WMOD00380

• Title: One Viewpoint Concerning Unit Sizes in the Development of Wind Turbines
• Date: 11/1/1981
• Weather Modification - Doc Type: Report

What is reasonable or unreasonable depends greatly, of
course, on estimates of cost-effectiveness and on how near or
how far we may be from the break-even point when wind-
power production will become profitable. Todd et al. [8],
have deduced that the combination of large windfarms sited
in especially windy places in the western United States with
hydroelectric pumped storage and long-distance transmission
of power to load centers would be cost-effective now in
comparison with new coal or nuclear-fueled base-load power
plants provided the windpowered generators now planned for
prototype production and testing before 1984 attain the cost-
of-electricity goals of 3-3.7 cents per kWh3 set for them by
the Department of Energy. If this situation should be verified,
it is possible that even the most sanguine estimates of wind-
power capability might be eclipsed, An issue paper prepared
for the Department of Energy [9] on the subject of an ac-
celerated program of wind power identified national goals for
the year 2000 of installed capacities ranging from 16 GW for
"business as usual" to 150 GW for an accelerated program.
The report on inexhaustible ene1rgy resources prepared for the
Department of Energy [10] projected 60 GW of installed
capacity as an "optimistic realistic" estimate for the year
2000 on the way to an 80 GW ceiling by the year 2010, with a
"realistic maximum" of 150 GW. In an assessment of en-
vironmental factors of wind energy conversion [11], The
Energy Research and Development Administration projected
consumption of electrical power from utility networks in the
year 2000 at 2400 GW, of which about 7 percent or 170 GW
might be the average power generated by wind with an ac-
celerated program of wind power development. At a capacity
factor of 0.34, this would correspond to 500 GW of installed
windpower capacity.
In view of this indefiniteness, the following calculations
embrace a range of goals from 12-1200 GW of installed
wind power capacity. This range is sufficiently broad to test
the sensitivity of energy cost to size of the goal from the trivial
to beyond what is considered likely.
Inseparable from any estimation of the nation's future
windpower goal is how far one must go in the direction of
lower mean windspeeds in ordet to fulfill it. One can estimate
from Figs. 2 and 3 that the cost of wind energy from an 8-m/s
site is almost double that from'a 1O.7-m/s site, and that the
optimum machine size for the former is about half that for the
latter. Presumably, the available sites with the highest wind
energy potential will be occupied first, going to progressively
lower winds peed sites until the goal is fulfilled. For example,
the potential from a 1O.7-m/s site is twice that from a 8-m/s
site, as shown in Figs. 2 and 3.
How rapidly the total windpower harvest would increase
with lower winds peed sites is suggested by Fig. 4, which shows
one estimate of how windpower potential and exposed land
areas increase with decreasing windspeed. If a given
proportion of the potential I~nd area in each winds peed
category were utilized for wind(arms, one might conclude by
comparing the respective cumulative capacities show on the
bottom scale of Fig. 4 that, of all the power harvested from
wind farms exceeding 8 m/s average winds peed , less than 1
percent of the power (and less than 1/2 percent of the op-
timum-size machines indicated by Fig. 2) would be in wind-
farms exceeding 10.7 m/s. If even lower windspeed sites came
into use - say down to 6.7 m/s - these trends might be
further accentuated.
For the purposes of the present investigation, the
calculations for both the 8 and 1O.7-m/s windspeeds were
performed.

~81 dollars, in a 7,2-m/s wind regime. Material presented at public
meetings on the Wind Energy Systems Programs Plan from U .S, Department of
Energy Docket Number CAS-RM-8l-404,

308/VoI.103, NOVEMBER 1981

CUMULATIVE AREA WITHIN 17 WESTERN STATES, km2
I 10 1<1 103 104 10~ 106
I I I I I I I

1
I

CUMULATIVE FRACTION OF AREA
10-6 10-5 10-4 10-3 10-2 0,050,1 0,3 0.5

1500

1000
'l' 800
E
~ 600
>
~
en 400
z
UJ
C
Q:
~ 200
~

~

~"'c'I.
:q(q",
~-94G
~lt .
''''0
s...~
~04}'
4",~
"'Clt,~}'
~-9
..,.~,
GIf}'
, '1>'"

UJ
Cl
~ 100
UJ
~

0,01 QI 10 100 300 600
CUMULATIVE PEAK POWER CAPACITY, GW (electrical)
Fig.4 Cumulative distribution of potential windpower as a function of
cumulative area exceeding given minimum windspeed, after Todd et al.
[81. Anemometer heights are from available records by Reed [121.

Estimation of the Experience Rate. The experience rate is of
crucial importance in estimating the cost of fulfilling the
windpower goal. Key questions concern the experience rate
itself and the length of the production run for which it may be
considered valid.
Experience leads to cost reduction in three ways. The first is
the learning curve of personnel engaged in a particular
manufacturing operation: it has a typical value of 70-80
percent of the starting cost for each doubling of production,
but is progressively erased by personnel turnover and design
changes. The second way is by a program of cost reduction
consciously adopted and conscientiously followed by a single
producer for a single product line. The classic example is that
of the Ford Motor Company, which achieved an 85-percent
manufacturing experience rate for its Model T between 1908
and 1926, terminated largely by a change in popular demand
from open to closed cars. This period embraced production of
about 13 million cars and more than 8 doublings of
production from a starting level of 40,000 per year [13].
The third way, yielding an overall experience curve, is
through continual competition, technological development,
and managerial and capital adjustments affecting an entire
industry or group of industries. In a work regarded as
authoritative [14], the Boston Consulting Group examined
experience curves for a variety of industries through 1968 and
derived from them a universal law of cost reduction over
entire industries and families of related products, with most
of the industry-wide experience curves lying between 70 and
85 percent per doubling.
Limitless decrease in cost implied by this simple experience
law poses a paradox. It is resolved when one realizes that
production cannot indefinitely continue to grow ex-
ponentially. When production becomes constant, progress
down the cost curve slows and after a while becomes in-
significant because the doubling period becomes very long.
Furthermore, no industry enjoys eternal life; the experience
curve for the buggy-whip industry came to an end without
necessarily violating the law.

Transactions of the ASM E