 200 250 300 3500

5

10

15

20

25

Per

cent

age

of O

bser

vatio

ns

Height (m)

2

4

6

8

1030

210

60

240

90270

120

300

150

330

180

0

Turbine Characteristics

0 5 10 15 20 25 300

2

4

6

8

10

12

14

16

18

20

Wind Speed (m/s)

Rot

or R

PM

Tucker Model – Collision Probability

Example OutputInputs Results

Avoidance Probability of Collision FatalitiesTower, Nacelle Tucker Flight

Height Total Rotor Tower Nacelle Year 20 Years

0.75N

Low 0.0001571 0.0001303 0.0000142 0.0000127 0.122 2.447High 0.0000259 0.0000230 0.0000015 0.0000014 0.020 0.409

YLow 0.0000545 0.0000273 0.0000140 0.0000132 0.044 0.871High 0.0000079 0.0000052 0.0000013 0.0000014 0.006 0.128

0.90N

Low 0.0001423 0.0001306 0.0000062 0.0000056 0.112 2.249High 0.0000244 0.0000233 0.0000006 0.0000006 0.019 0.384

YLow 0.0000393 0.0000286 0.0000055 0.0000052 0.032 0.636High 0.0000063 0.0000051 0.0000006 0.0000006 0.005 0.103

0.99N

Low 0.0001300 0.0001292 0.0000003 0.0000005 0.103 2.067High 0.0000244 0.0000243 0.0000000 0.0000001 0.019 0.383

YLow 0.0000270 0.0000260 0.0000005 0.0000005 0.022 0.445High 0.0000053 0.0000052 0.0000000 0.0000001 0.004 0.085

Relative Risk of Collision

Among Structure

Types

Major Assumptions: (1) Equal Avoidance Of Turbine And Guyed Structure, (2)Flight Perpendicular To Swept Area And 2 Directions Of Wires

TURBINE 65 M RD92 M MAX HEIGHT

COMM TOWER105 M HEIGHT1.25 MILES OF WIRE

BIRD WITH A 1-FT WINGSPAN HAS 3 TIMES THE LIKELIHOOD OF COLLISION WITH THE GUYED STRUCTURE THAN THETURBINE

Met Towers

• No fatalities observed at the permanent met tower at Nine Canyon

3 TIMES HIGHER MORTALITY AT GUYED TOWER COMPARED TO TURBINE AT FOOTE CREEK RIMONLY 5 GUYED TOWERS

Gehring 2004 – guyed comm. Tower 10 times higher mortality than unguyed towers

Illustration of accumulation of fatalities under future wind development scenario in the

Columbia Basin

• Fatality estimates from 7 studies in the region• Population size from BBS data and PIF modeling• Background mortality from literature• Population effect?

Raptor Nesting Impacts

• Short-comings – low sample size for

individual projects– Conditioned on

distribution of nests at a particular site

– Short-term vs long-term

Raptor Nesting Impacts• No suggestion of an effect

– Montezuma Hills, CA (Howell and Noone 1992)– Foote Creek Rim, WY (Johnson et al. 2000)– Altamont, CA (Hunt and Hunt 2006)– Nine Canyon, OR (Johnson et al. 2003)– Stateline, OR/WA

• Possible Effect– Buffalo Ridge, MI (Usgaard et al. 1997)– 261 Km2 surrounding facility density of 5.94/100 Km2

– No nests present in the 32 Km2 facility, 2 predicted• Caution: all studies relatively short-term

Radio/GPS-telemetryJ. Watson, WDFW

Habitat ImpactsHabitat Impacts

Strickland and Morrison

Potential Sources of Habitat Impact

• Direct loss of habitat• Indirect loss of habitat from

behavioral response to wind plant facilities

• Long-term impacts• Short-term impacts

Predicted Impacts Due to Habitat Disturbance from planning documents (e.g. EIS)

• Temporary (construction) impacts from roads, pads, substation, etc. (estimated)– 0.4 to 3 acres/turbine

• Permanent (operations) impacts (estimated)– 0.7 to 1 acres/turbine

• Permanent footprint 5-10% of site (BLM 2005)

• Mitigation – can be clearly defined

• Game species (e.g., elk pronghorn, mule deer, sage grouse, prairie chickens)

• State and Federal T&E (e.g., Washington ground squirrel)

• Other Protected species (e.g., golden eagle nesting, prey habitat)

Habitat Loss (direct and indirect)

• Grassland Songbird Displacement Studies– Buffalo Ridge, MN (Leddy 1999, Johnson et al. 2000):

Small scale displacement (~80-100m)– South Dakota: 1 of 3 species (grasshopper sparrow)

showed reduced density within 150m – (Schaffer and Johnson 2008) – Stateline: Grasshopper sparrow showed displacement

effect within 50m– Oklahoma: No displacement for grassland species as a

group (O’Connell and Piorkowski 2006) – Minnesota – no apparent impact to Greater Prairie

Chickens• Ongoing and potential studies of bird displacement

(e.g., Stateline, N & S Dakota, and prairie chicken in Kansas)

Completed and Ongoing Displacement Studies

Habitat/Displacement Impacts for other bird species

• Mountain Plover – Foote Creek Rim, WY declined at wind plant, a reference area, and regionally

• Canada geese – Top of Iowa no displacement in corn fields• Migrating raptors – Canada study suggests no avoidance

by raptors during migration with one possible exception (NWCC 2008) – low power.

• Europe– Some species unaffected while certain waterfowl, shorebirds, and

songbirds avoid turbines (e.g., European golden plover, northernlapwing, Eurasian curlews)

– Pink-footed goose displaced up to 600m

Radar Tracks of Migrating Birds through Nysted Offshore Windfarm for Operation in 2003

Response distance:day = c. 3000mnight = c. 1000m

Avoidance Behavior can be Significant

WEST, Inc.

Summary• No big surprises in cases where mortality data

collected in the area/region• We know a great deal about rates in some

landscapes (e.g., Ag) and some regions (e.g., PNW)

• Lack of replication of fatality studies, particularly in the northeast and south-central (ongoing studies will be valuable additions)

• Limited studies with pre-construction diurnal avian use data and post-construction fatality limit predictive ability– Raptor use and mortality appear to be related

Summary

• With the possible exception of Altamont, avian risk from individual wind projects appears to be a risk to individual birds and not populations

• Wind turbine lights not currently documented as a significant attractant for birds

• Micro and macro siting best ways at reducing mortality

• Unguyed met towers• Indirect habitat loss and cumulative

Impacts of most concern

WEST, Inc.

Priority Research NeedsPriority Research Needs• Better synthesis of existing information• Fatalities and habitat-related impacts in

unstudied and new locations and unstudied species are needed

• Raptor nest impacts, sensitive species impact understanding (e.g., ferruginous hawk)

• Cumulative impacts• Models for prediction of impacts and risk• Determine mitigation effectiveness

Questions?