Why is Americas Availability Lower than Europes?

Joe Stevens and Keir Harman

Garrad Hassan America, Inc.

AWEA Asset Management Conference

January 2008
Agenda

Introduction
Definition of Availability
Specific Availability Challenges
Availability Analyses
Ramp up of Availability
Summary
Introduction

Wind farm performance is always a hot topic

More so lately due to recent indications of
underperformance
Lenders and equity expressing concerns.
What role does Availability play in this?
Is Availability in the US worse than in Europe?
Perception is that it is...
Definition of Availability
Availability
Time during which the System is available to produce
Availability =
Entire time within a specified period

Machine Availability
Limited to the turbine system only.
This is the focus of the Turbine Warranty multiple definitions.
Includes deduct for items outside the control of the turbine.
Often includes deduct for annual turbine maintenance.

System Availability
Includes Machine Availability.
Includes the balance of plant through to the point of interconnect.
Includes all maintenance and items not under the control of the
turbine.
Availability Challenges
US Market
Parts availability
Competition with growing wind
construction market
Component manufacture is still
growing or dependent upon foreign
supply
Learning curve for each new facility
with small number of temporary
experienced crew
Project Staffing
High level of employee turnover for
operators
Explosion of opportunities for those
with any experience - workers
market
20 to 40 turbines per technician crew
(2 techs)
Relatively few trained technicians
nationwide
European Market
Wind construction market is generally
stable
Component manufacture is generally
well established and regionally
available
Larger experienced staff typically on
site for new facilities
Longer term employment
Stable growth of industry normal
migration of opportunities
10 to 20 turbines per technician crew
(2 techs)
Larger number of trained technicians
available
Availability Challenges
US Market
Manufacturer Technical Support
Technical support can be difficult to
track down if foreign based
US based support may be dependent
upon foreign engineering support =
delay in response
Project Construction
1 to 2 weeks for turbine erection and
commissioning
Shorter reliability runs PTC driven
Miscellaneous
Climate can be extreme high and low
affects performance of crews
Many projects quite remote
Larger projects 100 + turbines
common = difficulty managing issues
European Market
Technical support in same time region
and easier to bring to bear
Technical support is easily referred to
main engineering group = same day
response
2 weeks + for turbine erection and
commissioning
Sometimes 250 to 500 hours and
maintenance prior to reliability testing
Rare for projects to have both high
and low extremes
Large population centers within
nominal distance for most all projects
Smaller projects 10 to 40 turbines
common = fewer issues to manage
 Availability Data Global Market
100.0% 250
Annual moving average Quarterly moving average 240
230
99.0% Monthly moving average Count of wind farm years (2nd y-axis)
220
210
97% 98.0% 200
190
180
97.0%
Wind Farm System Availabilty [%]

170
160

Number of wind farms

96.0% 150
140
130
95.0%
120
110
94.0% 100
90
80
93.0%
70
60
92.0% 50
40
30
91.0%
20
10
90.0% 0
0 1 2 3 4 5 6 7 8 9 10
Years of wind farm operation

www.garradhassan.com/services/amos
 100%
Availability Data US / Europe Compared
US & Canada - Annual moving average
99% Europe - Annual moving average

98%
Wind Farm System Availabilty [%]

97%

96% 97%
95%

94%

93% Decreasing number of wind farms

92% = decreasing statistical significance

91%

90%
0.0 1.0 2.0 3.0 4.0 5.0 6.0
Years of wind farm operation
 Availability Data distribution of results
Majority 97% to 99%
Mean of the data = 96.6%
Includes ramp up data
20%

18%

16%
GH standard availability assumption 97%
14%

12%
% of population

10%

8%
Few <95%
6%

4%

2%

0%

100%
81%

82%

83%

84%

85%

86%

87%

89%

90%

91%

92%

93%

94%

95%

97%

98%

99%
80%

88%

96%
A v e ra g e a n n u al a v ail ab ili ty
 Availability Data Availability vs. wind speed
Cumulative % of time for an 8 m/s site
high wind
low wind Cumulative nominal energy delivered for an 8 m/s site
faults
Average availabilty over all 'curtailed' wind farms
maintenance
Average availabilty over all 'non-curtailed' wind farms

100.0% 100.0%

90.0%

95.0% 80.0%

Cumulative proportion of energy delivered

70.0%
Availability (RTA) [%]

90.0% 60.0%

Time to energy 50.0%

weighting conversion
85.0% factor = 1.0 40.0%

30.0%

80.0% 20.0%

Time to energy
10.0%
weighting conversion
factor = 1.4
75.0% 0.0%
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

w ind speed [m/s]

25 + wind farms
Limitations of Data
The data does have some limitations that must be recognized:
The data set for the US projects is less than 40 projects.
Even fewer US projects with more than 2 years operational data.
Several Owners/Operators are represented, but not all.
Graphs show results only of data made available to be updated
as more project data becomes available.
Much data from < 1 MW turbine wind farms, more so for wind
farms with 5 + years of operation.
No weighting given for wind farm size or wind turbine size.
Where only turbine availability provided, 99.5% adjustment to get
system availability.
Availability ramp up
A base availability ramp up in the energy assessment can be inferred
as follows:

Years of Operation Annual System Availability

Year 1 93.5%
Year 2 95.5%
Year 3 96%
Year 4 and beyond 97% (??)

The availability ramp up in the financial model must be evaluated on a

case by case basis.
Availability Data Results
The data does have some limitations
Conclusions that can be drawn:
Socioeconomic and geographic challenges are different between
US and European markets.
Unable to demonstrate that long term availability is, in fact, less
than long term projections.
Indications are that ramp up periods, as assumed in financial
models, may need to be adjusted.
Energy weighted availability generally in line with run time
availability (not including curtailment).
Conclusions that can not be drawn:
Not clear that the US Market is, in fact, performing materially
worse than the European Market. Additional data must be
obtained.
Not clear what generic long term availability assumptions can be
made. Requires evaluation of more project specific information.
A comprehensive conclusion regarding availability impact on
project underperformance cannot yet be ascertained.
Questions?

Additional Info at:

www.garradhassan.com/services/energyassessment
www.garradhassan.com/services/amos