Wind Energy Conversion Systems PDF

https://ntrs.nasa.gov/search.jsp?
R=19740008644 2019-11-09T13:51:08+00:00Z
NSF/RA/W-73-006
December 1973
WIND ENERGY
CONVERSION SYSTEMS
WORKSHOP PROCEEDINGS
\
. ,
NSF/NASA
JUNE 11-13,1973 WASHINGTON, D.C.
NSF/RA/W-73-006
WIND ENERGY
CONVERSION SYSTEMS
WORKSHOP PROCEEDINGS
WASHINGTON, D.C. JUNE 11-13, 1973
Edited by Joseph M. Savino

Workshop Chairman
NASA-Lewis Research Center
Sponsored by the
National Science Foundation
RANN - Research Applied to National Needs - and the
National Aeronautics and Space Administration
Grant AG-465 December 1973
For sale by the National Technical Information Service
Springfield, Virginia 22151
Price $6. SO
FOREWORD
The wind, until this century, has served mankind faithfully since
the early Chinese and Persian civilizations. It drove his ships and
his windmills which ground the grain, pumped the water, and generated
some of the electricity. It was the discovery of oil, the invention
of the steam and the - internal combustion engines and the large-scale
implementation of central powerplants and transmission systems that
caused a rapid decline in the use of wind as a power source so that
today the wind is used only in remote regions where other sources of
energy are too expensive.
The windmill improved slowly through the centuries and then only
as fast as new technology developed. The greatest improvements in size,
power output, and efficiency occurred in the time from the late 1800*s
and through the 1950Ts when gradually increasing capability in aero-
dynamics, electrical power generation, and structural design appeared.
From 1900 through 1950 there was great interest in many countries such
as Russia, Germany, England, France, Denmark, the USA and others in
developing large-scjale wind-driven generators. Many were built and
successfully operated, but none were cost competitive with the energy
supplied by coal and oil fired steam plants and hydroelectric plants.
It appears that there never was a sustained effort to develop wind
generators that were cost effective. The reason for this might well
have been because fossil fuels were abundant and very cheap. Further-
more, steam plants supplied energy on demand whereas wind generators
could not because of the uncertainties of the wind itself. Thus, it
seems, there was no great motivation for developing large cost competi-
tive wind plants. As a result, all interest in developing large wind
generators diminished rapidly so that by 1970 practically no one was
engaged in a significant sized effort on large wind generators.
In recent years there has been a growing awareness in some quarters

that the bulk of the U.S. future energy needs are not going to be met by
hydroelectric plant and fossil fuel plants alone as has been the case in
the past. The present fuel shortage makes it quite apparent that future
energy needs will have to be supplied from a variety of sources such as
coal, oil, nuclear, geothermal, direct solar, and, of interest here, the
wind.
In 1972 a Solar Energy Panel was organized jointly by NSF and NASA for
the purpose of assessing the potential of solar energy (including wind
energy) as a national energy resource. One of the conclusions drawn
by the Panel was that sufficient energy could be derived from the winds
to supply up to 19 percent of the predicted annual electricity require
ments by the year 2000. The Solar Energy Panel also recommended that
ill
the federal government take a lead role in implementing programs to
develop the economic systems that would utilize the wind as an energy
source.
Since NSF and NASA were both deeply interested in the wind as a
possible source of nonpolluting and inexhaustible energy, it was decided
in March 1973 to hold a workshop as a first step in following up the
work of the NSF/NASA Solar Energy Panel. The purpose of this workshop
was to bring together for the first time in more than a decade all those
persons who were actively interested in wind power and as many of the
pioneers as could be found to try to determine what was the state of the
art of wind energy systems technology and what direction the future
efforts should take. Anyone who had any significant knowledge of or
experience with wind-driven power systems was invited to make an illus-
trated oral presentation. In addition, invitations to participate in
the discussions were sent out on a selective basis to representatives
of the utility industries, government agencies such as the Federal Power
Commission, and the Rural Electrification Administration, industries
that might be involved in the design and production of the systems and
components, and user interests. A total 83 participants attended the
workshop. A list of the participants and their addresses is included at
the end of these proceedings.
Each participant making an oral presentation was requested to sub-

mit a short summary of it plus a few figures for publication in these
proceedings. A few of the summaries submitted turned out to be short
papers which were quite informative. Rather than mail them back to the
authors asking them to condense the paper to a summary, it was arbitrarily
decided to publish them as submitted primarily to save time so that the
proceedings could be published as soon as possible.
After each oral presentation, a question-and-answer period was

allowed between the audience and the speaker. These and the presentations
were recorded verbatim on tape. Since all discussion after the presenta-
tion was impromptu, the transcript of the tape required considerable
editing. For this reason, those persons in the audience who participated
in the discussion are not identified. Rather a format where a Q for
question, A for answer, and comments from the participants was used.
The workshop was structured to have technical sessions exclusively

on the first and second day. On the evening of the second day, the
participants convened in separate groups to discuss, assess the state of
the art of the particular area, and to draw some conclusions as to the
direction future work should take in those areas of potential advancement
or insufficient information. A chairman was selected to head each group
and to draft a summary report for presentation to all participants on the
iv
morning of the third day. On the third day the summary reports were pre-
sented by the chairman of each evening session and a discussion of these
summaries occurred. This was followed by a programmed panel discussion
by the aforementioned representatives of utilities, government agencies,
manufacturing and user industroe s. A question-and-answer period follow-
ed the panel discussion.
In the afternoon of the third day, the NSF/NASA Wind Energy Program
as it stood at that time was presented and discussed after which the
meeting adjourned.
The workshop was proclaimed to be a success almost unanimously by

the attendees. The interest and enthusiasm was high throughout the
entire three days. From it emerged the feeling that wind power was an
inexhaustible nonpoliuting energy source with a potential for once again
supplying an important fraction of mankind's energy needs. Compared to
other sources such as nuclear, fusion, and geothermal, wind energy con-
version systems are inherently simpler and could probably be made to be
as cost effective and reliable as existing fossil and nuclear plants.
The workshop participants believed that the cost of developing wind-
driven power systems should be quite low compared to the costs of
developing nuclear and other advanced systems and that what was needed
was concerted and systematic attack on the economic and technical
capabilities of wind energy systems. It is the intent of the NSF/NASA
five-year wind energy program to provide such a needed effort.
Page intentionally left blank
CONTENTS
Page
FOREWORD iii
INTRODUCTORY REMARKS, Joseph M. Savino . 1
OPENING ADDRESS, Alfred J. Eggers, Jr. . 2
WELCOMING ADDRESS, Seymour C. Himmel 4
SESSION I - IMPORTANT FAST DEVELOPMENTS
SMITH-PUTNAM WIND TURBINE EXPERIMENT, Beauchamp E. Smith 5

MOTION PICTURE HISTORY OF THE ERECTION AND OPERATION OF THE
SMITH-PUTNAM WIND GENERATOR, Carl Wilcbx 8
PERCY THOMAS WIND GENERATOR DESIGNS, Charles W. Lines 11
PAST DEVELOPMENTS OF LARGE WIND GENERATORS IN EUROPE,

Ulrich Hutter 19
INTRODUCTION TO VOIGT1 S WIND POWER PLANT, Joseph Tompkin . .<- 23
Guest Speaker: WHERE THERE IS A WIND, THERE IS A WAY,

Congressman Charles A. Mosher 27
SESSION II - WIND CHARACTERISTICS AND SITING PROBLEMS
NEED FOR A REGIONAL WIND SURVEY, Vaughn Nelson and Earl Gilmore ... 33
WIND POWER DEMONSTRATION AND SITING PROBLEMS, Karl H. Bergey ..... 41
SURFACE WIND CHARACTERISTICS OF SOME ALEUTIAN ISLANDS,

Tunis Wentink, Jr 46
WIND POWER RESEARCH AT OREGON STATE UNIVERSITY, E. Wendell Hewson . . 53
WIND DATA FOR WIND DRIVEN PLANT, Arthur H. Stodhart 62
SESSION III - ROTOR CHARACTERISTICS
AN INTRODUCTION TO THE PRINCETON SAILWING WINDMILL,

T. E. Sweeney and W. B. Nixon 70
THE USE OF PAPER HONEYCOMB FOR PROTOTYPE BLADE CONSTRUCTION FOR

SMALL TO MEDIUM-SIZED WIND DRIVEN GENERATORS, Hans Meyer 73
vii
Page
THE SAIL WING WINDMILL AND ITS ADAPTATION FOR USE IN RURAL INDIA,
Marcus M. Sherman 75
ROTOR DYNAMIC CONSIDERATIONS FOR LARGE WIND POWER GENERATOR

SYSTEMS, Robert A. Ormiston 80
THE EFFECT OF AERODYNAMIC PARAMETERS ON POWER OUTPUT OF WINDMILLS,,

W. Wiesner 89
VERTICAL AXIS WIND ROTORS - STATUS AND POTENTIAL, W. Vance . . . . . . 96
ADVANTAGES OF THE DIFFUSER-AUGMENTED WIND TURBINE,

R. A. Oman and K. M. Foreman 103
BUCKET ROTOR WIND-DRIVEN GENERATOR,

Howard H. Chang and Horace McCracken 107
SESSION IV - ENERGY CONVERSION SYSTEMS
WIND-POWERED ASYNCHRONOUS AC/DC/AC CONVERTER SYSTEM, D. K. Reitan . . 109
AN ELECTRICAL GENERATOR WITH A VARIABLE SPEED INPUT - CONSTANT

FREQUENCY OUTPUT, H. Jack Allison 115
VOIGHT VARIABLE SPEED DRIVE, Joseph Tompkin 121
SESSION V - ENERGY STORAGE
ENERGY STORAGE USING HIGH-PRESSURE ELECTROLYSIS AND METHODS FOR

RECONVERSION, William L. Hughes 123
USE OF HYDROGEN AND HYDROGEN-RICH COMPONENTS AS A MEANS OF STORING

AND TRANSPORTING ENERGY, Walter Hausz 130
STATUS AND APPLICABILITY OF SOLID POLYMER ELECTROLYTE TECHNOLOGY TO

ELECTROLYTIC HYDROGEN AND OXYGEN PRODUCTION, W. A. Titterington . . 135
SUPERFLYWHEEL ENERGY STORAGE SYSTEM, David W. Rabenhorst 137
BATTERIES FOR STORAGE OF WIND-GENERATED ENERGY, Harvey J. Schwartz . . 146
ENERGY STORAGE BY COMPRESSED AIR, George C. Szego 152
SESSION VI - SMALL WIND POWER SYSTEMS FOR REMOTE AND INDIVIDUAL APPLICATIONS
EXPERIENCE WITH JACOBS WIND-DRIVEN ELECTRIC GENERATING

PLANT, 1931-1957, Marcellus L. Jacobs .... 155
REVIEW OF THE WINDPOWER ACTIVITIES AT THE BRACE RESEARCH INSTITUTE,

T. A. Lawand 159
viii
Page
WIND POWER SYSTEMS FOR INDIVIDUAL APPLICATIONS, Henry M. dews . . . . 165
ECONOMIC CONSIDERATIONS OF UTILIZING SMALL WIND GENERATORS,

Robert Dodge 170
WIND UTILIZATION IN REMOTE REGIONS - AN ECONOMIC STUDY,

James H. VanSant 174
.SESSION VII - WIND POWER SYSTEMS FOR LARGE-SCALE APPLICATIONS
TECHNICAL FEASIBILITY STUDY FOR THE DEVELOPMENT OF A LARGE CAPACITY

WIND POWERED ELECTRICAL GENERATING SYSTEM, Ralph E. Powe 177
THE OREGON STATE UNIVERSITY WIND STUDIES, Robert E. Wilson 180
FRENCH WIND GENERATOR SYSTEMS, John M. Noel 186
A PROPOSED NATIONAL WIND POWER R&D PROGRAM, William Heronemus .... 197
SESSION VIII - TOWER STRUCTURES
A COMMENT ON TOWERS FOR WINDMILLS, H. P. Budgen 204
SOME EXTEMPORANEOUS COMMENTS ON OUR EXPERIENCES WITH TOWERS FOR

WIND GENERATORS, Ulrich Hutter 206
WIND MACHINES, P. B. S. Lissaman 208
COMMITTEE REPORTS
WIND CHARACTERISTICS AND SUING 209
ROTOR CHARACTERISTICS 213
ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS 218
APPLICATIONS 222
PANEL DISCUSSION 225
NSF PRESENTATION, Frederick H. Morse 240
NASA PRESENTATION, Ronald L. Thomas 244
ATTENDEES 254
ix
INTRODUCTORY REMARKS
Joseph M. Savino
Lewis Research Center

Cleveland, Ohio
Our nation is beginning to experience energy shortages. It has
been predicted that oil and natural gas are going to be much less plen-
tiful in the future. Coal is plentiful, but the energy derived from it
is going to be more expensive. Nuclear energy may supply a significant
portion of our energy requirements, but it will not fill the gap com-
pletely. Clearly, the energy needs of the future are going to be met
by a variety of sources instead of the few sources we have right now.
Alternatives to coal, oil, and gas must be developed. Wind is one
alternative which is inexhaustible and nonpolluting. The wind is within
easy reach; we do not have to spoil the land digging for it, nor are
there any undesirable waste products. But to make wind energy practical
requires a sustained effort.
The National Science Foundation and the National Aeronautics and
Space Administration have begun a cooperative program to support such
an effort. This workshop is but the first step in the program.
The purposes for holding this workshop are several:
(1) We want to bring together those people and organizations who
are interested in wind-derived energy.
(2) We want to promote an exchange of ideas and information among
the individuals present.
(3) We want to determine the present state-of-the-art and the pos-
sible direction of future efforts.
The agenda reflects our best attempt to identify all the persons
who are presently conducting investigations in some phase of wind ener-
gy, and to identify some of the pioneers who made important contribu-
tions to wind turbine technology.
OPENING ADDRESS
Alfred J. Eggers, Jr.
Assistant Director for Research Applications
Washington, D. C.
It certainly is a pleasure to welcome all of you to this first

workshop on Wind Energy Conversion Systems.'
This is the fourth of the seven major areas of the National Science
Foundation's Solar Program in which we have held such sessions as this.
The other areas include solar thermal conversion and photovoltaic con-
version and their applications to, for example, the heating and cooling
of buildings.
We are particularly happy that the National Aeronautics and Space

Administration is cosponsoring this workshop in the important area of
wind energy conversion. The purpose of all of these workshops is to
bring together people in the various disciplines and with various
points of view that are vital to the interchange we must have to ensure
us that the solar energy program is moving forward in the right direc-
tions.
It's particularly important, too, to do so in the wind energy con-

version, which, as you well know, is an ancient technology that has es-
sentially lain fallow for more than a generation, at least in this
country.
The solar energy program, more generally, is a major focus within

the National Science Foundation's program on Research Applied to Na-
tional Needs, or RANN, as it is sometimes referred to. The purpose of
RANN is to focus research on problems of national importance with the
objective of contributing to their practical solution. Our responsi-
bilities are far broader, however, than the physical sciences and the
technological challenge. We must point each program along lines that
relate to the needs of the ultimate user, whether the user be in agencies
of the Federal, or State, or local governments or the private sector,
including especially, of course, industry. Consequently, we must devote
full attention to user needs and such factors as environmental aspects,
the esthetics, regulatory matters, economic, and social concerns. We
need to consider other matters required to bring a technical idea
through all the phases of study, research, proof-of-concept experiments,
transfer to users, aggregation of investment capital, and actual ap-
proval by responsible government agencies.
Fortunately, we do not have to solve all these problems today as we

launch this research effort to apply technology in the area of wind
energy conversion, but we do need to keep them in mind. Our attention
should be focused on those technological solutions that seem to us to be
most likely to survive this process and actually deliver megawatts on
the line when the process is complete. That is what I mean when I use
the term "systems analysis." Let us look at all elements in the system,
including, for example, the unions. We should remember that our nation's
energy problems are very severe and that the need for clean, new energy
sources will probably increase as we move through the present decade,
into the 1980's, and well beyond.
No single solution is likely to provide the whole answer to those

problems. Different regions of the country will have the opportunity to
use regional resources to meet regional needs. This fact is becoming
clear in other parts of the RANN program. One example is geothermal.
energy. Geothermal resources are available in certain regions and con-
tribute to the needs of those regions. The same is true of wind energy,
as well as, perhaps, other forms of energy. In view of the stringency
of our national energy problems, we should not fail to cover any such
bets.
As you know, the joint National Science Foundation and the NASA
Solar Energy Panel reported on the potential of wind energy in its re-
port published last winter, under the auspices of the Office of Science
and Technology. Some of the numbers mentioned in that report are quite
impressive: More than 300 billion kilowatt hours a year from off the
shore of New England; 180 billion off the mid-Atlantic seaboard; 210
billion in the Great Plains; 190 billion off the Texas Gulfcoast; and
perhaps HOO billion along the Aleutian chain.
As you also know, a wind machine built in 1940 at Grandpa's Knob,

Vermont, generated about 1% megawatts of electricity but had to be shut
down because of the fatigue failure of one of the blades. We've come a
long way since that time in our materials and structural design capa-
bility. I am sure we have methods available today that would have
solved those problems. So there is a clear opportunity to pick up where
the technology stopped and move forward with deliberate haste. Let's
concentrate on keeping the capital cost down. We will also need eco-
nomical systems for energy storage to provide energy when the wind is
not blowing.
I was pleased indeed to observe the breadth of the technical issues

you have highlighted in this conference concerning wind characteristics
and sighting problems, rotors, conversion systems, energy storage,,
small systems, large systems, and tower structures. Now, when you are
through considering all these technical matters, please remember to
return to the basic reasons why we're here. The name of the game is
utilization. This is not just another opportunity to do our technical
thing. Think megawatts on line and all the things that have to be done
before they will be on line. In other words, think total systems.
We are very pleased to be moving ahead in this program with NASA

and taking advantage of their know-how in aerodynamics and structures.
WELCOMING ADDRESS
Seymour C. Himmel*
Deputy Associate Administrator for Technology
Washington, D. C. 2054-6
Gentlemen, I won't take up very much of your time other than to

extend a welcome from NASA to you in this vital activity.
We, of course, find that, when one addresses the energy problem,
a lot of. the technologies that we deal in in the National Aeronautics
and Space Administration are involved with and applicable to the solu-
tion of the energy problem. In particular, we feel that our work in
aerodynamics, materials, power conversion systems, and the like can
play a role in the development of practical wind energy systems.
From our agency's viewpoint there are no technological barriers

to overcome. There are no breakthroughs that have to be made other
than a very vital one: the economics of wind energy systems. The prob-
lem must be approached in a systems engineering fashion, which I always
like to describe as "solving your problems subject to a number of
boundary conditions.n
We are very pleased to be collaborating with the National Science

Foundation in attacking this problem, and I believe this represents the
first formal exercise between the two agencies, bringing in the outside
world, so to speak, in addressing this particular issue. We would like
to contribute to the solution of the nation's problems, in this case
the energy problem, and we feel we have the people and the facilities to
assist in doing this. By collaborating with the National Science Founda-
tion and gentlemen like you from industry and universities we feel we
can make notable advances and rapidly, as we define the problem, try out
solutions, and ultimately solve it.
We know that there is lots of wind available. The question is how

to harness it economically.
Again, I would like to welcome you and look forward to a most

productive session and a most successful and aggressive program.
*Now Director of Aeronautics, Lewis Research Center, Cleveland, Ohio

SMITH-PUTNAM WIND TURBINE EXPERIMENT
Beauchamp E. Smith
Retired President #
S. Morgan Smith Company
York, Pennsylvania
I am honored to speak to you about the Smith-Putnam wind turbine
experiment since it was carried out 30 years ago. Also, I actually had
very little direct participation in the project other than to see that
the bills were paid - which, incidentally, at times presented some real
problems.
I assume that many of you here have read Palmer PutnamTs book
"Power from the Wind". It was written at my request to try to summarize
the story of our efforts to test the feasibility of large scale utiliza-
tion of the natural energy available in wind for the production of elec-
tricity in commercial quantities and to learn something of the economics
involved.
Putnam was introduced to us through Mr. Tom Knight, then commercial
vice-president of General Electric in Boston, in the early fall of 1939.
Mr. Knight and Howard Mayo, the long-time manager of our Boston office,
had often worked together on hydroelectric projects throughout New
England. They were both concerned about the dwindling market for hydro
because most of the commercially feasible sites had already been de-
veloped. Our organization was already deeply involved in the promotion
of pump storage developments. It seemed to us that wind power in com-
bination with pump storage would be a natural partnership. And, if it
could be proven technically and economically sound, it would give us
both a new product and an expanded market for our existing lines of
hydraulic turbines and pump turbines, which in 1939 supplied the life-
blood for our company.
We fully realized that in undertaking the project we were taking a
real risk. We were then a comparatively small family-owned company so
it was easy to get a quick decision. Our board of directors voted to
take the gamble and the project was born.
Putnam had already done much preliminary work and had gathered to-
gether a very knowledgeable group who had been assisting him on a part-
time basis as their free time permitted. Having secured a sponsor,
Presentlv A11is-Chalmers.
Putnam undertook the task of organizing the group of eminent scientists
and engineers who would be responsible for the selection of the site
and the design of the prototype test unit. Dr. J. B. Wilbur of Massa-
chusetts Institute of Technology served as chief engineer of the project
in collaboration with George A. Jessop, chief engineer of our company.
Unfortunately, neither Putnam nor Wilbur, both of whom are still active,
are able to be here today because of other long-standing prior commit-
ments.
"Power from the Wind" sets forth the story of the project very com-
pletely, so I will not attempt to elaborate except to explain why so
much of the fabrication of the unit had to be farmed out. World War II
was already being fought in Europe and our manufacturing facilities in
York were completely loaded with orders both for our regular products
and also for various military items. So, when the time came to actually
begin the manufacture of the experimental unit, we had no capacity avail-
able in our own facilities, and all components had to be farmed out for
manufacture by other companies in order to meet the delivery requirements.
While model tests, design, and site selection were underway in the
early months of 1940, it became fairly obvious that the United States
would eventually become embroiled in the war. In this event materials
and manufacturing sources would probably no longer be available for a
project such as ours with little or no chance of being assigned a pri-
ority rating. This prospect necessitated making design decisions so
that orders could be placed for forgings and other critical items in
short supply long before final studies were completed. Unquestionably
the calculated risks involved, forced upon us by our timing, ultimately
contributed to some of our structural and mechanical problems and to the
final failure of one of the blades which brought our test program to an
untimely end.
Considering the very difficult working conditions prevailing both

in factories and the field, I feel that our team turned in a most re-
markable performance in producing and installing such a complex mecha-
nism in a very limited time. The blades of the test unit were rotated
by the wind for the first time on August 29, 1941, just 23 months from
the time of our first conversation with Putnam and Knight in Boston in
1939.
PutnamTs book sets forth the many problems encountered during the
tests, and how they were at least partially solved. Electricity was
generated in commercial quantity and delivered to a utility transmission
network, the first synchronous generation of power from the wind, when
the unit was phased-in to the lines of the Central Vermont Service
Corporation at 6:56 P.M. on October 19, 1941.
The project thus proved by actual demonstration the feasibility of

generating electricity in useful quantity from the wind. What it did
not prove is that this can be done on an economically feasible basis!
Putnam, in his book, sets forth the trials and tribulations, delays
because of component failures and difficulties in securing replacements,
and all the other problems that plagued the experiment until the blade
failure, which occurred in a known weak spot, at 3:10 A.M. on March 26,
After this blade failure, S. Morgan Smith Company, with its lim-
ited financial resources, reluctantly made the decision that it could
no longer continue to finance the project. The test unit was disman-
tled and removed from the site, the patents and patent applications
were dedicated to the public domain, and the investment was written off
to experience.
"Power from the Wind" was published in 19 U8 as pur final contri-

bution to the public. I and, I am sure, all of the many dedicated
individuals who labored so hard and contributed so much to the experi-
ment found it a fascinating episode in their lives. And, I believe,
most of us still harbor the hope that some day, somehow, someone will
revive interest in carrying on further research and experimentation in
this field.
With shortages of power developing all over the world, with the
growing realization that the worldrs fuel reserves are not inexhaust-
ible, and with the knowledge that our present known methods of using
our dwindling fuel reserves are damaging our environment, I believe the
time has come for another close and hard look at wind power as at least
a partial solution to some of these problems.
I am delighted that this workshop has been convened and I am very

hopeful that you learned gentlemen gathered together here in Washington
will produce some novel and useful approaches which will lead to fur-
ther research and experimentation and to a solution which will make
power from the wind a practical and useful energy source for mankind.
MOTION PICTURE HISTORY OF THE ERECTION AND
OPERATION OF THE SMITH-PUTNAM WIND GENERATOR
Carl Wilcox
Allis-Chalmers*
York, Pennsylvania
A color movie presented scenes at various stages in the assembly

of the major subsystems of the Smith-Putnam wind generator such as in-
stalling the rotor blades and the rotating platform at the top of the
tower. In addition, scenes are shown of the wind generator in operation.
DISCUSSION
Q: What safety factor did you use in the design of this system?
A: We used the safety factor, I think, of 1%, plus an ignorance factor
of 1% - just like an airplane. You can't afford to have too big a
safety factor. We tried to do most of the erecting work on windless
days during spring. We worked any hour of the day or night if the
wind wasn't blowing.
Q: Why were the rotor blades made of stainless steel instead of alumi-
num?
A: I think the answer is we had the Budd Company build the blades, and
they build out of stainless. The grids were stainless, but the spar
that runs through is not stainless. It was cortane.
Q: How closely did the blades match each other in weight?

A: The blades matched very closely; in fact, they both weighed within a
very few pounds of each other. After the assembly we, of course,
could not weight them on top of the hill, and they did not seem to be
out of balance at all. We made some balance checks later on, and
there was no problem.
Q: Did you ever build a scale prototype?

A: No.
Q: What was the length of the chord of the blade?

A: 11 feet, U inches.
Q: Was there any twist to the blade?

A: Yes. The blade was twisted about 5° in three sections, being
straight between each one.
Formerly the S. Morgan Smith Company.

Q: How long did actual installation take?
A: The tower was started in early 1941. We turned it over in August of
the same year, so the erection was pretty fast.
Q: Did you encounter any high winds when you were installing the blades
to shut down the work?
A: No, we didn't. This was done in July, and there was very little
trouble with wind at that time.
COMMENT 1: Carl, I would like to point out that we had some rather ex-
pert forecasting. It was a great help to us.
COMMENT 2: Yes, we had some MIT's meteorological department forecast the

weather day by day, even hour by hour if we needed it, so that you could
plan on windless periods for erection.
Q: Did-you feather the rotor blades under a high wind or let it rotate?
A: We feathered it under high winds. The scheme of operation was to
set the blades at about 14° with no wind. As the wind velocity in-
creased, the rotational speed increased, and at approximately rated
speed, the blades were rotated to the design angle.
Q: What was the rated speed?

A: The rated speed was 28.7 rpm. And then as the wind velocity picked
up, the generator was put on the line and the power increased until
you got the rated power of your generator. At that time you would
start to pitch the blades or start towards feather to control the
power output at your rated low. Originally, it was planned that at
60 miles an hour we would go to full feather and take the unit off
the lines. However, we found that under certain conditions we ran
in to 70 or 75 mile an hour winds. The reason for cutting it off at
60 was a feeling that your gust energy was too great in those cases
to control, but that did not work out to be the case.
Q: What was the tower natural frequency?

A: I don't know as I can answer that.
Q: How well did you control speed?

A: We controlled speed with a Woodward governor. The generator ran at
600 rpm, and we controlled the speed as a function of generator out-
put, and we have some charts where generator output is very smooth.
It took a lot of adjusting on the governor and so forth to get it
that way.
Q: What was the minimum wind velocity that you could operate at full f
power?
A: The minimum velocity at full power was about 30 miles an hour.
Q: What percentage of the time were you below minimum wind velocity?
A: I think about 30 percent.
Q: How did you check the balance to the blades?
A: We just checked the balance by its effect on the yaw motion of the
housing at the top of the tower and on stress readings, and we found
there was very little difference between the weight of the two of
them. We added some weights in one case on one side but it didn't
seem to make very much difference.
Q: Why did you position the rotor on the downwind side of the tower?
A: There were various reasons; if you make a dynamic study it looks
like that is the place to put them.
Q: What was the icing problem, if any, and what did you do about it?
A: Icing was one of the things that worried us a little, but that's why
the system was down on a lower hill (Grandpa's Knob) where it was.
We found we did not get too much icing especially on the blades.
The blades collect some ice, but during rotation the ice would break
up. The idea, I think, is to make your blades flexible enough to
break ice off them.
Q: What's objectively the character of the vibrations involved here

when the unit was operating? Can you give any G-levels or anything
like that?
A: No. There was motion, of course, from every revolution. There was
motion in the pitching drive all the time.
Q: Do you know what frequencies it was?

A: It was twice per revolution.
Q: Would you be able to summarize the cost of this project?

A: Do you mean total cost? I think I'll let Mr. Smith answer that ques-
tion. I'm sorry, I didn't look it up. The total cost was about a
million, three-quarters of a million, somewhere like that. We had
great cooperation from various component suppliers. The Budd
Company contributed a great deal to the designs; American Bridge
Company built the tower and also handled the actual erection. We
paid the costs. The total cost was just over a million, somewhere
around that.
Q: Do you have a recollection of the total weight of this machine in-

cluding the tower and the blade factor and the lot?
A: The total weight was about 500,000 pounds.
10
PERCY THOMAS WIND GENERATOR DESIGNS
Charles W. Lines
Federal Power Commission

Washington, D. C. 20426
For the benefit of all present, I would like to begin with a gen-
eral description of the organization and responsibilities of the Federal
Power Commission. The Federal Power Commission is a federal regulatory
body, administering the National Gas Act and the Federal Power Act, and
is comprised of five commissioners appointed by the President and con-
firmed by the Senate and supported by a staff, which includes members
of all professions and activities - legal, engineering, economic, and
many others.
As such, the Federal Power Commission has no responsibility,

authority, or funds for any research and development activity. It,
nevertheless, remains interested in all these activities and does con-
tribute indirectly to the advancement of the art and science. Through
the Uniform System of Accounts (the reports made to the Commission
about the cost and expenditures of regulated industry and its actions
through rule-making processes and hearings) the Commission can give en-
couragement to the regulated industry. One such example is allowing
for the cost of research and development activity in the determination
of allowable consumer rates (that is, wholesale rates) so that the costs
may be recovered by the utility which makes the expenditure. The Commis-
sion and its staff is also interested in any development that would af-
fect, of course, the various aspects of reliability, adequacy, economics,
and things of that nature.
But I would emphasize that anything I say that might be construed

as an opinion is mine and not that of the Commission. Any Commission
activities always are a result of hearings in rate, or similar, cases,
or are produced in the form of orders made public and followed by
industry.
Mr. Thomas devoted about 10 years, in addition to his other duties,

to a detailed analysis of wind power electric generation and its effect
on the electric utility industry. He actually produced, and they were
published by the Commission, four monographs, the first one was titled
"Electric Power from the Wind" (March 1945). (Notes on these monographs
appear at the end of this paper.) This first monograph was prompted by
the 1941 to 1945 construction and operation of the 1,250-kilowatt instal-
lation on Grandpa's Knob near Rutland, Vermont, which we have just seen.
This installation was integrated with the Central Vermont Public Service
Corporation, and, as we know, it suffered fatigue in blade failure.
11
As Mr. Smith and Mr. Wilcox have advised, it was not rebuilt because of
economic considerations.
Now, Mr. Thomas envisioned wind power electric generation for use
on interconnected utility networks firmed up by hydroelectric storage
facilities in order to overcome the firm power deficiencies of wind
driven generators. He used to a great extent the economic data from the
GrandpaTs Knob operation, and he concluded that between 5,000 and 10,000
kilowatts were necessary for economic viability.
In the first monograph, he described the twin wheeled, two bladed

propeller design for a 7500-kilowatt unit and a twin wheeled, three
bladed propeller design for a 6500-kilowatt unit. In order to overcome,
in part, the difficulty in coupling a variable speed, wind driven mechan-
ical source to a synchronous speed, alternating current commercial system,
Mr. Thomas* design proposed a wind driven, direct current generator,
electrically coupled to a dc to ac synchronous converter. He calculated
the cost, based on extrapolations and estimates from Grandpa's Knob, to
be $68 per kilowatt capacity for the 7500-kilowatt unit and $75 per
kilowatt capacity for the 6500-kilowatt unit. Mr. Thomas concluded that
with certain assumptions the economics warranted the collection of wind
data in greater detail and specifity than that then accomplished. He
also suggested testing propeller designs in wind tunnels over and above
that done, and, in addition, the necessity of constructing a full-size
7500-kilowatt prototype.
I would like to emphasize that in not only Mr. Thomas' first work
but the three that succeeded it, and probably in all the works and
treatises of people writing in the same area, that there is almost uni-
versal agreement that wind data including duration curves be acquired
over a very wide area.
In March of 1946, the second monograph was published. It is de-

voted primarily to the detailed design features of the twin-wheel 7500-
and 6500-kilowatt wind powered generators he discussed, in general, in
the previous monograph. He commented on the March 1945 shutdown of the
Grandpa's Knob installation and reiterated the desirability of larger
units, between 5,000 and 10,000 kilowatts for utility operation.
In January 1949 the third monograph was published. In it Mr. Thomas

compared the detailed aerodynamic designs of the Grandpa's Knob unit, an
English designed unit, and his design. It's interesting that in this
monograph Mr. Thomas modified, for comparison purposes, his 19 M6 design
of the. 7500-kilowatt unit by increasing the blades from two to three for
each wheel, shortening each blade, and increasing the designed rotational
speed. In all cases the general overall height of Mr. Thomas' design
approximated that at Grandpa's Knob, though it has been postulated and he
mentioned designs that might go up as high as the Eiffel Tower or the
Washington Monument, 500 or 600 feet. In this third monograph Mr. Thomas
again emphasized that, because of intangible characteristics and uncer-
tainties of extrapolation, wind tunnel tests and full scale prototype
12
construction were imperative to efficient design.
In February 1954 the last of Mr. Thomas' works was published. In

this last work Mr. Thomas digressed from the detailed technical design
features that he dealt with in his first three works. This work is
largely general in nature, commenting on utilization or integration of
wind generated electric energy in an electric utility network. He dis-
cussed generally the possible benefits of firm and secondary power, de-
rived from wind power generation, when supported by large interconnected
electric systems. By this time steam electric generating units having
250,000-kilowatt capacities were in operation, and the economic benefits
of economy of scale of these units presented at that time a stiff chal-
lenge to competing electric generation sources. Mr. Thomas moved away
from comparative cost base justification for windpower generation in
favor of more general statements. In other words, windpower electric
generation would be justified if the cost would be no greater than that
then being produced by modern steam plants. So, if I may insert
at this time from my readings of the literature that you can find today,
there are no real good present day costs that one can put a handle on to
make an economic comparison. Mr. Thomas also stated, regarding wind
powered electric generation, and this is to a degree at variance with
his earlier implications, that economies of scale were a questionable
attainment and implied that capacities of 2,000 to 4,000 kilowatts might
yield maximum economic benefits. He had moved away from the 5 to 10,000
kilowatt installations, and there is no reason given why he made his
statements in what he had written. But he did say that the 2,000 to
4,000 kilowatt units might be a maximum size. Again the implication is
that there is a lot of study needed in an economic way.
I believe that no one in this room, or no one with any construction

background whatsoever, would question the technical feasibility of con-
structing a windpowered electric generator. We have actually seen one
work, and the fact that the blade suffered that failure is really no
criticism, since in development activities things of that nature happen
all the time. But even today there is a lack in economic justification
for the use of the wind as a source of energy.
In his previous publications Mr. Thomas made some reference to the

use of windmills to generate mechanical power, not necessarily associated
with conversion to electric energy. In this last work he also commented
on a windpowered waterpump and some general applications, including its
use as a pumping source for hydroelectric pump storage operations.
That, essentially, gentlemen, constitutes the works of Mr. Thomas.

They were rather lengthy, and in deference to the program I have ab-
stracted them briefly and the abstracts are given at the end of this paper.
I might comment from my own reading that, while it is technically

feasible to construct such a plan and to integrate it into a central
station operation, in the economic evaluation of any generating source
two main costs must be borne in mind. One is the cost of the fuel, which
13
is a production cost for the energy so generated, electric energy. In
the case of wind it essentially is nothing; therefore, it enjoys that
advantage. Now, because of the nature of electric energy and its use by
the consumer it does not allow, in general, a storage of that energy
except in an indirect form. It can be stored in hydroelectric storage;
it can be stored in fuels which then are called on in accordance with
the demand of the customer. That is the serious defect in getting a
good, firm power worth to electric generation if it's merely integrated
into an operating system. And it appears, from what I have seen, the
costs are the biggest obstacle to overcome to the adequate use of this
source.
DISCUSSION
Q: You mentioned there were two costs. One thing that I didn't detect
in your comment was the cost of eliminating the traces that had made
the power, or to use the word "pollution." Do you see this becoming
a factor, or do you see a shortage in fuels becoming a factor that
would influence the balance of economic trade -off for this power?
A: We are very aware of the economic cost of environmental controls on
methods of generating electricity. In the end the consumer, the
purchaser of the energy, will be the one who actually determines
what steps are taken in these regards.
We have coal resources that can be exploited for all the foreseeable
energy needs until such times as our breeder nuclear reactor program,
or our fusion program, or anything else comes into being. So what
we are really asking, are we not, is what will we pay in an inter-
mediate time period for the home environmental freedom that wind-
power offers?
Now,even though there is nothing in the sense of air pollution or

water pollution that windpower generation implies, it does require
certain land uses, and there are certain esthetic aspects of mile
after mile of windmills scattered over the landscape. Those are
environmental costs also. But, to answer your question, I think the
consuming public by the cost of what they buy will be the eventual
decider of what environmental degradation they will stand.
There is another thing, too. We are prone to, I think, attribute

too great an advantage to some things and consider that the money
supply is inexhaustible. After all, money is a resource and does
come from production and other efforts, and if we waste money, then
we are implying a resource waste in other areas.
Q: Could you tell me where I could get these four pamphlets?

A: They have long been out of print. I have four copies, which are
actually the file copies. The Office of Information of the Federal
Power Commission can arrange to have them reproduced.
Q: What are your impressions on the subject of storage, not just on the
wind energy system, but the solar energy system?
A: Whenever I look at pumping energy into the electric utilities system,
I relate it really in terms of the proportion of windpower that could
be supplied in proportion to the total power output of the utility.
If the proportion of windpower is relatively small, I really don't
see the need for storage, because you have basically three systems
in the utility where the intermediate and backing systems essentially
can perform the storage function. Obviously, if a large proportion
of wind energy is being pumped into the utility, then is there a
storage problem. I disagree that carte blanche storage is a problem.
It depends on how much of the energy is being supplied.
There is a cyclical value to energy produced in the electric utility

operation; in other words, the cost of energy varies second by second.
The economic computers that load and unload the various components of
generation are so programmed that they evaluate the incremental cost
of energy as it's produced on cyclical basis.
If I might use a rather basic evaluation, if you are generating

under pump-hydro, or even run-of-the-river hydro, that has a nonline
at the time worth of 2 mills, then any energy produced by wind
power will not" have a higher value in the planning phase.
Q: Do you have any information on the attempt to get money to build a

larger prototype in 1951? I have a copy of the Congressional Record
in which the hearing was reported, but I have not been able to get
any information on what happened in that attempt.
A: I have no other information except the hearing record myself.
Q: I was wondering if Mr. Smith had any comments to say in collaboration

with the Federal Government. Were there ever any attempts to work
with Mr. Thomas of the FPC?
A: We had a good many sessions. We came to New York very frequently
and discussed this with Mr. Thomas. The problem Mr. Thomas had was
that his economic evaluations were rather broad.
COMMENT: I am manager of a municipal light plant in Massachusetts.

With regards to the economics and to energy storage, let me say that, at
the time that these wind generators were being made the total demand for
electric energy was quite different than it is today.
For example, total electric living was not being pushed to as great
degree at that time. Today they are pushing a total electric home, as
an example. In the total electric home, your greatest need is for heat,
almost 70 percent.
Therefore, the greatest potential of th e wind generator is to generate

heat directly, not to go to electricity and then make heat. The effi-
ciencies of a direct conversion from mechanical to heat energy, such as
in our community, are greatest when it is the coldest. That's when the
wind is blowing the hardest and all the heat is blown out of the buildings,
15
and therefore we would have a direct need at that time for the greatest
amount of heat and a peak demand for electrical needs.
This appears to have significance. I mean heat storage is becoming a

great thing - using heat exchanges, heat pumps, and storing up by water
in tanks for big commercial buildings - all of this in order to combat
the problems. If you buy power on peak, the cost of power is greater
than off peak, so all these things figure into peaking and storage which
is being substantially pushed already to reduce the demand costs.
The following are abstracts of Percy H. Thomas' four monographs which

were the results of his studies on the potentialities of wind power.
ELECTRIC POWER FROM THE WIND

March 1945
Percy H. Thomas

Washington, D. C.
This monograph by Mr. Thomas on the general subject of power from

the wind was prompted by the 1941-1945 construction and operation of a
1,000-kilowatt installation at Grandpa's Knob near Rutland, Vermont, on
the system of the Central Vermont Public Service Corporation. This unit
suffered a blade failure on March 26, 1945, and was abandoned because of
economic considerations.
Mr. Thomas envisioned wind powered electric generation for use on

interconnected utility networks, firmed up by hydroelectric storage
facilities in order to overcome the firm power deficiencies of wind
driven generators. Using certain economic data from the Grandpa's Knob
operation, the author concluded that units of a size between 5,000 and
10,000 kilowatts were necessary for economic viability. The author de-
scribed a twin-wheeled, two-bladed propeller design for a 7,500-kilowatt
unit, and a twin-wheeled, three-bladed propeller design for a 6,500-kilo-
watt unit. In order to overcome in part the difficulty in coupling a
variable speed, wind-driven mechanical source to a synchronous speed
alternating current commercial system, the design included a wind-driven,
direct-current generator electrically coupled to a dc to ac synchronous
converter. The author's calculated costs were $68 per kilowatt of capac-
ity for the 7,500-kilowatt unit, and $75 per kilowatt for the 6,500-
kilowatt unit.
Mr. Thomas concluded that, with certain assumptions, the economics

warranted the collection of wind data in greater detail and specificity
than that then accomplished, the testing of propeller designs in wind
tunnels, and the construction of a full size (7,500 kW) prototype.
16
THE WIND POWER AEROGENERATOR — TWIN WHEEL TYPE
March 1946
Percy H. Thomas

Washington, D. C.
This monograph is devoted to the detailed design features of the

twin wheel, 7,500- and 6,500-kilowatt wind powered generators discussed
in the previous monograph.
The author commented on the March 1945 shutdown of the 1,000-kilowatt

Grandpa's Knob unit, and he reiterated the desirability of larger units
having capacities between 5,000 and 10,000 kilowatts for utility opera-
tion.
AERODYNAMICS OF THE WIND TURBINE

January 1949
Percy H. Thomas

Washington, D. C.
In this monograph, the author compares the detailed aerodynamic de-

signs of the Grandpa's Knob unit, an English design, with his design.
Mr. Thomas modified his 1946 design of a 7,500-kilowatt unit for this
comparison by increasing the blades from two to three for each wheel,
shortening each blade, and increasing the designed rotational speed.
Mr. Thomas again emphasized that, because of intangible character-

istics and uncertainties of extrapolations, wind tunnel tests and full
scale prototype construction were imperative to fix a design.
17
FITTING WIND POWER TO THE UTILITY NETWORK
February 1954
Percy H. Thomas

Washington, D. C.
This is the last of the four monographs published by the Federal

Power Commission relative to the studies made by Mr. Thomas in the 1944-
1954 period regarding wind powered electric generation.
The author's prior works involved highly technical and specific de-
sign matters, particularly in the field of aerodynamics. This shorter
work was largely general in nature, commenting on the utilization, or
integration, of wind generated electric energy in an electric utility
network. He discussed generally the possible benefits of firm and sec-
ondary power derived from wind powered generation when supported by
large interconnected electric systems. By this time, steam-electric
generating units having 250,000-kilowatt capacities were in operation,
and the economic benefits of economy of scale of these units presented,
at that time, a stiff challenge to competing electric generation sources.
In this monograph, Mr. Thomas moved away from comparative cost based jus-
tification for wind powered generation in favor of more general state-
ments; i.e., justification would be sufficient if steam generation costs
were met. He also stated, regarding wind powered electric generation
and to a degree at variance with earlier implications, that economies of
scale were of questionable attainment, and he implied that units having
2,000- to 4,000-kilowatt capacities might yield maximum economic benefits.
In previous publications, Mr. Thomas had made some references to the

use of wind mills to generate mechanical power, not necessarily associated
with conversion to electric energy. In this work, he also commented on
a wind-powered water pump and some general applications, including its
use as the pumping source for hydroelectric pumped storage operations.
18
PAST DEVELOPMENTS OF LARGE WIND GENERATORS IN EUROPE
Ulrich Hutter
University of Stuttgart
Stuttgart, Germany
This presentation describes the more important large wind-driven

power systems that have been proposed or built in this century in the
various countries of Europe. Some of these are shown in the accompa-
nying figures 1 to 9. The physical size, maximum power output, and
other characteristics of each system were described with the aid of
slides. The most important of the large-size wind-driven plants in
Europe were built in Germany, England, Denmark, France, and Russia.
Also described, in some detail, was the 100-kilowatt wind-driven

generator that was designed and built by the author in cooperation with
the Allgaier-Works of Wurttenburg, West Germany (figure 10). A short
movie was presented to show the 100-kilowatt Hutter-Allgaier machine in
operation. In figure 11 is displayed the measured output in kilowatts
for various wind speeds. For comparison, the data of Andreau-Enfield-
Cables wind generator system are shown.
30 m 0. 100 KW DC RUSSIAN WIND -

TURBINE. YALTA BLACK SEA 1931
Figure 1 Figure 2
19
130 KW. 21 m diameter
ELECTRICITE DC PRANCE »
Figure 3 Figure 4
.j-
24 m 0. 100 KW ENFIELD-ANDREAU WIND-
DRIVEN GENERATOR ST ALBANS 1953
31 m DIAMETER, 800 KVA, NOGENT LeROI,

FRANCE, 1958-1960
Figure 5 Figure 6
20
ameter.
plant of J. Juui
SYDSTSJAELLANDS ELEK- 15 m 0 , 100 KW. JOHN BROWN
TRICITETS AKTIESELSKAB
WIND-TURBINE, ORKNEY, ENGL
Figure 7 Figure 8
10 m DIAMETER, 10 kW, U. HUTTER - ALLGAIER 35 m DIAMETER, 100 kW, U. HUTTER -

DESIGN, WEST GERMANY, 1950-1960 ALLGAIER DESIGN, WEST GERMANY,
1961-1966
Figure 9 Figure 10
21
oi C
J3 O
•M -H
O H
rH
a a
Ol
O
h A
0 E
•M b
m o
h<*
01 h
C 01
oi a
C
01
•3Stog
•H
3 fc
^ co
o eft
0 o
H O
h a h
01 o
•H T3 4J
ID C CO
bO CO h
rH 01
rH T3 C
•a 01 oi
I Ol M
Ol CO TJ
•P C
•P "U -H
3 C S
3 (0
Ol 01
* 01 iH
•P M.Q
(0 co
O 01 I
> T3
•M CO «-l
3 Oi
ft CU -H
°C 30)
O M
H 3
CO O
O
0) i-l
CO
01
4J JC
C 4J
CO
H 4H
22
INTRODUCTION TO VOIGT'S WIND POWER PLANT
Joseph Tompkin
Salem, Oregon
The late Hellmut R. Voigt's manuscript entitled "Von der Windmuhle

zum Winderaftwerk" — ("The Design and Operation of Wind Power Plants")
comprises 230 pages and 270 illustrations. One of the leading authori-
ties that certified the document was Professor Flugel, Hannover, Germany.
This investigation has been examined by experts of leading universities
(found in the bibliography of the text) who all testified to the cor-
rectness of the author's findings of his own wind power plants. Other
engineering authorities who personally certified Hellmut*s document were
Professors Pantell, Witte, and Schleichert.
The Cyclone D-30 is a high-speed three bladed wind turbine (diame-

ter, 30 m) that operates at a height of 50 meters. The blades are rig-
idly connected to the hub and the revolutions of the turbine change
linearly with the wind velocity, maintaining a constant speed ratio of
u (blade tip velocity) to v (wind velocity). This "ideal" wind tur-
bine holds its high efficiency over the full predetermined range of wind
velocities. The three generators installed in the gondola generate either
dc or ac current. In case of dc installation, the turbine turns by means
of a transmission, a set of dynamos. In case of ac installation, the tur-
bine turns by,means of an infinitely variable speed drive (patented), a
set of synchronous three-phase generators. The generator-gondola with its
streamlined shaft can turn around the tapered mast top by means of a
journal bearing and a king pin. This motion is controlled automatically
by two wind rosettes in such a way that the wind turbine always opposes
the wind direction. The mast, a truss and shell construction, pivots
about a central foundation and is held in position by steel cables. The
wind turbine is equipped with an aerodynamic brake system, and the blades
are equipped with an anti-icing system. The low-speed control wind tur-
bine (3-m diameter) located at the tip of the gondola, controls load and
speed of the wind turbine by means of a differential regulator driven by
both the main and the control wind turbine (patented). The total effi-
ciency of the wind power plant is 66 percent.
Based on the wind conditions at Cuxhaven, Germany, the maximum out-

put is 720 kilowatts at a wind velocity of 16 meters per second. The
total installed electrical capacity is 750 kilowatts, and the power out-
put pear year is 2,125,000 kilowatt-hours.
The turn-key installation cost of a wind power plant is not a func-

tion of its installed electrical capacity as in steam or diesel power
plants. The cost is determined by its towering structure, which has to
23
withstand maximum wind gusts of 50 meters per second (115 mph). The
installation cost of a wind power plant should be plotted versus the
wind turbine diameter. Such a curve leads to the conclusion that a plant
with a 30-meter wind turbine diameter represents the optimum in costs.
The price per output rises quite rapidly for smaller wind turbine diam-
eters, as well as for larger ones. In 1951, the turn-key installation
cost of a wind power plant Cyclone D-30 was about $244,000.
The operating cost of a wind power plant is not a function of the
installed capacity but instead depends on the number of wind power plants
operated in a certain area. The more plants that are interconnected to
form a wind power central, the lower the operating cost. For instance,
in my feasibility study at Cascade Locks, Oregon (Bonneville Dam sector)
one unit produced over 14,000,000 kilowatt hours per year for approxi-
mately 68.3 percent of the time per year at about 4.3 mills per kilowatt-
hour. If 20 units were located, the cost is reduced to approximately
0.8 mill per kilowatt-hour.
The cost savings of a nonweather vaning aggregate over a weather
vaning aggregate amounts to $25,000, or about 10 percent of the total
installation costs. The difference in operating costs is almost negli-
gible.
Mechanical coupling of wind turbine and water pump is prohibitive,
due to the tremendous torques developed by a wind turbine of considerable
size.
We have included the production of hydrogen from the electrolysis
of water for producing electricity by the hydrogen fuel cell. The in-
stallation of this system would supply the needed electricity during
peak load demands or augment power during the year, if wind velocity is
below the prescribed velocity (8 mph).
Some technical data of the TORNADO wind power plant DIOO are as
follows:
Diameter of wind turbine, ft 100
Operating height of wind turbine, ft 150
Total height of wind power plant, ft 200
Diameter of generator gondola, ft 10
Diameter of supporting tubular tower, ft 7.5
Height of observation platform, ft 90
Diameter of air rosettes, ft 12
Speed range of wind turbine, rpm 15 to 60
Utilized wind velocity range, mph 9 to 36
Average annual power output:
kW 250
HP 340
Maximum power output:
kW . 700
HP 950
Annual net current output, kWH . . . . . 2,125,000
Normal voltages:
dc 440
ac . . 525
Total efficiency of wind power plant, percent 66
DISCUSSION
Q: What year were those dollars?

A: 1951 dollars.
Q: I didnft quite understand the meaning of the 66 percent efficiency.

A: This is the total overall efficiency, including mechanical.
Q: From the total energy?

A: Yes.
Q: Assuming an ideal windmill is 100 percent?

A: Yes.
Q: Which is an efficiency of about 60 or 70 percent ideally?

A: The ideal is 59.3.
Q: 59.3 times about 60 then.

A: Yes, so it takes it down to about 40.
Q: About 40?
A: The overall power coefficient.
Q: Let's go through that just once more to make sure everybody agrees.
If you have a hundred energy units in the wind, an ideal windmill
will give 59.3 percent. Now you are saying you can expect 66 per-
cent of that 59?
A: True.
Q: Since Dr. Tomkins didn't have any slides, I'm not sure whether he
built one of these or what.
A: No, this was design based on a small unit. This was simply the
design. The two units were built for the 50-foot units, not the 100.
/
Q: I would like to hear some comments on the availability of some of
the older windmills. For instance, if somebody wants a 50-kilbwatt
unit, are any of these available to be rehabilitated? Where can you
buy a 50-kilowatt unit - who deals with used windmills?
A: The unit just discussed was not an actual unit but a prototype. How-
ever, the data that have been compiled were on actual units (12%
feet, 25 feet, and 50 feet). Mr. Voight didn't live to realize a
100 foot unit.
Q: Is it being built now?

A: No. There seems to be a lack of interest as far as funding is
concerned.
25
Q: Are there windmill units that are available as a production kind of
unit or as a used unit?
A-l:The machine just shown in the film (the 100-kW Hutter-Allgaier
machine) can be produced, but it would be much more expensive than
under the conditions we were working on then. We built about 1800
plants of 10-meter plants (10 kW). There is a fine market, like
used cars, for these plants. If anybody has a plant to sell, there
are five buyers. There is a market. Plants of the size I have
shown are available. It's a question of delivery and price.
A-2:I have been doing some research to find out if there are any in pro-
duction. Those I found are small-scale units up to about 5 kilowatts.
There is a company in Switzerland building a 5-kilowatt unit with a
5-meter diameter blade, and they are in regular production. You can
order one and get it in 6 weeks. As far as I know, this 5-kilowatt
unit is the largest unit presently in production in any quantity.
Q: Do you know the cost?

A; The unit itself, the wind plant, and the control system cost about
$1,900. Freight to this country is not a large factor - about $200.
Q: Now we are selling a minimum power. They need 14 knots of wind,

have a diameter of 9.2 meters, and they deliver ac current. They
can be used to either charge batteries or produce wattage. We can
accept orders for 4.21 machines.
Q: This machine is in production, U.I kilowatts, and 9 meters in

diameter?
A; Yes. In 2 weeks I'll have photographs of this machine. Last month
we tested 4.6-kilowatt machines.
A; There is a machine made in Germany which will generate a maximum of

400 watts ac, rectified dc for charging batteries. It is very use-
ful for household purposes. There is another one in Italy which
will generate 1000 watts, very much the same thing, but it's dc.
There is also a machine in production which can generate about 40
horsepower dc in a 30-mile-an-hour wind.
26
Guest Speaker
WHERE THERE IS A WIND, THERE IS A WAY

Honorable Charles A. Mosher
Congressman, Thirteenth District, Ohio
When I received Dr. Savino's invitation to be your speaker, I must
admit I was at first surprised and amazed. My instant reaction was to
ask, "Why me?" I protested that certainly few people in Washington are
less qualified than I to talk about the winds. But on second thought,
now I do understand why I am here today; I stand before you as "Exhibit
A", a living, panting demonstrator of wind power.
Just as a pun is alleged to be the lowest form of humor, similarly

a politician may be the lowest form of an energy machine. For example,
starting from an almost zero knowledge of the subject before us today,
but using my own wind power (lung power) - or hot air, you may say - I
will now turn that nothing into a 15-minute speech. And that, I submit
is a very real, though elementary, form of energy conversion.
Loosely harnassed, a vast amount of that low form of energy con-

version is best known as the Congress. And that is why some folks
irreverently refer to Capitol Hill as "Windy Knoll"; and the Congres-
sional Record, where the amounts and velocity of our energy conversion
are daily logged, is sometimes called "Cave of the Winds".
And, I want you to know that I have still another qualification

which you may not suspect. I'm not kidding, it is literally true that
back in 1906 I was born next door to a windmill factory, and I lived
there until I was 5 years old. That was in a small town out in Northern
Illinois. We lived next door to the Enterprise Windmill Co., and as a
kid, often visiting the friendly farms of that area, I knew well the
clank, clank, clanking sound of the windmills that pumped water on those
farms. One of my own favorite toys in those days was a model windmill
such as salesmen for the Enterprise Co. used in demonstrating their
products to prospective customers.
But historically, it is significant that immediately across the

street from that windmill factory there was located another factory,
one that my own family operated for three generations, where we built
corn shelters, grain elevators, hay presses and loaders, side delivery
rakes, manure spreaders, and, fortunately, gasoline engines. So, the
Enterprise Co. went broke, and we survived, because those gasoline
engines gradually took the place of windmills on most American farms.
And then, in turn, several years later those engines went out of use
27
when the rural electric cooperatives brought cheap, government subsidized
power lines into most of the farms.
Also, when I was a sophmore in high school, the dramatics event of
the year was our staging of an operetta called "Windmills of Holland".
I was in the chorus and much of the time played part of a windmill.
Suddenly, as I was preparing this talk, I remembered a verse from that
operetta. I won't sing it, but I believe it went like this:
"Touch a button, you or me, and then that great electricity
will do the rest, while we with zest will sit. and look our
very best."
The plot of the show, of course, was the demise of the windmills, dis-
placed by electricity.
So-since childhood, I have been somewhat aware of the waxing and
waning of wind power. But for HO years or more I had not given the
subject any serious thought until early last year when I received a
letter from an Ohio constituent who urged that Congress solve our na-
tional "energy crisis" by encouraging the location of huge windmills on
top of ail tall buildings. He suggested that the Empire State and all
other such skyscrapers, each could and should satisfy their own elec-
tricity needs by means of wind power generators.
Frankly, my staff and I assumed that constituent was some sort of
nut. I'm sure my reply to him was little more than a courteous brushoff.
But now that I've met you folks, I feel guilty about what I thought of
him. I promise you, I'm going to dig into our files and resurrect his
letter and take a more serious look at it. I expect to find it in the
file we have labeled "Crank Mail", but maybe now we should file it under
"Ideas Worth Considering".
And I judge that is precisely the significance of this NSF/NASA
Wind Energy Conversion Systems Workshop. You have gathered to take a
new look at some old ideas and technologies, long neglected, ignored,
laughed at, which in the contexts of today and tomorrow begin to look
very promising, and to me certainly very fascinating.
I am confident you are here to usher in a real technology resurrec-
tion, a very much deserved and needed second coming for wind power. In
the perspective of centuries of human history, I suppose this would be
no mere second coming but the umpteenth coming, only the latest of in-
numerable chapters in man's discoveries of how to make good use of the
winds. All of us have been raised on the wise old adage that "It's an
ill wind that blows no man good", and I take that wisdom to mean that
nearly every wind could be put to some good use. So, we count on you
who are here to see to it that that goal is accomplished.
I asked Frank Huddle, Senior Science Specialist at the Library of
Congress, to tell me where, when, and how human beings first learned to
28
control and convert the winds for their own uses. Obviously, no man to-
day knows the sure answers to those questions. But Frank is imaginative,
and he is confident that observant savages very early learned the ways to
harness the winds. Surely, primitive man (°r was it first a woman)
quickly noticed that a cold wind made him feel colder than cold without
wind; so he (and she) retreated into the caves, not only for protection
from animals but perhaps, even more so, from the winds.
And perhaps men first discovered fire by observing forest fires
caused by lightning. But Frank says it is just as likely that he saw
firey particles dropping from branches of trees forcibly rubbed to-
gether in a high wind, and so perhaps the wind taught him to make fire.
Certainly manTs discovery of the sail was one of the most important
technological innovations of all time; it converted his crude raft or
dugout canoe into an ocean-spanning transportation system. Frank suggests
the inspiration for that first sail may have come when primitive man
watched curled leaves being blown across the surface of a pond.
Having harnessed the wind for transportation, it was inevitable
that man should similarly harness it to grind his grains and pump his
water. I'm told that the windmills of Holland enabled the Dutch to re-
claim vast acreages from the sea; and in Yorkshire, England, they also
were used to pump water from the lowlands.
I already have mentioned that era in Midwest America when most
farmers used patent windmills to pump water from wells for themselves,
their livestock, and crops. But then came cheap electricity to do those
chores, and so not many of those old mills are clanking today.
So, always, as civilization advances, technology and economics in-
teract and whatfs new replaces the old. But, often what's new is merely
an updated version of what's old. And that, I repeat, seems to be what
this Wind Workshop is all about.
Now, gentlemen, I'm going to conclude these remarks by tossing at
you a list of 4 or 5 personal opinions which perhaps have some bearing
on your work here. There's a great deal of interest and concern in the
Congress about today's so-called "energy crisis". It's a very popular,
fashionable subject on "Windy Knoll"...lots of speeches, hearings,
studies, reports, etc. But I do not pretend here to speak for the Con-
gress or for any other members. These opinions (hopefully somewhat
provocative) are strictly my own. I will state them with little or no
attempt to explain or defend; they are tossed to you just for what they
may be worth.
FIRST OPINION - I am convinced that we in the U.S.A. should decide
right now, as a matter of national policy, to free ourselves from any
dependence on oil or natural gas; we should completely back away from
those fuels as major energy sources.
Now, I don't pretend to know exactly when that revolutionary change
might be fully accomplished, but I would hope it could be largely under-
way before the end of this century, less than 30 years away.
My present guess as to a time table is that for the short run, the
next 8 to 10 years, we will be forced to scramble in every direction for
our energy using a lot of undesirable expedients, such as unhappily in-
creasing reliance on Mideast oil and temporarily postponing some of the
desirable, stricter environmental standards. But, I emphasize, that
should be a short-term temporary situation.
In the intermediate period, from 1980 to past the end of this cen-
tury, we must encourage an increasing reliance on our still immense coal
resources (by goal gasification and liquefaction) as well as construction
of conventional nuclear fission power plants (with increasing emphasis
on safety and environmental controls) and then the breeder reactors, as
quickly -as they can be proved practical.
But for the long run, a third stage from the year 2000 on, certainly
our energy goals must emphasize thermonuclear fusion, and most important
of all, ultimately a major reliance on solar energy.
And am I not correct that the energy in the winds is in fact a form
of solar energy, a product of solar heat beating down on our earth and
sea surfaces? So, it seems to me entirely reasonable that your big goal
in this important three-day workshop should be a major change in that
timetable I have just outlined. Perhaps a dramatic shortening of the
timetable could be brought about by bringing on line commercially feasi-
ble wind energy conversion systems (and thus, a form of solar energy)
well before the end of this century, long before any of us have thought
possible!
SECOND OPINION - Obviously, the success of that revolutionary
shift to new energy sources can be accomplished, and hastened, only by
means of a massive, diversified, but selective and coordinated, energy
research and development effort, probably including some so-called
"crash" programs. And, of course, that R & D will require federal appro-
priations at levels and at a pace not yet contemplated in any budget
proposals of which I am aware.
Within the next year or two there must develop a concentrated em-
phasis and momentum for energy related research if we are to have any
chance at all of bringing on line in practical, commercial form those
alternative energy sources that will be so necessary by the turn of the
century.
I am guessing there might be general agreement that the prime can-
didates for considerably greater R & D funding should be the following:
(1) Coal stack gas removal,
(2) coal gasification and liquefaction, plus vastly improved
30
techniques for mining safety and environmental protections in mining,
(3) fast breeder reactors, with increased emphasis on alternatives
(gas cooled?) to the currently emphasized liquid metal fast breeders,
(4) long term nuclear waste disposal technology,
(5) thermonuclear fusion,
(6) solar energy,
(7) pollution controls, and
(8) energy conservation technologies, including new concepts in
building construction, more efficient storage and transmission of elec-
tricity, and surely more efficient, economically feasible, productive
systems for recycling wastes.
Much of that R & D effort will be extremely sophisticated - far out
stuff, terribly costly, and at best a big gamble, adventuring into the
unknown. I believe those big investments are necessary, even though they
are a gamble.
But the point I reiterate right now is this: In our fascination
with sophisticated and costly new technologies, we will make a tragic
mistake if we ignore those great opportunities that exist in new uses of
older, familiar and relatively simple technologies...and, of course, by
that I mean it is very important that we adequately fund this fresh,
innovative look at wind power. I repeat the point made early in these
remarks, that innovation more often than not means a new, imaginative
look at old information and old experience.
THIRD OPINION - My third opinion is a quickie, merely to express
my doubt that there exists in Washington today, either in the Congress
or in the Executive Branch, a sufficient understanding or adequate, ef-
fective, decision-making machinery to provide the aggressive leadership
and national policy decisions which are desperately needed in the realms
of science and technology...and especially needed to solve our energy
problems.
I see some hope in our authorization of the Office of Technology
Assessment, as a new staffing arm of the Congress. If and when it is
funded, OTA should provide innovative impetus. I also see hope in the
new presence of Charles DiBona and his energy staff at the White House
level. And I believe the Administration's reorganization proposal makes
good sense, that we create a new umbrella Department of Energy and
Natural Resources. But, as yet, I see no sign that OMB [the Office of
Management and Budget) is likely to approve really adequate R & D fund-
ing in the near future. However, I can assure you that there are at
least a few of us in the Congress who are aware, and pushing for the
level of determination and effort we believe is imperatively needed.
Time itself is a major human resource, a major national resource.
We must use it wisely, effectively, vigorously; we must not fritter it
away, and that is why I am so heartened by your efforts here.
And, of course, the winds are no respecter of national boundaries
31
or national sovereignty. They blow alike on the just and the unjust.
Certainly, in the winds we have a superb opportunity for sharing the
fruits of scientific and engineering effort with all mankind. So, I am
heartened to know that there are representatives here from several other
nations, as well as our own.
I salute you all!
And now in closing I have a slogan to suggest for your workshop.
It results from my very strong feeling that in our national science
policy today we are somewhat lacking in sufficient commitment; we need
a greater sense of purpose and urgency, a sense of the will to overcome
our problems.
All of us know that old saying "Where there is a will there is a
way". So I suggest the guiding motto for this workshop should be a
slight variation on that theme:
"Where there is a WIND, there is a way!"
32
NEED FOR A REGIONAL WIND SURVEY
Vaughn Nelson
West Texas State University
Canyon, Texas
and
Earl Gilmore
Amarillo College
Amarillo, Texas
The economically favorable utilization of wind power on at least a
modest scale will most likely occur in those regions of the U.S. where
the greatest potential exists and where the intermittent nature of the
wind speeds is small. Golding and others (ref. 1) have emphasized the
need for accurate measurements designed specifically for the purpose of
estimating wind energies, but in the U.S. only a small amount of work
has been done. Thomas (ref. 2) indicated that the Southern Great Plains
is a region over which wind speeds are significantly greater than in al-
most any other part of the nation, a fact that is common knowledge to
residents of the area. The area is large, flat, and accessible (both
financially and physically).
The general wind characteristics as indicated by data from the
National Weather Service are as follows:
(1) The average wind speed is high (table 1) . The 31-year mean
for Amarillo, Texas, is 13.7 mph (anemometer height is 23 ft.), and the
wind speed is greater than 15 mph 35 percent of the time.
(2) The average wind speeds are consistently high throughout the
year with the strongest winds in the spring.
(3) The wind occurs both night and day with a small diurnal varia-
tion. The low and high averages by time of day for any month differ by
approximately 3 mph from the average value during windy months.
(4) The duration of calm periods (zero speeds) is short. For
Amarillo, Texas, from 1968 to 1972 there were only two, 9-hour calm
periods and six, 6-hour calm periods. The lowest daily average for the
5-year period was a speed of M.3 mph, which was on one of the days with
a 9-hour calm period. The wind speed frequency curves (fig. 1 and 2)
show that over 90 percent of the time the wind speed is greater than 5
mph. In fact for 1970-72 (3-hr observations) the wind was 9 mph, or
greater, approximately 80 percent of the time.
(5) High wind speeds are also common (table 2) with gusts of over
33
80 mph. During the spring of 1973 gusts to 100 mph caused extensive
damage in the area.
Preliminary calculations made from National Weather Service data

(Amarillo, Texas) give an indication of the energy in the wind, 153 to
212 kW-hr/(ft2- yr) for 1970 to 1972 (table 3) . Comparable results were
obtained for other years (fig. 3 and 4, data from the 1950's), and a
somewhat smaller value was obtained for Oklahoma (ref. 3).
There are several limitations that must be recognized in using

weather station data to evaluate the wind power potential. Wind speed
is known to vary with height. Obstructions are sometimes found close
enough to anemometers to affect their readings. And the method of
taking an average reading may indicate too little energy. Relatively
small increments in wind speed values lead to large differences in the
calculated values of the energy. For example, systematic errors of 2
and 4 mph give calculated energy differences of approximately 32 and
69 percent respectively. (See fig. 3 and 4.) Such differences in the
available energy magnitude may well determine whether wind energy cap-
ture is economically feasible.
We are beginning a wind energy survey based on the data compiled

at the 10 National Weather Service Stations within 275 miles of Amarillo
(fig. 5) . The wind -survey will provide data from which the wind energy
potential can be estimated for an integrated network. The type of infor-
mation calculated will be the statistical characteristics of the wind
and the time correlations between the stations. As stated earlier, this
area is large (area of the circle is greater than the combined areas of
New England, New York, and Pennsylvania), essentially flat, and acces-
sible; and it encompasses the high wind region.
We are also planning to collect data at sites in the immediate

vicinity of Amarillo. These data could then be correlated with the data
from the National Weather Service.
REFERENCES
1. Proceedings of the United Nations Conference on New Sources of

Energy, Rome, Aug. 21-31, 1961 vol. 7, Wind Power. United
Nations, 1964.
2. Thomas, Percy H. : Electric Power from the Wind. Federal

Power Commission, Washington, D.C., 1945.
3. Energy Storage..Key to Our Economic Future, Oklahoma State

University, School of Electrical Engineering.
DISCUSSION
Q: When you have a Weather Bureau Station a different height than you
like, how feasible is it by relatively short-term measurements at
different heights to get the effect?
34
A: He wants to know if you have Weather Bureau data at a certain height
then how do you get data for different heights. There have been some
tables put out on the difference in height. Say you took wind speed
measurements at 10 feet. Then, what can you estimate the wind speed
to be at 100 feet. About three different tables have been published,
and they really don't agree. I talked to a friend of mine who is in
meteorology, and he said, for example, between 10 and 100 feet just
add 60 percent. Now, I don't know how accurate that is.
Q: My question was about the measurement. Can you do better by particu-

lar tables?
A: I'm not sure actually. Once you get a correlation between heights,
your Weather Bureau data will give you a good idea of the wind ener-
gy potential in that region.
Q: I have a question and a comment. I believe that in the Western part

of your region there is quite a bit of high ground there, and, with
the speed-up factors on high terrain the British investigations
show, I wondered if you were going to get really much in that area.
The second comment: In the Texas area near Dallas there is a high
TV tower that has been instrumented. You might do some extrapola-
tions from that as it is one of the tallest instrumented towers on
the continent.
A: Yes, I have seen some data from that tower. It is in a region of

low wind down there. There is a range of mountains between us and
Albuquerque, and we don't expect much wind from that region. In
other words, it's only when you get out on the high plain that
you get high wind velocity essentially the year round.
Q: I was wondering whether you people had done any work on instrumenta-
tion that would measure wind energy directly. There has been some
suggestion that the anemometer that measures velocity is not what
you're interested in. What you're really interested in is v^.
A: Some people in England measure wind energy directly.
Q: Is that going to be used?

A-l:We don't know yet. Probably.
A-2:Just to a,dd to your reply, the type of the anemometer that the
French used for the wind service during the 1940's and 1950's was
an integrated type of meter. It gave out kilometers per meter
squared. This has certain characteristics itself. It acts as an
energy machine so it doesn't necessarily have the same characteris-
tics. There is a mass of data for France in kilowatt hours.
TABLE 1
MEAN WIND SPEED FOR AMARILLO, TEXAS (1942-72)
Jan. 13.2 July 12.4
Feb. 14.3 Aug. 11.9
Mar. 14.5 Sept. 13.1
Apr. 15.4 Oct. 13.0
May 14.8 Nov. 13.2 13.2
June 14.4 Dec. 13.2
Average = 13.7 mph = 6.09 m/s = 21.9 km/hr

*
TABLE 2
SPEED OF FASTEST MILE FOR AMARILLO, TEXAS
(Number of Days
50 mph
mph km/hr and i
1968 50 80 2
1969 59 94 3
1970 59 94 2
1971 54 86 1
1972 64 102 2
1973 65 104 3
36
TABLE 3
ENERGY (kW-hr/m2) IN THE WIND*
Month 1970 1971 1972

\
1 107 171 179
2 86 272 141
3 180 310 164
4 186 274 240
5 131 277 185
6 150 217 117
7 73 110 129
8 51 102 94
9 120 163 159
10 156 136 137
11 228 133 158
12 184 125 149
Totals (kW-hr/m2-Yr) 1652 2290 1852

Mtf-hr/ft2-Yr 153 212 172
*Values calculated from 3 hour observations of wind speeds,

anemometer at 23 ft., from National Weather Service, Amarillo, Texas.
37
Average Speed
600
WIND SPEED FREQUENCY .CURVE
at
The Amarillo, Texas National
Weather Service Station.
Data for this graph have been

taken from the Summary of Hourly
Observations, 1951 - 1960 and
smoothed to one mph speed intervals
and to one hour time intervals.
Anemometer height above ground: 33 feet
000
16 18 20 22 24 26 28 30 32 34 36 38 40 42
WIND SPEED (miles per hour)
FIGURE 1.
600 Average Speed

WIND SPEED FREQUENCY CURVE
at
The Lubbock, Texas National
Weather Service Station.
500
Data for this graph have been
taken from the summary of Hourly
observations, 1956 - 1960 and
smoothed to one mph speed intervals
and to one hour time intervals
B!
Anemometer height above ground: 68 feet
o, a
x
8
X t.
M Q)
3 a.
100
000
2 46 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42
FIGURE 2.
38
35 40 45
FIGURE 3.
10 15 20 25 30 35 40 45 50
FIGURE 4.
39
1
<U
4J
C
<U
O
H
I
WIND POWER DEMONSTRATION AND SITING PROBLEMS
Karl H. Bergey
University of Oklahoma
Norman, Oklahoma
I would like to discuss the nature of the work being done at the
University of Oklahoma on wind-power generation and on systems associ-
ated with the use of nonpolluting energy sources. Our activities are
primarily student-oriented programs aimed at specific educational goals.
They result from a conviction that wind power represents the most prac-
tical way to harness solar energy. The fact that Oklahoma is more than
generously endowed with wind is also clearly a factor.
At the same time, we concentrated on the development of a nonpol-

luting urban car that could use the resulting electrical energy directly.
We saw the two programs as being both educational and a contribution to
an increasingly critical social problem. We also felt that such pro-
grams encourage engineering students to think in terms of social goals,
an area which must become increasingly important in engineering educa-
tion.
In order to insure a general acceptance of alternatives to fossil

fuels, research in energy generation and in compatible energy usage must
be carried on in parallel. In the long run, there must be alternatives
to the "fuel tank" energy economy on which our present transportation
systems are based.
Speaking first of the transportation aspect, figure 1 shows the OU

Urban Car, a two-seat electric vehicle with an on-board auxiliary gen-
erator for extending the range and providing power for heating and cool-
ing. The range of the car is 25 miles on battery power alone and 50
miles with the on-board charger operating. A design goal was to avoid
the excessive weight problem which has reduced the performance and ef-
ficiency of many electric vehicles. Our test program includes measure-
ments of performance, drag, rolling friction, battery discharge
characteristics, and various other technical features. It also includes
an operational test program to determine the acceptability of the concept
in regular daily use.
The windmill studies have included technical and economic feasi-

bility studies and a hardware program for a small generator that can
provide overnight charging for the OU Urban Car. The blades for this
mill have a diameter of 12 feet and are coupled through V-belts to a
conventional automobile alternator. Under average wind conditions, the
mill is capable of completely charging the car batteries in 8 hours.
We have recently embarked on a program of modeling the dynamic re-
sponse of windmill systems using real wind characteristics and consid-
ering variable inertia, aerodynamic characteristics, and energy storage
systems. We plan to look both at the short-term response of the system
and the long-term energy balance associated with specific demand patterns.
In this connection, OU is fortunate in having the National Severe

Storms Laboratory of NOAA located on campus. NSSL has a 44-station
recording complex in Oklahoma and Texas, as shown in figure 2 (ref. 1).
These stations record wind direction, wind speed, pressure, temperature,
and precipitation. The number of stations per unit area is very nearly
an order of magnitude greater in the Oklahoma City-Norman area than is
currently available in most synoptic data. The purpose, of course, is
to provide an accurate portrayal of the surface winds during thunder-
storm activities. It is particularly valuable for investigating the
siting and performance of wind-power devices.
In addition, NSSL has instrumented the WKY-TV tower to obtain in-

formation on the variation of wind characteristics with height. It is
located within the area of high-station density for the NSSL surface
network. In addition to ground-level measurements, the WKY-TV tower is
instrumented at six levels from 146 feet to 1458 feet. Continuous
measurements of wind speed, wind direction, and temperature have been
recorded since 1966 (ref. 2). The information has been digitized and is
available both as annualized velocity and directional distribution and
in the form of short-term velocity distributions with intervals of 2
seconds, 10 seconds, and 1 minute. It is these data that serve as one
input to our windmill response modeling program.
Since the program is just starting, I believe it is too early to

discuss the project in any detail beyond commenting that our goals are
to investigate the effect of short-term gusts on the windmill/storage
system characteristics and the effect of long-term variations on energy
storage requirements with various cut-in speeds and installed capacities.
I would like now to comment briefly on what I see as some of the

research needs in the field of wind power.
First of all, I do not believe that we need general concept studies.

The concepts exist and, for the most part, the technologies also exist.
It seems clear that the development of wind-power systems can be thought
of in terms of relatively straightforward engineering programs. The
critical element is the willingness of government and industry to make
the necessary commitment to policies that will encourage the development
and use of wind power. This support need not be expensive, certainly
quite inexpensive when compared with AEC funding, for example, and
could take the form of development grants and subsidies. Both are con-
sistent with past government efforts to promote the use of emerging
technologies. Perhaps even more important, the expression of government
interest can have a multiplying effect on independent research carried
out by industries and in the universities.
42
I would like to emphasize particularly the role of practical demon^
strations rather than an extensive program of concept and feasibility
studies. The subject is old. It has been reviewed and investigated by
innumerable competent people over the years and, as Dr. Hutter's presen-
tation showed, the technology itself has reached a high level of develop-
ment in Germany and in other parts of the world.
As an example of the early recognition of some seemingly-new con-

cepts, I would like to quote J. B. S. Haldane, the eminent British
biochemist, writer, and teacher. In 1920, in a small book entitled
"Daedalus or Science and the Future," he wrote:
"Ultimately we shall have to tap these intermittent but

inexhaustible sources of power, the wind and sunlight.
The problem is simply one of storing their energy in a
form as convenient as coal or petrol."
He went on to say that the energy problem would be solved when:
"...at suitable distances, there will be great power stations,

where during windy weather the surplus power will be used for
the electrolytic decomposition of water into Oxygen and hydro-
gen. These gases will be liquified and stored in vats, vacuum
jacket reservoirs, probably sunk in the ground."
Note the word "intermittent." For the eventual optimization of

windmill/storage systems, we need to know how intermittent and we need
to know in some detail. The usual velocity information is obtained at
relatively wide intervals of time, which permits a comparative, and
therefore rational, basis for siting wind power stations. It does not
provide the detailed information needed by the designer to fine-tune the
optimization process. The point is that detailed wind information is
desirable for the long-term applications of wind power, but its lack need
not stand in the way of early demonstration programs.
From the purely technical standpoint, we need better information on

the performance of windmills in the vicinity of fixed obstacles and when
operating in tandem. We need better methods for modeling boundary layers
up to 500 feet in a variety of terrains and with relatively few data-
points. We need to develop low-cost blade and tower structures, reli-
able control systems and efficient storage methods.
Perhaps even more important is the need to consider the matter of

social impacts and public acceptance. The appearance of individual mills
is important and the siting must be reasonably unobtrusive and based on
careful public preparation.
I do not believe the latter point can be overemphasized. Only a

few years ago, nuclear power was being hailed as the ultimate source of
clean, nonpolluting and inexhaustible energy. Environmental considera-
tions and public objections have essentially halted new construction of
nuclear power plants, and the prospects for the immediate future are not
bright.
43
In this connection, a recent study at the University of Texas re-
viewed the siting problems of a variety of nuclear power plants and com-
pared their public acceptance with the approach taken during their early
development. The study found that the successful programs had followed
a consistent pattern of full public disclosure, even to the point of
announcing where the power plant would be located before completing the
plans and buying the land. Although the land costs were higher as a
consequence of this policy, the overall costs were considerably less than
those associated with strong and well-organized public opposition.
The point to be made here, I believe, is that conventional business

wisdom is not necessarily the best guide to decisions that have a strong
public impact. In the long run, there would appear to be no substitute
for full public disclosure and adequate public discussion prior to a
major commitment of funds.
Finally, I would like to address briefly the subject of how develop-

ment and demonstration of wind power might best be carried out.
I believe that we must avoid the trend toward technological over-

kill. By this I mean that we should not attempt to do the definitive
study in 6 months with 50 people. I recognize that there is often a
considerable pressure on funding agencies to come up with early judgments,
but if the information is hurried and if the judgments are not mature, it
is likely that more harm than good will result.
I make this comment because it seems to me that I have seen a grow-

ing tendency on the part of funding agencies to send out RFP's with ex-
tremely short response time and high manpower densities. I happen to
believe that this is not an efficient way to get at the type of long-
term problem which wind power represents. I do not mean that we should
neglect certain large or comprehensive studies, but I do believe it is
valuable to take advantage of the widespread interest in new power sources
by parceling out a relatively large number of small grants. Such grants
serve as seed money for other sources of sponsorship and, if issued in
the form of student support, for example, have a strong multiplying ef-
fect by involving interested faculty members without cost. Furthermore,
such programs give the sponsoring agencies a calibration on the interest
and competence of the various universities and organizations, which is
bound to be helpful when considering commitments to larger or more
expensive programs.
It is important to strike a reasonable balance between research,

system studies, and demonstration projects.
Aside from the technical programs, we should place a strong emphasis

on public policy issues, the effect of the technology on society, and the
interaction of State and Federal regulatory actions.
REFERENCES
1. Operations Staff, The NSSL Surface Network and Observation of
Hazardous Wind Gusts. NOAA Technical Memorandum ERL-NSSL 55.
2. Crawford, K. C. and Hudson, H. R. Behavior of Winds in the Lowest

1500 Feet in Central Oklahoma: June 1966 - May 1967, ESSA Tech-
nical Memorandum ERLTM-NSSL US.
Figure 1
NSSL SURFACE NETWORKS - 1973

AVK PNC evo
TNK O
EN.
SWO
OKM
EJU
+
EU, WKY* CHO
+ 6 OKCT1K+
CSM
SMO
-* LHS
ous 4-34*N
W-W -I- 96'W
"oliwS. «A. ANCTMiDTARV SITES
-t- NSSL SURFACE SITES
A NSSL TOWER SITE
6N5SL POPPLER-Z
Figure 2
SURFACE WIND CHARACTERISTICS OF SOME ALEUTIAN ISLANDS*
Tunis Wentink, Jr.
University of Alaska
Fairbanks, Alaska 99701
Our long-range goal is to establish (and, hopefully, exploit) the

wind power potential of Alaska, one-fifth the area of the contiguous 48
United States (fig. 1). Important corollary questions are access to
"promising wind power sites for construction of test or permanent wind
machines and shipment of the wind-derived energy. The "packaged" form
of the energy must be considered for possible export to the "lower 48"
or use in Alaska. Also, the wind regimes involved will dictate wind-
mill design, including, of course, the economics.
A first step is our analysis of near-surface wind data (ref. 1)

from some promising sites accessible by ocean transport. We consider a
few Aleutian sites (ref. 2) here, to indicate probable velocity! regimes
and also present deficiencies in available data. Cold Bay and Dutch
Harbor are two such wind power sites, (fig. 2, - areas numbered 12 and
15, respectively (ref. 3).)
COLD BAY, ALASKA
Cold Bay has a 5-year average wind velocity of 15.1 knots (17.4 mph) .
An airstrip accommodating 747Ts, a harbor permitting 30-foot draft ves-
sels, and a large, wide, unshielded (N and S) plain ideal for a windmill
farm all make this area a prime candidate for initial Alaskan large-scale
wind power investigations. The monthly average cycle (ref. 4) is shown
in figure 3 (more on this later). Figure 4 gives the velocity duration
curve averaged for all months. All such curves mask short-term fluctua-
tions, like those of figure 5.
Cold Bay data show surprisingly little wind speed variations with
height. Simultaneous measurements at various heights near ground are
desirable, as at any site. However, a shift in anemometer height from
88 to 21 feet showed no significant change in monthly velocity distribu-
tion curves for comparable months.
*Supported by State of Alaska funds.

•^Speed and velocity are interchanged here. Wind direction is not
considered, but the winds are cyclonic and, while often from the SW, can
shift rapidly in all directions in most areas.
DUTCH HARBOR (UNALASKA ISLAND)
For Dutch Harbor the available wind data (figs. 3 and 6) are disap-
pointing, in view of the fact that it is the best and most sheltered
harbor in the Aleutian Chain. These data are a good "horrible" example
of data obtained for purposes other than wind power generation. The
present values are from an anemometer at the airstrip, which is shel-
tered on all sides from most strong winds and especially those coming
in from the sea, north and south. However, our 1973 on-site inspection
indicated that there are at least three areas in the vicinity that seem
ideal for windmills; long term weather data should be gathered there.
The Dutch Harbor and Driftwood Bay (also Unalaska Island) monthly
mean velocities (fig. 3) illustrate probably typical horizontal differ-
ences in the Aleutians for nearby areas. Driftwood Bay is 14 miles
northwest of Dutch Harbor. The yearly means are 8.3 and 9.3 knots,
respectively (ref. 5).
GENERAL AREA INFORMATION

Table 1 contains data for eight Aleutian sites, including frequen-
cies of selected velocity ranges. The 7- to 21-knot (8 to 24 mph) range
covers the cut-in to near peak power speeds for small available generators.
In general, in the Aleutians, winds for some degree of power generation
are available 77 percent of the time (averages of table 1). Since the
data available so far are from sites generally chosen for nonwind power
purposes, the values of speed and frequency are probably lower limits
for wind power planners.
Wind speeds to 140 mph are reported spasmodically in the North

Pacific Ocean - Bering Sea area, but are seldom verified. Peak veloci-
ties depend on location. The record at Cold Bay is 73 mph, at Amchitka
above 115 mph. At Amchitka winds above 70 mph can last for several hours.
REFERENCES
1. T. Wentink, Jr.: Wind Power Potential of Alaska: Part I, Surface
Wind Data for Specific Sites. IAEE, Univ. of Alaska, Fairbanks,
Alaska, June 30, 1973.
2. B. Garfield: The Thousand-Mile War. Ballantine Books, Inc.,

New York, N. Y., Paperback ed., 1971.
3. Interagency Arctic Research Coordinating Committee: Arctic Research

Logistics Support Handbook. National Science Foundation, Dec. 1972.
4. Climatolography of the U.S. #82-49, Decennial Census, Summary of Hour-

ly Observations (for various stations), 19S6-60 (GPO 1963) ; also
specific location Local Climatological Data, Monthly and Annual Sum-
mary. U.S. Dept. of Commerce, NOAA, National Climatic Center,
Asheville, N.C. 28801.
5. Unpublished (generally) data, Air Weather Service, NTC - Asheville,

as in ref. 4.
TABLE 1
TIME PERCENTAGE OF SELECTED WIND VELOCITY RANGES IN ALASKA
Annual % of Time % of Time

Location** Mean, Knots at 7-21 Knots above 21 Knots
Adak NS (AD)** 13.1 (1942-65) 60.9 16.2
Amchitka AFB (AM) 18.3 (1943-50) 60.1 30.7
Cape Sarichef AFS 13.7 (1952-56?) 57.5 18.5

(ID
Cold Bay* (12) 15.1 (1956-60) 64.0 24.0
14.8 (1965-71) Data discrepancies need resolution.
Driftwood Bay AFS 8.3 (1959-69) 56.9 1.9

(15)
Dutch Harbor NS 9.6 (1946-47,

(15) 1950-54) 53.3 6.9
St. Paul Island* 16.4 (1972) 68.9 24.2

(SP)
Shemya AFS 15.1 (1943-53)

(45) 16.2 (1950-72) Data discrepancies need resolution.
'Commercial Airport.
**
See Fig. 2.
DISCUSSION
Q: I am interested in the freezing rain potential of your area. Do

you think this might come to be a problem?
A: As I see the major problems they are three. Some of them can be
checked out early.
One is the mechanical behavior of a large windmill during peak

velocities. Can you feather quickly enough? I think the indica-
tions are that most of the properly engineered windmills will take
above 100 miles an hour. What will they do at 150 miles an hour?
I don't think there is any problem in towers. What I am really con-

cerned about especially in the Aleutians is the prevalence of this
high velocity mist they speak about, almost horizontal rains. Ex-
perience indicates that in the Aleutians it's almost impossible to
seal an electronic device against these winds. If you are going to
mount your generators on towers behind the blades, you may have to
pressurize the generators to keep them dry.
The problem of icing should be checked out. There are indications

in other parts of the world that icing is no problem. It is not
particularly cold in the Aleutians, but the snow is quite wet.
These are problems that will be faced.
A L A S K A SUPCHIMPOSED
O N T H E UNHID S T A T E S
FIG. 1 ALASKA (586,400 SQ. MILES) AND CONTIGUOUS UNITED STATES (3,022,400 SQ. MILES)
50
SURFRCE WIND DflTfl
(FROM HOURLY OBSERVRTIQNS)
STFTTION- COLO BRT. RK
TERRS- 1956-60
MONTH- BLL
RVERRGE VEL.- 15.1 KNOTS
CRLM- 3.0 Z
8
UO 60 80
PERCENT OF YEflR flBOVE V
SURFACE WIND SPEED VARIATION AT COLD BAY, ALASKA
(3-hour observation Interval)
Random data selection, from February 1972.
8
40 60
PERCENT OF YEflR flBOVE V
52
WIND POWER RESEARCH AT OREGON STATE UNIVERSITY
E. Wendell Hewson
Oregon State University

Corvallis, Oregon
The program of wind power research at Oregon State University com-

menced late in 1971 under the sponsorship of four Oregon Peoples Utility
Districts, those of Central Lincoln, Tillamook, Clatskanie, and Northern
Wasco. The interdisciplinary research team consists of faculty members
from the Departments of Atmospheric Sciences, of Electrical Engineering,
of Mechanical and Metallurgical Engineering, and of Physics, and from
the School of Oceanography.
There have been two primary thrusts of the research effort to date,
along with several/supplementary ones. One primary area has been an
investigation, in a preliminary manner, of the wind fields along coastal
areas of the Pacific Northwest, not only at the shoreline but also for a
number of miles inland and offshore as well. Estimates have been made
of the influence 'of the wind turbulence as measured at coastal sites in
modifying the predicted dependence of power generated on the cube of the
wind speed. Wind flow patterns in the Columbia River Valley have also
been studied but in less detail.
The second primary thrust has been to substantially modify and im-
prove an existing wind tunnel to permit the build up of a boundary layer
in which various model studies will be conducted.
One of the secondary studies involved estimating the cost of build-

ing an aerogenerator of the Smith-Putnam type at 1971 prices.
WIND PATTERNS AND SITING PROBLEMS
The wind patterns at a substantial number of coastal sites have

been examined. Some of the wind stations were established by the present
project, with locations chosen after examining the criteria set forth for
the Vermont Smith-Putnam location (ref. 1) and for various coastal sites
in Wales and Scotland (ref. 2). The locations of both the older and
newly established wind stations are shown in figure 1.
The terrain of the coastal areas of Oregon bears a marked resem-

blance to that of western Wales and Scotland where a comprehensive wind
power survey was conducted a number of years ago (ref. 2). The wind
patterns are substantially different, however. The pressure gradients
over the British Isles lead to west winds which blow up and over the
53
coastal higher ground. On the other hand, the pressure gradients over
the coastal areas of the Pacific Northwest lead to prevailing northwest
winds in summer and southwest winds in winter. The High Cascades beyond
the Coast Range provide an additional barrier which tends to promote a
lower level airflow parallel to the coast rather than perpendicular to it
both summer and winter. As a result, the most promising wind power sites
appear to be right on the coast or over nearby offshore waters.
Referring to figure 1, our research to date indicates that substan-

tially stronger winds occur at our lower wind stations at Cape Foul-
weather and at Yaquina Head and at the Columbia Lightship, which is
stationed several miles off the mouth of the Columbia River. Relatively
high winds also occur over many of the bluffs which extend for several
hundred miles along the coast of Washington, Oregon, and northern
California.
Another attractive site for wind power development is the Columbia

River Valley where limited studies only have been made. Topographic
features along a stretch of the river valley east of Portland are illus-
trated in figure 2. The valley sides rise to heights of 4000 feet mean
sea level (msl) to the south and perhaps 1600 feet msl to the north. On
September 5, 1972, serial pilot balloon observations were made at three
points in the valley, at Cascade Locks, Wyeth, and Viento Park whose
locations are shown in figure 2. Balloons were released and tracked at
each station at intervals of 1 to 2 hours during the day. The results
are presented in figures 3 to 5. Each figure gives isotachs of equal up
valley speed in knots. On this September day the winds in the valley
appear to average about 20 knots (23 mph; 10 m/sec) if one allows for
the tendency of the wind to increase with height. One purpose of these
serial pilot balloon measurements was to determine whether or not signif-
cant large-scale turbulence occurs in the valley. One such region of
turbulence is shown by the hatched area in figure 3 where a strong ver-
tical gust caused the single theodolite method to fail. Future observa-
tions in the area will use two-theodolite tracking which avoids this
problem. It is obviously important for wind power purposes to determine
the magnitude and frequency of the occurrence of such gusts.
Two of the more attractive areas for wind power development are the
offshore coastal waters and the Columbia River Valley, as suggested by
the power duration curves presented in figure 6. The British aerometric
survey referred to earlier showed one of the highest average wind speeds
to be 21 knots (24 mph) at Rhossili Down, Glamorgan, Wales at a height
of 633 feet (193 m) (ref. 2). The power duration curve for Rhossili
Down is shown in figure 6 for comparison with corresponding curves for
the Pacific Northwest. A power duration curve for an inland British
station at a height of 267 feet (81 m) having limited wind power poten-
tial is also shown in figure 6.
The coastal terrain obviously has a substantial effect in slowing

down the stronger offshore winds as the comparative data for the Columbia
Lightship and Astoria (fig. 1) given in figures 6 and 7 show. The wind
records at two of the newly established wind stations, Cape Foulweather
and Yaquina Head, do show high winds, but the instruments have not been
in operation long enough to permit the establishment of reliable power
duration curves. Details of these studies are available (ref. 3).
WIND TUNNEL STUDIES
The purpose of the wind tunnel studies is to develop appropriate

simulation methods to the point that it will be possible to locate at-
tractive coastal wind power sites from model studies without the neces-
sity for taking lengthy and expensive on site observations. Analyses of
various types of aerogenerators through the use of models will also be
undertaken. Work to date has been directed to modifying and improving
an existing return-flow wind tunnel to permit the buildup of a boundary
layer in which the model studies can be conducted. A 1 to 300 accurate
scale model of a portion of Yaquina Head on which one of our anemometers
is located has been constructed and is ready for insertion in the tunnel.
The enlarged tunnel, with its 5- by 4-foot test section 30 feet long, is
shown in figure 8. Other information is also available (ref. 3).
TERRAIN MODIFICATION
One area to be investigated by wind tunnel studies is the possibility

that terrain modification may result in appreciably augmented average
wind speeds. The concept that will be explored is illustrated in figure
9. The diagram is self-explanatory.
AEROGENERATOR ARRAYS
Another area for exploration is the pros and cons of arrays of aero-
generators, especially of inexpensive, mass produced vertical rotor units.
An inexpensive variant of the Savonius rotor is sketched in figure 10.
The vertical hemicylinders, in sections of appropriate length, would con-
sist of corrugated steel culverts cut in half. An array of such units is
sketched in figure 11. The system of vertical rotors is maintained by
guy wires as shown; the only compression members are the vertical shafts
of the rotors which are themselves stiffened by the four hemicylinders.
Each rotor drives directly, without the need for slip rings, a multipole
generator of modern design which is housed below grade in a suitable
enclosure.
Although the efficiency of such units is low, this may be more than
offset by low cost and the need for little maintenance. Such arrays may
be much larger than the one illustrated.
REFERENCES
1. Putnam, P. C.: Power from the Wind. New York, D. Van Nostrand, 1948.
2. Golding, E. W.: The Generation of Electricity by Wind Power. London,

E. & F. N. Spon, 1955.
55
3. Hewson, E. W., et al.: Wind Power Potential in Selected Areas of
Oregon. Report No. PUD 73-1, Oregon State Univ., Corvallis,
Mar. 1973.
DISCUSSION
Q: Professor, with respect to the electric utility industry, would you

care to comment with reference to future, what the common objectives
are?
A: The Oregon PUD Directors Association allocated $150,000 over three
years. At the end of the end of this period they will probably
consider a pilot program to get at least one windmill running.
Columbio
Lightship
Asjorio A
WICKIUP RIDGE
Seaside
TILLAMOOK HEAD
Mount Hebo
Lincoln City/* A C A N N E R Y MOUNTAIN

CAPE
FOULWEATHER
YAQUINA HEAD
Woldport
Older wind stations
Newly established
Florence wind stations
Winchester Boy
Fig. 1 Wind stations on or near the central and northern portions of the
Oregon coast.
56
n)
tu
O
4J
e
ai
§
M
I
I
O
10
n)
n)
a.
C
rt
M
0)
O
O
57
(It)
§
o
«
5
a
- 6000
-4000
- 2000
0
0 400 800 1200 1600
Hours during which Power Available is greater than Value Given by Curve
Fig. 6 Wind power duration curves for: Rhosslli Down, Wales; the Columbia
Lightship off the mouth of the Columbia River; Cascade Locks, Oregon;
Astoria, Oregon near the mouth of the Columbia River; and an inland
site in Great Britain.
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
1968
Fig. 7 Comparison of 1968 wind speeds at the Columbia Lightship with those
at the nearby coastal station of Astoria
59
Fig. 8 The exterior of Che test section of the wind tunnel; direction of
air flow is from left Co right.
+ fill
— excavation
Fig. 9 Proposed type of terrain modification for Che purpote of augmenting
average wind speeds
60
-Ck
I [
II II
|« — U 002
h
*:
«;
A
•a
-4.
61
WIND DATA FOR WIND DRIVEN PLANT
Arthur H. Stodhart
Electric Research Association

Surrey, England
To obtain the best performance with any type of machine, it is es-

sential to have a good understanding of the fuel used. It is also neces-
sary, for the broader view, to know where that fuel can most readily be
found, the rate at which it can be extracted and used, and the magnitude
of the resource. This is as true with wind as with any other fuel, ex-
cept that this resource does not reduce with use.
What data do we have on wind? We know that of the 1.5xlQl8 kilowatt-

hours (kW-hr) of solar energy falling on the Earth annually some 26000x
10^2 kW-hr are converted to air in motion and that a small fraction of
what is accessible in the first 100 meters or so above the Earth's sur-
face should provide a resource at least equivalent to the world water
power reserves.
Wind is not distributed evenly over the globe, being on the average
more plentiful in the temperate and polar latitudes and, almost everywhere,
higher in coastal areas than inland for the same type of terrain. (See
fig. 1) This same picture is shown in a more local context by the
"isovent" map of the British Isles. (fig. 2) This type of map is pro-
duced from standard MO (Meteorological Office) data, based on standard
MO anemometers, supplemented by spot readings in some sites and by visual
observations reported on the Beaufort scale. It relates to open situa-
tions in level country at 10 meters above ground. Its use for wind power
purposes is to indicate areas that are worthy of further exploration.
Why do we need to explore further? Simply because, if wind regimes

much better than those indicated by the isovent map cannot be found,
there is no economic case for wind power use at present day fuel prices.
What must be found are sites where the specific output of wind-driven
plant rated at wind speeds of around 15 meters per second (30 to 35 mph)
is between 3500 and 4500 kW-hrAW. Typical velocity-duration and power
duration curves for such sites are shown in figure 3. Analysis of a
large number of curves of this type shows that their shape does not
differ widely and that there is, for any given rated wind speed, a
roughly linear relationship between specific output and mean annual wind
speed (fig. 4). It is not, therefore, necessary to use elaborate re-
corder equipment at each and every potential wind power site.
An adequate, and economic procedure in any geographical area is to

make comprehensive measurements, including hourly mean values, at one site
62
and obtain only weekly or monthly integrations of wind at a series of
others.
The type of anemometer used is not of particular importance, but it

is preferable to use identical types for all the general wind survey work.
How should we choose these sites? All past experience indicates

that smooth-shaped hills with all-round exposure provide mean wind speeds
well above the average for the surrounding countryside. In temperate and
northern latitudes, hills below the normal winter snowline areipreferable.
It may be that studies of the ecological evidence, as suggested by Putnam,
will help in the initial choice of site, but such evidence is mot always
available. Typical hill sites in the British Isles show average annual
wind speeds 35 to 50 percent higher than at lowland measuring stations.
Characteristic of the best sites is a slope of about 1 in 3% (16°) in the
final few hundred meters approach to the summit and the absence of a flat
top to the hill. Conical shaped hills are preferred to ridges. Another
advantage with such hills is the reduction in the vertical wind gradient.
In level country of average roughness, and with neutral stability, the
exponent of the gradient is about 0.17 (i.e., V ~ nP-1? where V is the
mean hourly wind speed and h the height above ground). This is undesir-
able for aerogenerator operation since it imposes additional cyclic loads
on the wind rotor. On ideal sites, because of the compression of the
streamlines over the summit, this gradient is reduced, certainly in the
first 50 meters, to give values of the exponent of 0.05 or less. For
slopes of 1 in 6 the exponent is around 0.10 and for very shallow slopes,
it approaches the value for level country.
All of this information can be obtained from hourly mean wind speed
data.
To obtain useful data for structural design purposes requires, in

the first instance, more detailed information, such as can be provided
by sensitive anemometers installed in vertical and horizontal arrays on
one or more typical wind power sites. Since the earlier wind power work
was undertaken, considerable advances have been made in the collection
and analysis of short-term wind data and its use for structural design
purposes. The aerogenerator designer can benefit from these advances.
Associated with the period of 1 hour over which means are generally
taken, there is a continuous random signal comprising the fluctuations
of the wind about that value. These "gusts" need to be described, and
to do this the methods developed in communications and control engineer-,
ing, based on probability theory and statistical techniques, can be
applied. By separating the gust vector from the mean, the rms gust
speed and intensity of turbulence can be calculated. This leads to the
important conclusions that the former is virtually independent of height
and that the latter decreases with height (figs. 5 and 6).
Next, for a dynamical approach to wind loading, it is necessary to

describe the evolution of gust velocity in time and its variation in
space. The time structure of random signals (in. this case gusts) can be
63
described by the auto-covariance function, or by a normalised version
known as the auto-correlation function. The latter is a measure of the
information a gust component at one instant of time gives about the
value at a later time. Gust properties can also be described by means
of power spectra, an extension of Fourier analysis principles to non-
periodic random signals. The power spectrum of a signal can be defined
in terms of the contribution to the total variance coming from simple
harmonic components in a defined band width of the continuous spectrum
centred about a given frequency. It has been shown that the power spec-
trum of the longitudinal gust component can be fitted into a simple ex-
pression having as parameters the hourly mean wind speed at 10 meters
and the surface drag coefficient. Typical gust spectra are shown in
figure 7.
Results so far obtained confirm that the power spectrum provides a

description of the evolution with time of the random gust velocity ade-
quate for many structural loading problems.
Also important are the space average properties of gusts, and these
can be obtained from cross-correlations for zero-time lag, which provide
a measure of the relationship between simultaneous values of gust com-
ponents at different points. These can be combined with the time rela-
tionships to show correlations at different points for different time
lags (see fig. 8). Application of these methods to aerogenerator design
could overcome past difficulties of relating wind behaviour to structural
performance.
Finally, it is useful to accumulate data on extreme wind speeds in

order to assess the likely probability, or return period, of any given
value that may be relevant to the machine or tower design. However, be-
cause present techniques require the collection of maximum values of wind
speed over many years, it will initially be necessary to learn what we
can from less windy sites where such information is at present available.
The technique is straightforward; the highest values for each year of the
period are ranked in order from lowest to highest, a plotting position is
calculated, as is the reduced variate (see fig. 9). The results can then
be plotted (fig. 9(a)) and confidence limits drawn in. The hourly mean
values are immediately meaningful; the gust values are so only if the
term "gust" is defined. What may be more useful is to derive the proba-
ble values of short-term means from the hourly mean values. For level
country inland the following relationships are typical.
60 min 10 min 1 min 20 s 5s 0.5 s

mean . mean mean mean mean mean
1-° i-06 1 33
- 1 36
- 1 47
' 1 59
-
For good wind power sites the ratios of the shorter term means to
the hourly means will probably be less than these; a full examination of
wind data from such sites is lacking.
The relative importance of extreme wind speeds to wind driven plant
has not been established. Under these conditions the plant will be shut
down and loading on the rotor and the tower could well be less than under
full power conditions. Extreme data for low level sites in the UK are
shown in figures 10 and 11.
Summarizing, the use of simple, averaged data will provide informa-

tion on energy availability, facilitate site selection and enable appro-
priate operating ranges to be established for wind-driven plant. It
will also provide a basis for the prediction of extreme speeds. For
structural design purposes the more detailed shorter-term data are re-
quired, and more sophisticated methods of analysis must be utilized and
applied.
Fig. 1.—Availability of wind energy

Annual specific output of windmills rated at 2Smilejh
65
I
.s
iE 1S
5° I
,fc a
'* 1
1 I
c 2
C u
9 I
O
O O
O O
II
1 "S
S S
1
I
puiyv\
66
R.M.S. gust speed (mile/hj
4-5 5-0 3-3
1 1 = ODU ^ '£ 2OO 650 J
.6QO 6 & .
600 s
5 1 a
- 55O jj J; X
550 $
I I I
- 500 | 1 - 500 §
- 450 - 450
- 400 125 - 4OO
. 350 - 350
100 •
- 300 - 3OO
- 250 „ . 250
fl
- 200 200
- 150 *> * 150
• 100 * 100
• 25 *
> 50
* SO
, ft n 1 1 t
2-2 2-6 Q-10 0-12 O-14 0-16 O-16
R.M.S gust speed <m|s) Intensity of turbulence
Figure 5 R.MS, gust speed measured at Rugby

Figure 6 Intensity of turbulence measured at Rugby
I ''Of
Sale
O Cardington
7- Ann Arbor
+ Cranfield X Even correlation
0 Brookhaven
— Theoretical curve (£-lBOOm (SttOOfO)
O-OO02 O-OO05 0-O01 O'OOZ 0-OO5 0-01 002 0 OS 0-1 0-2 O-5
Wave number at 10 metres (cycles/ metre)
Figure 8 Cross correlation between long, component at 116-4 m (382 fi) and long, component
Figure 7 Comparison of the original and modified gust spectra at I00-0m(328 ft)
67
o
ho
CM
-8
-8
a:
IU "£•
100. *
i
^^N
< o
CO*
oo IA' 5f 00
c o c o m c o - g 5 C M c o i n N e D C N i n O < g i n « > O - - ' v« » - l o
K m T j - f o i O O — <Nco^.cDNcnOol'^'' On —
0 sO s13 - 0 O O O O O O O O O O O O O O
1 1 1 1 1 1 1 1
CM<N<->
oo in
Q 1 CK
Uro o
O_CJ) 1-
CO 05
Q7r Ê|i
> f< l >
CM CO l r > N C o O ' l ' ' N0<>OcMcoinNOOOcMCOinNOD
Tj- 00 o i c o O m c D o o N — i n O ^ o o c M c o O i n c n c o N — tn
•^~o
>^_
g
1- Q.
(| 0 0 — — NcMoiooco^j-^-ininincocoNNNoococncn
O 0 O O O O O O O O O O O O O O O O O O O O O
o 3
Set: «
Ul
co O — — C M N C D C i M c o c n c o i n O T j - ^ - ^ c o c o c M c o o o i n K
in in ^ - i n u > c o c o - ^ c O ' i i j - x j - - « < - > ( j - c o ^ i n ' * - ^ c o c o o o c o ^ -
cn o) Q} Ol U} <2) OT Ot O) OT C77 OT O7 QJ OJ O> W 01 O? O*> O) ift OT
— ^
X
<rO
CO
^ 1— 3 <-•
0
CL in cn O — C M C O C O ^ I n N C O O ) — C M C M C O O O O O — C M C O O D C O
_1< in in to to co c o t o t o c o c o c o t o N t t s N N r s r x C D o o o o o o o o
fc?
u
Z> I X
0
Z
1 — CM coM-mtoNcocnQ — cNoo-»mco|va>cnO — Noo
<E
IX
68
69
AN INTRODUCTION TO THE PRINCETON SAILWING WINDMILL
T. E. Sweeney and W. B. Nixon
Princeton University
Princeton, New Jersey
Generally discussed is the Princeton University interest in a wide

range of wind machines. Specifically discussed is one example of the
work - the Sailwing windmill. The aerodynamic characteristics of the
Sailwing itself are presented in condensed form and its natural appli-
cation to the wind machine is discussed. Past and present Sailwing
windmill configurations are shown and their relative merits are com-
pared. A section on a future promising configuration is presented and
its compatibility to advanced technology electrical machinery is briefly
discussed. Also included is a short bibliography.
BIBLIOGRAPHY
Sweeney, T. E. : Exploratory Sailwing Research at Princeton University.

Princeton Aero. Report No. 578, Dec. 1961.
Ormiston, R. A.: Theoretical and Experimental Aerodynamics of the

Sailwing. J. Aircraft, vol. 8, no. 2, Feb. 1971, pp. 77-84.
DISCUSSION
Q: What is a reasonable size for one of these sails? Could you get up
to 100 feet or so?
A: Possibly, but there is a crossover point. We had studied it in
reference to the entry body back when we were talking about Skylab:
fixed wing versus something you could fold up. If you send up heavy
big loads, the structure get so heavy, that you lose your whole
weight advantage. So we're pretty sure we're good at 25, and I
would bet 50. When you get to 100 I'm going to leave.
Q: What sort of loadings do you use in a sail application?

A: For the airplane type of application, 10, 12, up to 15 pounds per
square foot, rather light.
Q: On the aerodynamic comparison between the solid and the sailcloth

type of configuration, the difference there is that you built the
solid rotor the same as the sail one. If you did, it should have
the same characteristics.
A: Yes, the characteristics would be the same if a conventional metal
wing could crinkle and bend the way a sailwing does. The advantage
of the sailwing is its flexibility.
70
L.E. = Leading edge
I.E. = Trailing edge
Fiq.1 Fig.2
Sail wing types of blade
J I ! Lf
-/o S o 5" lo IS" 2& -I" 5 O 5 ie> If 2o°
Angle of attack.
Fig.3a Fig.3b
Aerodynamic characteristics of sail-wing airfoil
71
Princeton sailwlng wind generator
72
THE USE OF PAPER HONEYCOMB FOR PROTOTYPE BLADE CONSTRUCTION
FOR
SMALL TO MEDIUM-SIZED WIND DRIVEN GENERATORS
Hans Meyer
Windworks, Inc.
Mukwanago, Wisconsin
Windworks, under the sponsorship and direction of R. Buckminster

Fuller, began working on wind energy conversion systems in 1970. It was
decided, in view of the high cost per kilowatt output and the relative
difficulty of construction of conventional or existing wind plants, that
the first area of emphasis should be in making wind more accessible for
experimentation and use.
With this in mind, we began working with paper honeycomb for the
construction of conventional, propeller-type, windmill blades. Using
fairly simple techniques and conventional power tools, it is possible to
shape both simple foils (NACA 4415) and more complex foils (Wortmann
FX-60-126 and FX-72-MS-150A) with or without tapered plan forms
with or without varying profile. Still more complex geometries can be
developed using router techniques developed by Hexcel Corporation.
For blade diameters up to 30 feet, typical costs are as follows:
Honeycomb $ 0.14/ft2 of blade

Fiberglass/resin .90/ft2 of blade
$ 1.04/ft2 of blade
Construction time 8 to 12 hr/blade
Tolerances:
Paper t 1/32 in.
Aluminum + 1/100 in.
The first step of the process is to cut out the blade blank. A
block of honeycomb, in its compressed form, is mounted on a wedge and
run through a handsaw with the table at an appropriate tilt angle. It
is the combination of the wedge angle and the table angle that gives the
tapered plan form and profile shape.
Next the honeycomb is expanded on the shaft and jigged to give the
desired angles of attack. With the honeycomb fixed in position, the
blade is covered with a fine weave fiberglass cloth. Any surface quality
can then be achieved with filling and sanding.
73
The process, being both simple and low cost, "Ien4s itself particu-
larly to prototype work and tool making. In encouraging individuals to
use and experiment with wind energy, we hope to increase the support for
wind utilization which will be necessary for the acceptance of large-
scale developments in this country.
DISCUSSION
Q: In any of your units do you have any feathering mechanisms or do you

keep the angle constant?
A: No. But we do have feathering. We have two types: we have Popular
Science typical coning — incidentally, in the U.N. report there's a
Japanese windmill for which are shown test results for absolute con-
stant rpm with a mildly sophisticated feathering system. It's a
cone windmill running downwind with variable pitch, and they have
been able to run that at constant speed regardless of wind velocity.
We've worked on two different types: the coning straightaway, which

is an umbrella structure working against a spring with damping, and
the flyball governor wherein as the blades move out they rotate. When
the windmill shuts down, the blades go into a partial coning position,
which increases the starting torque. As it starts up, the blades
begin to rotate into maximum power, and then they rotate further into
full feather.
Q: What kind of increase in power did you get with the Venturi proto-
type?
A: We designed for a 50 percent increase in windspeed. We got about a
35 percent increase.
Q: What was the area ratio, minimum and maximum?

A: It had a 6-foot opening, a 5-foot diameter, a 7-foot exit, and an
80-foot length from entrance to exit.
THE SAIL WING WINDMILL AND ITS ADAPTATION FOR
USE IN RURAL INDIA
Marcus M. Sherman
New Alchemy Institute

Woods Hole, Massachusetts
A 25-foot diameter sailwing windmill was built in 1973 in a small

village near Madurai, Tamilnadu State, India. This windmill is the re-
sult of design research conducted in the U.S.A. at the New Alchemy
Institute—East (refs. 1 and 2) and in India at the Indian Institute of
Agricultural Research (refs. 3 and 4) and the Wind Power Division of the
National Aeronautical Laboratory (ref. 5). It is to be used mainly in
light winds during the dry winter months for irrigating small fields,
watering dairy cattle and supplying water for domestic use.
In many parts of India there are adequate supplies of ground water

which are unavailable to farmers during the dry season because of inade-
quate power resources for pumping. Three to eight horsepower diesel
pumps are frequently used, but they are expensive to operate because of
the high cost of imported oil and often must be taken out of service for
costly and time-consuming repairs. Efficient 5-horsepower electric pumps
are being used more and more as rural electrification proceeds, but only
well-to-do farmers can afford to buy and maintain them. Recently in
South India there has been a 75 percent power cut to the rural areas due
to heavy use in the cities and to overexpansion of the power grid without
a corresponding increase in supply. This power shortage means that there
are only 4 hours of electric pumping per day. This situation is expected
to worsen for the next 4 to 5 years until the Indian Government begins
operation of atomic power plants in South India. At the present time
bullock operated pumps remain the most common and reliable source of
-irrigation water for subsistence farming. Water for domestic use is
usually~-hand-lifted with a rope and bucket from open wells.
During the early 1960*3 the Wind Power Division of the National
Aeronautical Laboratory in Bangalore, Mysore, developed, tested, and pro-
duced two hundred 12-bladed fan-type windmills which demonstrate the
feasibility of using wind power to pump water to South India (ref. 6).
Several types of imported European and American multibladed windmills
have also been used to harness India's abundant wind energy resources.
However, due to lack of public awareness of the subject and the unavail-
ability of an even simpler and less expensive device, wind power remains
only occasionally exploited.
*
Cloth sails with a wooden framework have been used for hundreds of
75
centuries for transforming the useful energy of the wind into labor
saving mechanical work, especially grinding grain and pumping water.
The use of windmills spread from Iran in the seventh century A.D. to
coastal China where the application of the art of sailmaking signifi-
cantly improved the sophistication of windmill construction (ref. 7).
Heavy rigid wood windmill blades surfaced with cloth were increasingly
used throughout northwestern Europe so that by the seventeenth century
the Netherlands became the world's richest and most industrialized nation,
largely as a result of extensive exploitation of windpower with ships and
windmills. Cloth was a natural choice for windmill sails because of its
acceptance and wide use in sailing ships. It is lightweight, easy to
handle, readily and cheaply available, and most importantly it forms a
strong uniform surface for catching the wind when firmly supported at
three or more points.
In the Mediterranean region flour-grinding and oil-pressing mills

were rigged with six to twelve triangular cloth sails set on simple
radial spars. A three-dimensional array of guy ropes radiating from a
central spar projecting out along the axis of the main shaft suspended
the sails in position, rather than a heavy grid of wood as was used in
the traditional Dutch-type windmills. This sailboat jib type of rigging
was a significant improvement in windmill design which encouraged the
spread of windmills throughout the deforested Mediterranean countries.
The wind capturing area of these windmills was controlled by wrapping
each cloth sail around its spar. Though requiring daily rigging adjust-
ments and occasional replacement of tattered sails, the efficiency and
simplicity of these windmills resulted in their widespread use in Rhodes,
the Black Sea coast, the Aegean Islands, and Greece. In Portugal their
use was accompanied by the sound of whistles attached to the rigging, an
audible indicator of the wind at work. In the West Indies large sailing
windmills were commonly used for crushing sugar cane (ref. 8). Many
handcrafted windmills with eight triangular jib sails are presently
pumping irrigation water in the Plain of Lassithi, Crete (ref. 9). In
Japan four-bladed jib sail windmills are used to operate reciprocating
pumps which supply water to vegetable gardens. A high-speed aerodynamic,
two-bladed sail wing is being developed (refs. 10 and 11) . Further con-
struction simplifications may make it applicable to use in lesser
developed countries.
A windmill with four self-adjusting cloth sails was developed for

rural markets in less industrialized regions (ref. 12). Its relatively
complex design is limited because of the difficulty in connecting it to
a deep well pump. Unfortunately, it cannot be manufactured by hand using
local materials. Those people who are in a situation to most benefit
from a windmill are also those least able to pay for it. If the critical
moving parts were separately available, a small farmer could purchase the
remaining materials needed and assemble the windmill in his own village
using local skills and labor. This way a major portion of the money
spent would remain in the village.
The 8-meter-diameter prototype sail wing windmill recently erected
76
on a small peanut and sesame farm in a dry hilly region in South India
lifts 300 pounds to a height of 20 feet in 1 minute in a 10 mph wind.
This is accomplished by a rope passing over a 6-inch pulley on the main
drive shaft. This lift is used to lift soil and rock from the well being
hand dug below the windmill. The windmill will be set up to operate a
modified paternoster or chain pump like those used to drain mines in
England many years ago. Recently chain pumps have been rapidly replacing
the traditional square-pallet pump and the noria water lifting wheel
throughout China. A chain pump, easily and cheaply built, is more effi-
cient than most types of pumps. Most importantly, it operates well with
a low-speed, variable power source.
This sail wing windmill is made of a 1-meter-diameter bullock cart

wheel to which three bamboo poles are lashed in a triangular pattern
jA/ith overlapping ends. Each bamboo pole forms the leading edge of a
wing, and a nylon cord stretched from the outer tip of the pole to the
rim of the wheel forms the trailing edge. A stable and lightweight air-
foil results from stretching a long narrow triangular cloth sail over
that bamboo-nylon frame. This wing configuration, a hybrid of low-speed
eight-bladed Cretan sail wings and high-speed two-bladed aerodynamic sail
wings, produces high starting torque at low wind speeds. The bullock
cart wheel is attached at the hub to the end of an automobile axle shaft
which rotates in two sets of ball bearings. The shaft and bearing assembly
is mounted horizontally on top of a turntable. The turntable consists of
two circular steel plates separated with a raceway of ballbearings and held
together with a ring of eight bolts which encircle the bottom plate. A
1-foot diameter hole through the center of the turntable will allow the
chain and gaskets of the chain pump to go up and around the "squirrel
cage," which is mounted at the center of the automobile axle. If a recip-
rocating, deep-well piston pump were desired, the reciprocating rod,
rather than a chain, would go through this hole and the crankshaft rather
than an axle shaft would be mounted on top of the turntable. Since the
blades have a slight built-in coning effect and the axle or crankshaft is
mounted slightly off center from the centerline of the turntable, the
blades act as their own tail, trailing in the wind. Because the blades are
downwind from the tower, there is no danger of the bamboo poles bending in
a monsoon wind and hitting the tower. The tower is made of five 25-foot
long teak poles set in concrete at the base and bolted at the top to five
angle irons welded at a slight flaring angle to the bottom of the turn-
table. The tower tapers in towards the turntable at the top from a 7-foot
diameter at the base. It has cross bracing and a ladder.
It is hoped that other persons will continue to refine and adapt this
windmill to their own needs and materials. Please send all inquiries,
operating experience, and suggestions for improvement to: Marcus M. Sherman,
New Alchemy Institute—East, Box 432, Woods Hole, Mass. 02543.
REFERENCES
1. Eccli, Eugene: Alternate Sources of Energy Magazine, Mar. 1973,

page 17. Box 73, Kingston, New York 12401.
77
2. Sherman, Marcus M.: A Sail Wing Windmill in India. J. New Alchemists,
New Alchemy Institute, Woods Hole, Mass., 1973.
3. Neermal, T. H.: Indian Institute of Agricultural Research, Pusa

Institute, Machinery Development, New Delhi: private communication.
L|. Micheal, A. M. : Indian Institute of Agricultural Research, Pusa

Institute, Water Technology Center, New Delhi: private communica-
tion.
5. Gupta, Sen: National Aeronautical Laboratory, Wind Power Division,

Bangalore, India: private communication.
6. Harness the Wind with the WP-2 Windmill. Published by Wind Power
Division, National Aeronautical Laboratory, Bangalore, India.
7. Needham, Joseph: Science and Civilization in China, vol. IV,

Mechanical Engineering. Cambridge University Press, Cambridge,
England, 1965.
8. Chilcott, R. E.: Implications of the Utilization of Wiqd Power for

the Development of Small Caribbean Communities. Brace Research
Institute, McGill University, Montreal, Pub. No. R25. 1971.
9. Reynolds, John: Windmills and Watermills. Praeger Pu&iisjiers,

New York. 1970. ^--\
10. Kidd, Stephen and Garr, Douglas: Electric Power from the Wind.
Popular Science Magazine, Nov. 1972.
11. Fink, Donald E.: New Air Foil Design Method Developed. Aviation
Week and Space Technology, Nov. 1972.
12. Proceedings of the United Nations Conference on New Sources of

Energy, Rome, Aug. 1971, vol. 7, Wind Power. United Nations, 1964.
FURTHER READING
Merriam, Marshal F.: Is There a Place for the Windmill in the Less
Developed Countries? Working Paper Series No. 20, Technology and
Development Institute, East-West Center, Honolulu, 1972. Also:
Windmills for Less Developed Countries. J. Intern. Division of the
American Society for Engineering Education, TECHNOS, Apr.-Jun. 1972.
Merriam, Marshal F.: Decentralized Power Sources for Developing

Countries. International Development Review, vol. 14 no. 4, 1972.
Also: Working Paper Series No. 19, Technology and Development
Institute, East-West Center, Honolulu, 1972.
78
THE SAIL WING WINDMILL
AND ITS ADAPTATION FOR USE
IN RURAL INDIA
Figure 1
79
ROTOR DYNAMIC CONSIDERATIONS FOR LARGE
WIND POWER GENERATOR SYSTEMS
Robert A. Ormiston
U. S. Army Air Mobility R&D Laboratory

Moffett Field, California
Practical large-scale wind power generating systems must be competi-

tive in terms of energy cost. Low available wind energy per unit area
demands efficient aerodynamic design, a location with high mean wind
velocity, and proper integration with established power grids. Also es-
sential for low initial and operating costs are reliable, lightweight,
mechanically simple designs requiring a minimum of maintenance. To a
large extent, these qualities depend on the dynamic loads and vibratory
stresses of the rotor/tower structure that, for very large wind turbines,
will probably constitute the ultimate design constraints.
In many respects the dynamic properties of large wind turbines will

be similar to helicopter rotors where cost, reliability, vibration, etc.,
are also of primary importance. Although some differences exist, much of
todayTs helicopter rotor technology is applicable to the design of wind
power systems. Based partly on this experience, the following comments
are offered to provide some understanding of the dynamic properties of
large wind turbines and suggest some possible design concepts.
DETERMINATION OF VIBRATORY LOADS
Vibratory loads and stresses result from unsteady aerodynamic,

inertial, and gravitational forces which act on the rotor/tower struc-
ture. If this structure is ideally rigid, vibratory loads and stresses
can be easily determined from known applied forces. For real flexible
structures, elastic deformations contribute additional aerodynamic and
dynamic forces, and the determination of vibratory stresses is consider-
ably more difficult. If exciting forces occur at frequencies near the
natural frequencies of the structure, resonance may seriously amplify
dynamic loads.
Principal Structural Deformations
The elastic deformations of the rotor blades and tower structure

are shown schematically in figure 1. In this example the blades are
80
shown cantilevered to the rotor hub, but similar deformations occur for
articulated (hinged) blades. Structural deformations include flap and
lead-lag bending of the blade perpendicular and parallel to the plane of
rotation, respectively, blade torsion or elastic twist, vertical and
horizontal bending of the rotor shaft (not shown in fig. 1) , and bending
and torsion of the tower structure. The importance of these elastic
deformations will be dependent on the degree of flexibility of the rotor/
tower structure.
Rotor Blade Frequencies
The vibratory loads and stresses of a rotor system depend to a large
extent on the natural frequencies of the structure. Some understanding
of the dynamics of a single rotor blade can be obtained from the linear
bending-torsion equations (ref. 1) that determine the rotating natural
frequencies .
Flap: -(Twf) + El w"" + mw = L
z
— y
Lead-lag: -(Tv_L)' + El zv"" + m(v - fl2y) = Ly
Torsion: -GJ<(." - k2 /i (T<J>')' + mk2 HI * + mfi2|~k

I
2
Jllrt
- k2 ~U = M T
T I
o
where T f = -mn x.
Flap and lead-lag deflections are given by w and v, respectively,
and torsional deflection by <j> (See fig. 1.) The effects of centrifu-
gal tension and stiffening due to rotational velocity n are under-
lined. The remaining terms are due to bending stiffness (El) or torsional
rigidity (GJ) and inertial forces due to blade mass m. The forces and
moment L , L , and M, applied to the blade are caused by aerodynamic,
inert ia 1,z and^gravitational forces. When these applied forces are not
retained, the homogeneous equations define the blade natural frequencies
and mode shapes. Rotor-blade frequencies are typically displayed in
dimensionless form as a function of the normalized rotor speed. A typical
example is sketched in figure 2. The frequencies and rotor speed are
normalized by the nominal or rated operating speed ft . The frequencies
correspond to the fundamental and higher modes of bending and torsional
deformations and they increase with rotor speed because of centrifugal
stiffening. Also shown on this plot are frequencies of the applied blade
forces which occur at integer multiple harmonics of the rotor speed (such
as one per revolution, twice per revolution, ..., or, IP, 2P, ..., for
short) . These applied forces will exist whenever the rotor blade is not
uniformly loaded around the azimuth, for example, nonaxial wind components,
gravity forces, rotor disk tilt, or shaft precession. Generally, the ap-
plied forces diminish with increasing harmonic number.
The significance of this figure is that resonance and severe vibra-
tory stresses may occur when a blade natural frequency is close to the
frequency of an applied force. Therefore, the rotor blade must be de-
signed to avoid such resonances to achieve low fatigue stresses and long
life. One difficulty is that during operation below rated speed, or with
81
ungoverned wind turbines, it is virtually impossible to avoid a resonance
at some speed. This may preclude operation at that speed if severe vibra-
tory stresses result.
DYNAMICS OF LARGE ROTORS
The importance of flexibility for vibratory loads and stresses de-
pends partly on the degree of flexibility of the structure. Therefore,
the proper questions in discussing wind turbines are:
(1) what parameter best characterizes blade flexibility, and
(2) how does this parameter vary as a function of rotor size?
Perhaps the most appropriate parameter is the dimensionless fundamental
blade natural frequency, which depends on the ratio of blade bending
stiffness to centrifugal forces - g M/Q
This parameter establishes the condition for dynamic similarity for .
a wide variety of rotor blades having large differences in size, stiffness,
mass, and rotational speed. It does not, however, account for gravitational
forces. The blade natural frequencies are also a good measure of the
importance of flexibility on dynamic loads. For very high stiffness or
frequency, only low-energy, higher integer harmonic forces will be avail-
able to cause resonant vibratory stresses. The low frequency forces will
then act on the structure much as static loads. Lowering the blade stiff-
ness and frequencies will tend to relieve high "static" loads but will
increase the importance of dynamic response.
It is interesting to compare expected wind turbine blade fundamental
flap and lead-lag frequencies with conventional rotor and propeller blade
frequencies as shown in figure 3. The conventional fully articulated
rotor has very low fundamental frequencies because of the blade hinges.
The teetering helicopter rotor with a single hinge has a low flap frequency
and a moderately high lead-lag frequency. Other systems include the canti-
levered hingeless helicopter rotors and conventional propellers which are
relatively stiff. Structural information for large wind turbines are
nonexistent and therefore only estimated frequency values can be shown.
The lead-lag bending frequency is assumed relatively high in view of the
typical low operating speeds of wind turbines, and the need to stiffen
large rotor blades against gravitational stresses. Three possible wind
turbine configurations are shown:
(1) a teetering or coning hinge design to relieve aerodynamic
thrust and hub moments,
(2) a hingeless design to relieve blade root stresses with
elastic flap bending, and
(3) a stiff design to withstand aerodynamic loads directly.
These fundamental frequency values must be more precisely known before it
will be possible to accurately compare the dynamic load characteristics
of large wind turbines with other types of rotor systems.
82
Scaling Effects for Large Rotors
Sizing trends for wind turbine properties may be deduced from

dimensional analysis considerations. For purposes of comparison, a
constant level of aerodynamic efficiency at a given wind speed is assumed
which in turn constrains the blade tip speed Rft to a constant value.
For geometrically similar construction then, the rotor parameters will
vary with size, or radius R, in the following manner:
Parameter Proportional to
Rotor speed R~^

Blade mass R^
Centrifugal force R^
Solidity R°
Power ' R2
Thrust R
Centrifugal stress R
Aerodynamic stress R
Gravitational stress R
Dimensional natural frequencies R
These relationships show that the important aerodynamic and centrifugal
stresses are independent of rotor size, but that gravity stresses in-
crease in proportion to the radius. The dimensionless natural frequencies
remain constant, however, which means that dynamic response and resonant
characteristics will not be influenced by rotor size. Interestingly, the
power output increases with the square of the radius but blade weight in-
creases with the cube. This is an example of the "square-cube law" that
will eventually limit wind turbine size because of diminished power to
weight ratio. This factor as well as aerodynamic efficiency trade-offs
will alter the ground rules of geometric similarity and constant
tip speed as a basis for establishing trends for dynamic properties of
large rotors. For example, power losses due to aerodynamic profile drag
can be reduced by increasing the rotor solidity and reducing tip speed,
but only at the expense of increased blade weight and cost. Improvements
in airfoil lift/drag ratio will permit reduced solidity and higher tip
speeds. Increased tip speeds would be advantageous for reducing the
capacity of the speed-increaser gear needed to step-up the low rotor
shaft speed to the electrical generator speed. Although these trade-offs
are complex, it will probably be necessary to sacrifice some aerodynamic
efficiency to reduce blade size and weight of large wind turbines.
Therefore, thinner blades operating at higher tip speeds will tend to
reduce the dimensionless natural frequencies and so increase dynamic
response and the effects of flexibility for large rotors. And, inevita-
bly, gravitational stresses will be important for large wind turbines.
ROTOR CONFIGURATIONS
The choice of a specific rotor configuration can strongly influence
83
the mechanical complexity, vibratory stresses, reliability, and mainte-
nance cost of wind turbines. Therefore, the attributes of different
rotor concepts must be carefully weighed. Important configuration
properties include the number of blades, blade to hub articulation,
pitch control mechanisms, etc. Before discussing the dynamic character-
istics of several rotor systems, the various forces contributing to blade
vibratory stresses will be described.
Rotor Blade Forces
These include aerodynamic, inertial, and gravity forces. The major

aerodynamic loads are generated by the unsteady nonuniform wind environ-
ment. The mean axial wind component generates thrust forces which de-
flect the blades equally downwind (coning). Gradients in axial velocity
(the vertical gradient of the ground boundary layer for instance) produce
a tilting of the rotor disk with respect to the shaft. Velocity components
perpendicular to the rotor axis also produce disk tilting as well as higher
harmonic loadings. Nonuniformities in velocity peculiar to the wind tur-
bine location, the tower wake, and atmospheric turbulence will produce
important unsteady loads•. Additional unsteady loads, though probably small,
will be induced by the rotor wake vorticity that itself results fôm un-
steady blade loadings. Inertial blade loads include centrifugal tension
due to rotation, lead-lag, and flapping loads due to Coriolis forces
arising from blade oscillations, and gyroscopic forces due to precession
of the rotor shaft to maintain alinement with the wind velocity vector.
The primary lead-lag Coriolis loads result from tilting of the rotor disk
(flapping deflections). Finally, gravity loads may produce significant
lead-lag bending stresses for large rotors as noted above.
Hub Configurations
The importance of minimizing cost by reducing mechanical complexity

favors the use of a minimum number of blades and the elimination of
unnecessary articulation (blade attachment hinges) at the hub. Usually,
however, some articulation is required to reduce blade stresses, due to
aerodynamic loads, and blade flapping motion. Several types of rotor
hubs found on helicopters are shown in figure 4. The simplest two-
bladed teetering rotor is typical of current helicopters and allows simple
flapping freedom (disk tilting) to relieve IP aerodynamic loads. The
teetering hub does not provide individual blade flapping (coning) to re-
lieve thrust forces due to wind gusts. The coning hub relieves these
forces with an additional hinge, but these hinges must support the full
centrifugal force load as well as the lead-lag bending moments. A hub
configuration found on helicopters with three or more blades eliminates
individual blade flapping hinges by attaching the blades to a gimbaled
hub. This would not provide coning freedom for a wind turbine, but it
does provide relief for inertial lead-lag bending moments. The common
fully articulated helicopter rotor hub provides nearly complete relief
for the major blade loads by using individual flap and lead-lag hinges
for each rotor blade. However, this system is complex and, again, the
hinges must carry the full centrifugal load of the blade. The last
configuration shown in figure 4 is the hingeless rotor system in which
84
hinges are replaced by flexible spars that deflect elastically to re-
lieve applied loads. Because flap bending moments are transmitted to
the shaft, two bladed hingeless rotors would not be practical unless the
tower structure could withstand 2P vibratory hub moments due to rotor
disk tilting. With three or more blades, only steady hub moments due to
disk tilting are transmitted to the shaft. Although hinged rotor hubs
are currently in wide use, the hingeless rotor has definite advantages in
terms of reduced complexity and improved reliability; and with continued
development it is gaining acceptance for helicopter applications. Suit-
ability of the hingeless rotor for wind turbines remains to be established
and would require detail design and feasibility studies aimed particularly
at reducing vibratory loads and stresses. Especially attractive is the
use of molded composite materials to reduce fabrication costs and optimize
the blade structural properties.
Each of the rotor hubs in figure U requires additional bearings to

permit blade pitch changes for regulating power and feathering the rotor
in extreme wind conditions. A possible extension of the hingeless rotor
concept might permit the elimination of the pitch change bearing for maxi-
mum simplicity. A conceptual sketch in figure 5 shows the flexible canti-
levered spar with bending flexibility to relieve stresses and with tor-
sional flexibility accommodating blade pitch changes. This twisting may
not be sufficient to fully feather the blade, however, and alternate
means for dealing with extreme wind velocities might be necessary.
AEROELASTIC STABILITY
Practical rotor systems, including both helicopter rotors and wind

turbines, must avoid aeroelastic instabilities. These may stem from
several different but related physical mechanisms. Perhaps the best
known is classical bending torsion flutter encountered on fixed-wing air-
craft. This type of flutter occurs when relatively high-frequency un-
steady aerodynamic forces couple with the elastic flap bending and torsion
of the rotor blade to produce negatively damped oscillations. This is
generally precluded by proper mass balance. Another type of instability,
although less well known, can occur for cantilever (hingeless) rotor blade
configurations. This type of instability involves both flap and lead-lag
elastic bending as well as torsional deformations. It is primarily due to
the strong structural coupling between bending and torsion that is charac-
teristic of cantilevered rotor blades and can be avoided by proper tailoring
of the bending and torsional stiffness distributions. A typical example of
stability boundaries for a helicopter rotor of this type operating in hov-
ering flight is shown in figure 6. These boundaries show that instability
will be encountered above a certain pitch_ angle 9 for configurations
having various dimensionless torsional u). and lead-lag tuv natuaral
frequencies. This figure, taken from reference 1, is only indirectly
representative of specific wind turbine configurations. But it does in-
dicate that aeroelastic stability should be considered in the design of
large rotor systems.
85
CONCLUSIONS
Successful large, reliable, low-maintenance wind turbines must be
designed with full consideration for minimizing dynamic response to aero-
dynamic, inertial, and gravitational forces. Much of existing helicopter
rotor technology is applicable to this problem. Compared with helicopter
rotors, large wind turbines are likely to be relatively less flexible
with higher dimensionless natural frequencies. For very large wind tur-
bines, low power output per unit weight and stresses due to gravitational
forces will be limiting factors. The need to reduce rotor complexity to
a minimum favors the use of cantilevered (hingeless) rotor configurations
where stresses are relieved by elastic deformations.
REFERENCES
1. Hodges, D. H. and Ormiston, R. A.: Stability of Elastic Bending and

Torsion of Uniform Cantilevered Rotor Blades in Hover. Paper No.
73-405, AIAA, Mar. 1973.
DISCUSSION
Q: You say a helicopter rotor is apparently designed for much higher

frequency?
A: Yes. You couldn't take a helicopter rotor and turn it sideways and
have an efficient wind turbine. In principle there are many simi-
larities, but in terms of detail characteristics there are many
differences.
Q: I have one question and one comment. You did not in your presenta-
tion state the difference of material. Materials have natural
frequencies. The natural frequency is divided by the density. Per-
haps this will be used to evaluate rotors. Did you investigate this?
A: Well, we haven't done any work on that, but I have made that point
in my written comments. There is a tremendous potential for using
glass fiber components or epoxies or whatever kind of molding materials,
to tailor not only the aerodynamic configuration but the structural
characteristics as well. This is extremely important in terms of
aeroelastic characteristics, the blade frequency, vibration, and so
forth. It's a tremendous potential for a rotor, any type of rotor.
And it makes for much simpler construction. I think the work that
Professor Hutter has done is a good example of that, and from what
I've seen it looks a very good way to go.
The hingeless rotor I mentioned has no bearings or hinges, is made

with composite materials and, is quite simple. They are nothing like
typical rotors that we have nowadays.
COMMENT: There are problems of vibration, that would increase beyond

limits we have so far heard on the tests. Therefore, I see that vibra-
tion would be most serious in the development of some windmills. It
would be very difficult to erect these machines because a vertical rotor
can't have downhanging rotors.
86
A: I think you are referring to gravitational loads under static condi-
tions? Yes, that's going to be a problem. You don't have to go too
much higher in size before just the static deflections get to be a
problem.
In helicopter rotors, the plane of the rotor is normal to the gravity

field so the blades all drop evenly. The rotor can start up and gain
centrifugal stiffening, and the blades can be made much more flexible,
But for wind turbines, they have to start vertically. It could be a
real problem.
BLADE
DEFLECTIONS
LEAD US
TOWER 9P 8P
BENDING AND TORSION lORSIOH 7P
INTEGER MULTIPLE
EXCITATION
FREQUENCIES
NORMALIZED ROTOR SPEED,
Fig. 1 - Elastic deformations of Fig. 2 - Variation of rotor blade

rotor/tower structure. natural frequencies and excitation
frequencies with rotor speed.
87
TEETERING CONING GIMBALEO
2.5 IIHO b'2 b22 b>2
TURBINES
2.0 /TEETERBie
< ROTORS
CONVENTIONAL
1.5 PROPELLERS
STIFF INPLANE
IX) HINSELESS ROTORS
ja SOFTINPLANE
o .5 U HINGELESS ROTORS
2 a?
I ARTICULATED
' ROTORS
1.0 15 2.0
FLAP FREQUENCY, ww/0
Fig. 3 - Comparison of rotor systems Fig. 4 - Typical rotor hub configu-

according to dimensionless bending rations .
frequencies.
•o
o
r3
UJ
X 9
0-2
UNSTABLE
FLEXURE ELEMENT *"•"({?{_
BENDING FOR LOAD RELIEF STABLE
TORSION FOR BLADE PITCH CHANGES
24 6 8 10 12 14 16 18 20
NONROTATING LEAD-LAG FREQUENCY. w»NR
Fig. 5 - A hingeless rotor concept of Fig. 6 - Aeroelastlc stability of

simplified design. hingeless helicopter rotor blades
in hover.
THE EFFECT OF AERODYNAMIC PARAMETERS ON
POWER OUTPUT OF WINDMILLS
W. Wiesner
The Boeing Vertol Company

Philadelphia, Pennsylvania
SUMMARY
This paper reports the aerodynamic results of a study on wind power

generation the Boeing Vertol Company has conducted during this past year.
Windmill power output is presented in terms that are commonly used in
rotary wing analysis, namely, power output as a function of drag developed
by the windmill. Effect of tip speed ratio, solidity, twist, wind angle,
blade setting and airfoil characteristics are given.
INTRODUCTION
Several papers have been written on the performance of windmills,

but none, to the author's knowledge, have systematically examined the
effects on power output of the various aerodynamic parameters. Therefore,
the Boeing Vertol Company formulated a theory and computer program to
accept the various parameters that can be studied such as tip speed ratio,
solidity, twist, wind angle, blade angle, airfoil characteristics, etc.
The purpose of this paper is to present the effect of such variations on
power output of a windmill. All combinations possible are not the intent
of this study, but it is believed that an initial understanding of how
these parameters effect power can be obtained from the data herein,
METHOD OF ANALYSIS
The program used is our SR1BR Rotor Research Program which calcu-
lates 10 points along the blade at 12 evenly spaced positions of rota-
tion and sums the individual points to give the usual rotor parameters
such as thrust, drag, lift and power. The program is set up to calculate
induced velocity based on the disk loading at the particular point of
calculation and thus uses a nonuniform downwash program. The data re-
sulting from these calculations are all referenced to the product of
wind dynamic pressure, wind velocity, and windmill rotor solidity ratio.
SYMBOLS
d diameter of rotor, ft
P power output
89
q wind dynamic pressure, (1/2) pV
R blade radius, ft
V wind velocity, ft/sec
X rotor drag, Ib
ds stall angle of airfoil
ccun •on
*?uaerodynamic angle of attack at blade element at 0.90 blade radius
CD profile drag coefficient of blade at blade angle of attack = 0°
o
6^ blade twist (linear)
6Un • /nnblade
U
incidence at 0.70 blade radius
P air mass density
cr solidity, ratio of blade area to disk area
p tip speed ratio, ratio of wind velocity to rotational tip

velocity of rotor (V/WR)
u angular velocity of rotor
RESULTS
Comparison with Test Results
Figure 1 shows that the SR1BR Program calculations compare very

favorably with the power output measured values published by the Brace
Research Institute for their 32-foot-diameter, three bladed windmill.
Effect of Blade Angle
Figure 2 shows how power output P and drag X vary with blade
angle at 0.7 of the blade radius for the baseline windmill chosen for
this paper. This baseline1 performance is based on a blade airfoil sec-
tion that stalls at a 14° angle of attack, a tip speed ratio of 0.30,
and a solidity ratio of 0.20. It will be noted that there are two values
of blade pitch where the power is zero. Point (A) is where the relative
velocity is in line with the blade elements. Point (B) is where the blade
elements are fully stalled as will be noted by the value of angle of attack
J-Baseline parameters were arbitrarily established for example only.

They should not be accepted as optimum parameters. Future studies should
be conducted to determine optimum aerodynamic parameters for specific
applications.
90
at the blade tip of 19.3°, which is beyond the 14° stall angle of the
airfoil. The maximum power output P occurs at a blade angle of 16°,
but the important item to note is that at maximum power the blade angle
of attack at the 0.90 radius point is 11.4° - about 3° less than the
stall angle. Point (C) is maximum efficiency point (that where the
power for a given drag is greatest) . However, windmills should operate
at Point (D) because that is where the power output is greatest. The
tower structure must be designed for the drag at Point (B), unless great
care is expended in developing a drag limiting governor.
Effect of Blade Twist
Figure 3 shows that, for the baseline windmill operating at a tip

speed ratio of 0.30 with solidity of 0.20, the best linear blade twist
is 30°. Such twist will produce 25 percent more power than an untwisted
blade. It should be noted that optimum twist^ will vary with tip speed
ratio.
Effect of Solidity
Figure 4 shows that power output normalized to (qVd^c) decreases

with solidity, but figure 5 (based on the maximum power line of fig. 4)
shows that power output is maximum at a solidity between 0.20 and 0(. 40
for the baseline conditions of e - 30°^ o = 14°? and y - 0.30.
L S
Effect of Wind Velocity
Figure 6 shows drag and horsepower output for a 100-foot-diameter

windmill using the Points (A) and (B) design values from figure 5 for
solidity =0.20. It will be noted that a 45-foot per second wind will
generate 800 horsepower or 0.10 horsepower per square foot of rotor
swept area.
Effect of Airfoil Section
Figure 7 shows that if the airfoil stall angle is doubled without

any other changes, that the power output increase over the baseline con-
figuration with «„ = 14° is 20 percent while the windmill drag force
X is increased 25 percent at maximum power output. The maximum drag
that can be developed is almost doubled so windmills with high life air-
foils will necessarily have to be provided with stronger towers at an
increase in cost.
Effect of Wind Angle

Figure 8 shows that output power decreases rapidly if the wind
angle with respect to the shaft is very much greater than 12°. Thus,
windmills must be provided (as they have through the ages) with some
means to point into the wind for maximum power generation.
shown in this paper.
91
Effect of Tip Speed Ratio
Figure 9 shows that power output is dependent on tip speed ratio

and that for each operating tip speed ratio there will be an optimum
solidity to produce maximum power output. Boeing studies are continuing
in the effort to quantify such effects.
CONCLUDING REMARKS
(1) Boeing SR1BR Program gives results that compare favorably with
test data.
(2) Maximum power output of windmill occurs when a blade-element

angle of attack near tip of blade is about 3° less than airfoil stall
angle for the baseline case considered.
(3) Selection of the proper blade twist will increase power

output.
(4) For a given blade twist, airfoil and tip speed ratio, there
is a range of solidity ratios that will produce nearly the same power.
(5) Power output varies approximately as the cube of the wind

velocity. Drag of the windmill varies as the square of the velocity
(6) Higher lift airfoils for windmills will increase the power
output, but greatly increase the design drag value for tower design.
(7) Windmills should be designed such that shaft should point

into wind with tolerance no greater than ^12° for good power output.
(8) Solidity will vary with operating tip speed ratio to produce
maximum power output.
DISCUSSION
Q: In your computer program can you use that blade element theory?
A: Yes, it is very similar to the vortex theory or ones that are
developed on helical vortex analysis.
COMMENT: Certainly you have the solution where you have the slow
rotating shaft speed into a high speed by putting some small rotors on
the tip of the blades. This is the only position of application where
you have rated the towing efficients. If you have a fixed wing on an
aircraft, you want to give out from the resistance or the drag of the
wind the maximum power. This is your efficiency. This efficiency is in
anqther sense involving the ordinary windmill. But the towing efficiency
is important for the transformation of speed.
OF
E.PRECT OF
)n*5rrnrre. TO
fee BS
sc
4C
30
lo
10 zo 30
R&. 3 £ FFe cr OF
OF SOUO\TN
93
EFFECT or Sou PITS ON
0
ttl 1.0
.8
.C,
o/
p
o/
til
I .Z.
.10 .20 .30

_g-
AMD
4.
b
1
b
01.
10 ZO 30 -=VO SO 60 70
OF T\f>
. 8 EFFSCT OF V/IHD SPS.g.0
OF
,s
•p
tfd-S
O IO 20 30 40 SO
.a
V/iwo Ancur VI»TH TiPSpeeo J M.
95
VERTICAL AXIS WIND ROTORS - STATUS AND POTENTIAL
W. Vance
Advanced Concepts Division
Science Applications, Inc.
La Jolla, California
Except for a rather inventive period in the 1920's, the approach

taken to extracting power from the wind has been that of using blades
or vanes rotating about a horizontal axis with the plane of the blades
essentially perpendicular to the wind velocity vector. The two devices
shown in figure 1 were patented almost 50 years ago and have as a com-
mon element a vertical axis of rotation. The first device, patented by
Msr. G. J. M. Darrieus in 1931, has received some recent study by the
National Aeronautical Establishment in Canada; the second was developed
and patented by Mr. S. J. Savonius in 1929. These rotors share a per-
formance characteristic which differentiates them from the horizontal
axis wind rotors, namely, their ability to operate equally well regard-
less of the direction of the wind. This characteristic is important
because it permits the rotor to extract the energy of a given wind or
gust instantaneously regardless of any rapid changes in wind direction.
Considering that the energy available from the wind is proportional to
the cube of the velocity, the feature of not having to take time to head
the machine into the wind may well provide additional energy extraction
capability over that of a horizontal axis rotor. It is also likely that
the elimination of a heading control and servosystem will tend to reduce
acquisition and maintenance costs and improve reliability.
Although a number of applications were developed for the vertical-

axis rotor, the concept never became popular. Horizontal axis machines
were improved over the years and have received substantial attention,
perhaps largely due to the availability and advance of propeller theory.
We believe that the time is right to take a hard look at the vertical
axis machines to see if recent advances in aerodynamics, structures, and
materials technology might not place these concepts individually (or
perhaps in combination) in a favorable light in comparison with the
horizontal axis wind rotors.
To maintain brevity, we will concentrate on the S-rotor for the

remainder of the presentation. The configuration of the original S-rotor
shown in figure 2 resulted from some 30 or more wind tunnel and field
tests conducted by Savonius wherein he varied some of the parameters of
the rotor.
Essentially, the device operates (at least during part of its
96
rotation) as a two stage turbine wherein the wind impinging on the concave
side is circulated through the center of the rotor to the back of the con-
vex side, thus decreasing what might otherwise be a high negative pressure
region. The flow is indicated in figure 2.
Savonius applied his wind rotor to water pumps, ship propulsion, and
building ventilators, all with some success. In addition, he also showed
the feasibility of using the energy in ocean waves to drive the rotor.
This last application was developed subsequentially as an ocean current
meter and is available commercially. Very good current measurement capa-
bility exists in a region of from 0.05 to 5 knots.
In reviewing the work that has been done on vertical axis rotors, we
have concluded that there are a number of development alternatives that
should receive some attention from the standpoint of both test and analysis.
Figure 3 indicates some of these alternatives. The effects of aspect ratio
(the ratio of rotor height to diameter) and the number of vanes will be
discussed in detail below. The issue of the profile of the rotor has not
been investigated, at least in terms of large (50 ft high or greater)
machines. Questions have arisen concerning whether more of the area of
the rotor should be at the top to catch the higher wind speeds or whether
the area should be at the bottom to provide a more uniform torque distri-
bution along the height. The rotor camber and thickness distribution also
need to be optimized. Our own limited amount of test data have indicated
that the amount of venting between the rotor vanes has a very significant
effect on the rotor speed for a given wind speed.
Figure it presents some of the results of a preliminary analysis of

the impact of rotor aspect ratio on rotor acceleration. Most of the
rotors in use have relatively low aspect ratios (refs. 1 to 3). If we
look at the rotor's ability to accelerate as defined by the ratio of the
torque on the rotor to its polar inertia, it can be shown that this char-
acteristic improves in proportion to the square root of the aspect ratio
as aspect ratio increases. Clearly, there must be limits to this trend
due to structural or other considerations. Furthermore, constant-speed
performance may impose other requirements.
Test data are shown in figure 5, which indicates the static torque
obtained for the two- and three-vaned rotors shown as a function of wind
direction. The torque diagram for the two vaned S-rotor has a consider-
able irregularity that, could make it difficult to start under some orien-
tations. The addition of the third vane smoothes the torque diagram to
some degree and apparently increases the torque per revolution, but also
increases the polar inertia of the rotor, which may offset the increased
torque when starting under low wind conditions. Whether two or three
vanes will be optimum remains to be resolved. It is also likely that the
torque diagram for a rotating rotor may be considerably different from
that of the static case described.
The S-rotor may be located in any area where a horizontal axis rotor
might be sited. However, the nondirectionality of the S-rotor may be put
to use more effectively on sea coasts where the diurnal variation of the
97
wind could be readily accepted. In considering this basic application,
it occurred to us that it might be possible to generate an artificial
on-shore breeze through the appropriate use of solar energy in the desert.
Figure 6 shows a concept of such an artifice. A set of S-rotors are
placed circumferentially around a circular area whose surface is made
such that the air over it is heated to a higher temperature than the air
outside of it. A flow will be established from outside of the heated
area to replace the rising heated air. By locating the rotors in the
throats of suitably contoured areas, it may be possible to extract con-
siderable energy from the resulting accelerated air. It is recognized
that this.is an ambitious concept. In essence, we are trying to produce
our own wind in sufficient quantities to make a cost-effective power
system. Analysis and test techniques must be developed to verify the
feasibility of this system concept.
Another application of S-rotor might be -in remote areas such as the

one depicted in figure 6. In the Arctic and many other places in the
world, empty oil drums might be used for rotor vanes. In some underde-
veloped countries it may be possible to construct the rotor from indigenous
materials. The actual siting of the rotor in a village or base camp would
depend on knowing where strong winds persisted without regard to their
direction. This simplification coupled with low costs (for the rotor)
might make the S-rotor a valuable asset to the community. It should also
be noted that vertical axis rotors might be of considerable value in meet-
ing instrumentation and power needs for research on the surface of other
planets.
In conclusion, we believe that the potential of vertical axis rotors

has not been exploited in recent years and that a comprehensive program
including design, analysis, and test could yield devices of equal (if not
better) cost-effective.performance than that of horizontal axis rotors.
We further believe that applications of these rotors should be considered
simultaneously with their development to ensure the practical utility of
the wind machines.
DISCUSSION
COMMENT: There are two units currently being manufactured in Switzerland

which use this principle. One is a 50 watt unit; the other is a 250 watt
unit. The design is a slight variation of the Savonius rotor principle.
The larger unit has been in production now for about 5 years. They are
being used very successfully, particularly on top of radio towers where
a little power is needed for a booster amplifier to take the signal down
to the house. They worked very well for that purpose.
COMMENT: The company, Electro GMBH Company, also produces the 6-kilowatt
standard generators. The man that runs that company is very interested
in vertical axis design. He has experimented a lot with them. He has
come up with, it seems, a quite successful unit for a small-scale, very
simple in design and virtually no maintenance whatsoever, and no prob-
lems with regulations in high wind, and so on. There has been one built
98
as a matter of fact, out in the Scripps Institute of Oceanography not too
far north of where we are.
Q: You speak of one of these machines being 100 feet high. What is the
Largest model you know about?
A: I haven't heard of any near that size. An important question is
whether this type of rotor be scaled up to larger sizes? My answer
is: I don't know.
Q: How big you have seen any size.

A: About 15 or 20 feet high.
Q: In the thirties one about 100 feet high was constructed in New
Jersey.
VERTICAL AXIS WIND ROTORS
DARRIEUS ROTOR BACKGROUND

PATENTED IN 1931 (US AND FRANCE)
CURRENTLY UNDER STUDY AT NATIONAL AERONAUTICAL
ESTABLISHMENT, OTTAWA, CANADA
CHARACTERISTICS
EFFICIENCY ~ 35%
TIP SPEED TO WIND SPEED ~ 6 TO 8
POTENTIALLY LOW CAPITAL COST
CURRENTLY NOT SELF STARTING
S-ROTOR BACKGROUND
PATENTED IN 1929 (US AND FINLAND) BY S. J. SAVONIUS
CURRENTLY USED AS AN OCEAN CURRENT METER
OTHER APPLICATIONS SHOWN FEASIBLE
CHARACTERISTICS
TIP SPEED TO WIND SPEED ~ . 8 TO 1.8
EFFICIENCY ~ 31%
SELF STARTING
VERTICAL AXIS ROTORS OPERATE INDEPENDENTLY

OF WIND DIRECTION AMD THUS HAVE A POTENTIAL FOR
HIGH EFFICIENCY IN CHANGING WINDS
FIGURE 1
S-ROTOR DEVELOPMENT
CONFIGURATION
END PLATES PROVIDE
STRUCTURAL MEMBERS
CONDITIONS FOR 2 DIMENSIONAL FLOW
COMMON ASPECT RATIOS < 3
SHEET METAL VANES
FLOW CONDITIONS
PERFORMS SOMEWHAT LIKE 2 STAGE TURBINE
FLOW TO BACK SIDE OF ADVANCING VANE
REDUCES NEGATIVE PRESSURE
SUBSTANTIALLY CONSTANT AREA FOR
AIR FLOW
DEMONSTRATED APPLICATIONS
WIND DRIVEN WATER PUMP
WIND DRIVEN SHIP PROPULSION
BUILDING VENTILATORS
OCEAN WAVE DRIVEN WATER PUMP
OCEAN CURRENT METER
FIGURE 2
DEVELOPMENT ALTERNATIVES
Number of Vanes Vane Profile
Vane Thickness Rotor Configuration
Separation
UOverlap
FIGURE 3
100
EFFECT OF ASPECT RATIO ON ROTOR ACCELERATION
ACCELERATION •> AR 1/2
Sa
© 2
0 1 2 3 4 5 6 7 8
AR
ASSUMED CONSTANTS
ROTOR AREA WIND VELOCITY

ROTOR WEIGHT BEARING FRICTION
AVERAGE LOAD 2-DIMENSIONAL
PROVIDING TORQUE FLOW
AVERAGE LEVER ARM/ROTOR RADIUS
FIGURE 4
STATIC TORQUE PROFILE
180°
Wind
18<F
270°
FIGURES
101
w ^§
o<
w
K Z
HH
i CO
O
n
HH
8 ^^
°^
S^
OT g
K *
O 2
H W
Ko g
F
CQ
K
l« K O
O
z S § fc g
2 <; 0
il
ô
gs K HK g W
AR
S.8 W j
I §s
wo
§S
Hw
Oo
K
CO
S
» 3
H
O
H
O
g£rig 5S
H
^ s o S go
w
CO CO
K
102
ADVANTAGES OF THE DIFFUSER-AUGMENTED WIND TURBINE
R. A. Oman and K. M. Foreman
Research Department
Grumman Aerospace Corporation
Bethpage, New York
The fact that substantial performance advantages can be realized by

the use of a shroud and diffuser on a wind turbine was recognized in the
1950's. The work of Lilley and Rainbird (ref. 1) indicates that diffuser-
augmented turbines can produce up to twice the power of unshrouded tur-
bines of the same diameter, and our independent analysis indicates the
same range of performance improvement. Even more attractive is the po-
tential for operational flexibility afforded by the diffuser, enabling
useful power generation at lower and higher wind velocities, and simpler
control features.
Figure 1 shows an artist's conception of a diffuser-augmented wind

turbine. The basic function of the diffuser is to convert the kinetic
energy of the flow downstream of the rotor into a pressure rise. This
lowers the pressure level behind the rotor, and makes it possible for
the rotor to capture airflow from a free stream tube area that is greater
than that of the rotor itself. The inlet area need not be large, as the
stream tubes will converge naturally to the inlet if the diffuser is
sufficiently effective. The optimum conditions for such a system are
very significantly different from the familiar ideal optimum for an un-
shrouded rotor (cf. Glauert, ref. 2) . The flow velocity through the
rotor is typically 20 to 60 percent greater than the free wind velocity
as opposed to 67 percent less than the free wind for the unshrouded case.
In addition to offering more output per unit rotor area, this fundamental
change in stream tube configuration enables practical rotor designs to
operate even at very low wind speeds. The presence of the diffuser also
offers the opportunity to accommodate to very high wind speeds without
the need for variable pitch in the rotor blades. These large performance
and operational advantages may be sufficient to overcome the cost of the
large diffuser, especially in applications for which storage is a signifi-
cant cost factor.
Our one-dimensional analysis differs from that of Lilley and Rainbird

in that the drag of the shroud does not enter explicitly into the per-
formance prediction. Figure 2 shows that stations used in the analysis,
and indicates schematically that the upstream capture area is greater than
that of the rotor, but smaller than that of the rotor stream tube far down-
stream of the diffuser exit. We define the ideal power coefficient C_ =
O i
-L
Ap 0 _V_/(l/2)pV , which through Bernoulli's equation, continuity, and a
^ O £. O
statement of losses in inlet and diffuser can be expressed as
103
C = 1
IPL - Ki -
where e = VO^Q* î ^s ê ^n-'-e"t l°ss coefficient, nn is diffuser

efficiency, Cp^ = (Pl| - pQ)/(l/2)pV^, and X = Ag/A^.
The optimum velocity ratio and the corresponding ideal power
coefficients are shown in figures 3 and 4, while the off-optimum per-
formance is shown in figure 5, all for Ki = -C ^. Because of the
very low exit velocity, viscous entra inment downstream of the exit
should make it possible to operate with slightly negative Cp 4 , in-
creasing ideal performance beyond that indicated. It is to be under-
stood that no real turbine could achieve these ideal figures, but the
relationship between shrouded and unshrouded systems should be at least
as favorable to the diffuser-augmented type as that shown. The effect
of the shroud in reducing tip losses is not accounted for in this analy-
sis, nor is the reduced swirl loss due to radial expansion in the diffuser
and the reduced optimum disk loading of the diffuser system (Ap2o/(l/2)pV£=
2/3 versus 8/9 for the unshrouded optimum) .
The relative advantages of a diffuser-augmented wind turbine will

be sensitive to the type of application; that is, the size of unit, the
economic value of a broader operating range, and the local wind spectrum.
Technical issues that need better definition are the relationships be-
tween diffuser efficiency and diffuser geometry with a turbine exhaust as
input flow, the range of Cp 4 that can be achieved with a practical ex-
ternal contour as a function' of V4, diffuser and support construction
costs, optimization of X for given applications, and trade-offs for
rotor design factors such as pitch control and disk loading as affected
by the diffuser.
REFERENCES
1. Lilley, G. M. ; and Rainbird, W. J. : A Preliminary Report on the

Design and Performance of Ducted Windmills. Technical Rep. CA119 ,
The British Electrical and Allied Industries Research Association,
Great Britain, 1957.
2. Glauert, H. : The Elements of Aerofoil and Airscrew Theory. Second

ed. , Cambridge University Press, 1948.
DISCUSSION
COMMENT: Our calculations show very similar results. One point, though,
which I am sure the speaker is aware of is that a ducted rotor is never
better than a free rotor that has an area equal to the area of the duct
exit.
A: Right. The equivalent rotor would be between the two. It can be

better in that it can have a very much lower wind velocity cut-in speed.
104
That could be more important than the power per unit size.
Q: Did the speaker ever consider actually making the duct rotate with
the rotors? What might the losses be in a system like that?
A: On what axis?
Q: Just fasten the rotor tips to the duct itself, so that you removed
all your clearance and mechanical problems, and have the duct go
round.
A: I think the bearing problem would bother us more than the tip losses
at the wall would.
Figure 1
I
0
Figure 2
. 105
I
I
106
BUCKET ROTOR WIND-DRIVEN GENERATOR
Howard H. Chang
California State University

San Diego, California
and
Horace McCracken
Sunwater Company
San Diego, California
One of the common types of impellers for moving large volumes of

air at low velocity is the bucket rotor. Perhaps we have overlooked the
possibility of using this design in reverse for the extraction of power
from the wind.
To get some preliminary feel for this, a unit with rotor 4 feet in
diameter and 4 feet long has been built (see figure 1). It has deflec-
tors on the top and bottom to guide the wind into the top half. The
lower deflector also shields the back side of the bucket from the wind,
thus reducing the reversing wind force. The present rotor is fixed in
direction facing the predominant wind; it may also be mounted and in-
stalled with a tail boom to follow the direction of the wind.
Mounted on a trailer towed by an automobile, this unit has been

tested at wind speeds from 15 to 40 mph. The mechanical energy produced
was measured using a rather crude dynamometer. The maximum power at 40
mph was measured to be 0.14 horsepower. Further improvements in the con-
figuration design will undoubtedly improve the performance.
As compared with the ordinary propeller-type rotor, the bucket rotor

is limited in rotational speed since the tip rotor speed can never exceed
the wind speed. However, it does not present the blade fatigue problem
that the ordinary rotor does, and it perhaps causes less sight pollution.
The deflector vanes also provide a venturi passage to capture greater
wind flow. The bucket rotors can be strung together end-to-end up to
thousands of feet long to produce large amounts of power.
DISCUSSION
Q: This is simply a Savonius rotor on its side with a couple of vanes

out front to direct the flow. The disadvantage is that itrs direc-
tional.
A: It is directional, certainly, but because it is directional, the
107
vanes here have the Venturi effect that simply improve the efficiency
of the thing.
Q: Right, but, if radial vanes are mounted on a Savonius rotor it cap-

tures the wind from all directions, the effect is the same.
A: But the Savonius rotor has a vertical axis. However, if you are free
of directional influence, you cannot put in those guide vanes.
Q: You could put them in as part of your supporting structure. They

would hold the upper bearing of the vertical axis. I don't see how
your rotor is any improvement over the Savonius rotor with^guide
vanes.
Figure 1
40 MPH
.1
*.
H
.01
.001 -I L J I
20 40 60 80 100 120 140 160
RPM
Figure 2. - Test results of the bucket rotor wind-
driven generator.
108
•
WIND-POWERED ASYNCHRONOUS AC/DC/AC CONVERTER SYSTEM
D. K. Reitan
University of Wisconsin
Madison, Wisconsin
The two asynchronous ac/dc/ac systems shown in figures 1 and 2 are

in the process of being modelled at the University of Wisconsin. The
figures contain the main information and are somewhat self-explanatory
such that only a brief explanation is contained herein. The system of
figure 1 can use a variable or constant Hertz alternator drive such as
can be provided by wind power. The system of figure 1 generates (ref. 1)
variable Hertz ac, rectifies this with a frequency independent three-phase
two-way, six-pulse bridge-rectifier (see inset) operating with constant
current control, and parcels this constant IR among pumped-storage, dc
loads, and the high-capacity 60-hertz bridge inverter, also operating
usually under constant current control (C.C.C.). The 60-hertz bridge
inverter comprises a stable link to the power company to either supple-
ment them from wind energy, storage, or from a combination of both at a
preset desired current. In that the rectifier and inverter are identical,
they are "converters" and can operate in either mode depending on the
silicon-control-rectifier (SCR) firing angle a- Thus, if needed, the
inverter can go over into rectifier mode and pump back into storage.
Tests were run at a wide range of wind-bus hertz and with rectifier C.C.C.
set at higher and lower levels than the inverter C.C.C. - storage taking
up the slack. The system is not "sensitive" in that, for given C.C.C.
settings, Sw3 can be opened and closed at will - suddenly - activating
or cutting off the inverter current to the 60-hertz power company bus.
The system is presently modelled as a three-phase ac generation/dc/three-
phase 60-hertz inversion, but a single-phase 60-hertz C.C.C. inverter
will be built also.
The system of figure 2 employs the same rectification but from a

60-hertz alternator arrangement (refs. 2 to U) . This system is missing
the high-power 60-hertz inverter tie to the large backup supply of the
power company, and it is thus meant to be a self-contained electric supply.
System 2 has the option of mainly dc output, some sinusoidal 60-hertz from
the wind bus and some high harmonic content 60 hertz from the 800-watt
inverter. Work is presently under way to do some wave shaping on inverters
of this type to investigate the harmonic tolerance of various appliances
(ref. 5).
Figures 3 to 8 are mostly self-explanatory and show the instantaneous

waveforms of voltage found at the indicated places in figure 1. Of note
are the following:
109
(1) "Zero" voltage exists during conduction of an SCR.
(2) The "safe" negative, but large, voltage exists across a
rectifier SCR during its off period.
(3) The potentially troublesome large positive voltage (shown
in fig. 7) exists on an inverter SCR during its off time.
(4) The harmonic content in the bridge voltages increases in all
cases when both the SCR firing angle and commutating angles are greater
than zero.
(5) The firing angle of an inverter SCR must be such that the
conduction period (120° + y ) is over soon enough to allow the deion
angle Y or a commutation failure will occur. This is the reason for
the build-in inverter constant extinction angle control (C.E.A.), which
will override the inverter C.C.C. when necessary.
(6) Every commutation between a pair of SCR's momentarily results
in a direct short circuit on one of the line-line voltages (fig. 8).
For example (see inset, fig. 1), a commutation from SCR#1 to SCR#3 causes
a short on line-line voltage Ea^ for a duration of p . The firing
order of the converter shown is 1, 6, 3, 2, 5, 4.
Figure 9 shows the input and output of a six-pulse bridge-rectifier
as supplied by a 100-percent field-modulated alternator (refs. 2 to 4) .
Pertinent operation is evident or noted directly on figure 9.
REFERENCES
1. Clews, Henry: Electric Power from the Wind. Solar Wind Company
booklet and brochures on wind generators, RFD 2, East Holden,
Maine 04429.
2. Bernstein and Schmitz: Variable Speed Constant Frequency Generator
Circuit Using a Controlled Rectifier Power Demodulator. AIEE,
Paper No. CP60-1053, San, Diego, Calif., Aug. 1960.
3. Allison, Ramakumar, Hughes: A Field Modulated Frequency Down Con-
version Power System. IEEE Industry Applications Meeting
Philadelphia, Pa., Sept. 1972.
U. Lindsley, E. F. and Luckett, H.: New Alternator Delivers 60-Cycle
Power at Any Speed. Popular Science, Jul. 1973, page 38.
5. Grateful acknowledgment to the Sperry Univac Corp. and the Univ.
of Wisconsin Research Committee for their encouragement and
support to continue this project.
110
DISCUSSION
Q: With the rectifiers you were describing, is it possible to charge

two frequencies?
A: No. I had a little 14-pole generator; 14 poles develops 60-cycles
at 514 rpm. I ran it over 5000 rpm and operated it with no delay
and with delay. The rectifier is independent of the frequency. It
will pump into storage constant current control. All solid state
controls will operate at that frequency.
Q: We found the converters were extremely expensive.

A: They are expensive, but I would like to add that this inverter-con-
verter scheme, this particular graph's type bridge, is what's used
in the present-day, large-scale dc power links. There are a dozen
or so operating within the world, and there is practically no limit
on the size of these converters. With the present state of the art,
this rectifier and this inverter can be built for 3000 megawatts.
Q: I believe this solves the problem of the variable at one speed.

A: The rectifier is independent of the frequency. The inverter will
operate strictly at 60 cycles and tie right onto the existing supply.
Q: Is that generally cheap in a system - are there any hydraulic

variations?
A: I'm assuming that somebody can supply the proper wind generator, and
I take it from the electrical end.
• Automatic Variable Pitch Self-Feathering C.C.C. « constant current control

1 • 1 Relatively Constant-Speed, C.E.A. • constant extinction
QUIRK-type Propeller " ISO rpm angle control
Frequency Independent Three- Rural

To Power
Phase, 6-Pulse "Bridge" Residence .
Ccmpany
RECTIFIER with smoothing Rural AC Loads
Network
inductors L^c* (Diodes or Residence
SCR's depending upon C.C.C.)- DC Loads
30 or 10
Bus
115 volts
AC
Three-Phase or
60 Hz
Single-Phase SCR
Bridge CONVERTER
Power
(Inverter-Rectifier)"
Company
Secondary
Service
V^gp - Power Conpany Bus RHS volts
Swl - Circuit breaker *1. etc. During off-peak periods the

SCR Inverter can operate In
Rectifier mode to charge bat-
teries if no wind available
FTCURZ 1. Ron-Synchronoua AC/DC/AC Pumped and Pumpback Storage Wind-Energy System to Supplement Rural Power Company Supply
111
3
CO
I
ut)
T
800 watt
Inverter
41 • O
js -a m
I
41 U
4J 41 1-1 .a M o
CO 4) H= a a •
4) P. il 4) -H 4) U
(ti CO w oo js -a
I IM
<w <u o
i-l C rrl
HI a) r-<
CO U 4)
0.
§ p
o o :
u O CL,
112
FIGURE 6. Full-Load Rectifier SCR Voltage
FIGURE 5. No-Load Rectifier SCR Voltage a # 0 a =0 V = 17°
For same full load V is less because of Positive SCR voltage,

greater commit at ing voltage when a > 0. ;gate must hold
SCR off.
FIGURE 7. Full-Load Converter SCR Voltage — a = 25°, V - 7°
Remarks: 1. The system of Fig. 1 operates

in a normal manner with rated V^ty on the
wind-bus and rated VACP on the Power Co. bus,
for any level of Storage Battery System bus
voltage from 10% to 100% of normal. The oper-
,wt—• ation for lower Storage bus voltages merely
changes to larger firing angle <C and a diff-
erent C.C.C. setting for the same constant
current.
Repeating Each notch and 2. For low Storage bus-voltage (due to temp-
line-line ent occurs at orary partial outage) the harmonic content
short circuit 'some SCR commu- increases, the SCR voltages go lower and
tating pair shift to the left in all the above figures,
and the volt-arnpere-reactive requirement of
FIGURE 8. Full-Load Line-Line Voltage both. AC busses unfortunately increases. This
would be a temporary situation, usually.
113
•8
*
CO
•
o\
X*
AN ELECTRICAL GENERATOR WITH A VARIABLE SPEED
INPUT - CONSTANT FREQUENCY OUTPUT
H. Jack Allison
Oklahoma State University
Stillwater, Oklahoma
ABSTRACT
A new type of rotary energy conversion device for obtaining a de-
sired constant frequency output independent of the speed of the prime-
mover has been developed arid tested, using the technique of field
modulation and solid-state alternator output processing. This paper
describes a 10-kilowatt prototype field modulated frequency down con-
verter system designed, built, and successfully tested at Oklahoma State
University. Experimentally obtained performance characteristics are pre-
sented and discussed.
PRINCIPLE OF OPERATION
A conventional three-phase synchronous machine of basic frequency
fr will have induced voltages of frequencies (fr + fm) and (fr -
f^ when excited with an alternating current of frequency fm; fr ;>
fm. When such three-phase voltages are individually full wave rectified
and their outputs tied in parallel, an output voltage containing the
following components results:
(1) dc component
(2) Ripple of frequency 6 fr
(3) Full-wave rectified sine wave at the frequency fm;
V 1I sin ait
m 'I where to = 2irf
mm
The dc component is, in general, proportional to the reciprocal of
the modulation frequency ratio m, where m = fr/fm. For values of m
greater than 10, this component becomes negligibly small. The resulting
full-wave rectified sine wave can be converted to a sine wave voltage at
the modulating frequency fm by using a suitable switching circuit em-
ploying controlled rectifiers.
DESCRIPTION
Figure 1 shows a simplified schematic of the frequency down con-
verter system. It is built around a high-speed high-reactance high-
frequency three-phase alternator. Both rotor and stator are laminated
to minimize the iron losses. The six stator leads are brought out and
115
three full-wave bridges are connected as shown, one across each of the
phases. The outputs of the bridges are tied in parallel across the load
through a silicon controlled rectifier switching system. Tuning capacitors
C are connected across each of the stator windings to decrease the excita-
tion requirements (both watts and vars) at the rotor terminals. The main
switching process is accomplished by the four controlled rectifiers, SCR1
through SCR4. The commutating circuit consisting of 1*2 and C2 and
the controlled rectifiers SCR 5 and SCR 6 aid in this switching process,
especially when the load is not purely resistive. In addition to filtering,
capacitance Cg enables the handling of lagging power factor loads by the
system.
The field is excited by an ac power source of frequency fm. Since

this frequency fixes the output frequency, care must be exercised in the
design of this part of the system. In case the system is required to be
completely self starting, an inverter dc source combination must be used
to excite the field. The dc power might come from an exciter alternator-
rectifier unit mounted on the same shaft as the main generator.
The 10-kilowatt 220-volt single-phase 60-hertz prototype designed

and built has 16 poles and runs at around 7000 rpm. This corresponds to
a frequency fr of about 930 hertz. For a modulating frequency of 60
hertz, m is between 15 and 16. The dc component associated with this
value of m is essentially negligible and causes no problem in the
switching action of the SCR circuitry. The rotor diameter is 6 inches,
and active iron length is 2.5 inches. Overall dimensions of the generator
are a diameter of 9.75 inches and a length of 7:75 inches. The electronics
associated with the system can be arranged compactly. A photograph of
the 10-kilowatt prototype system is shown in figure 2.
PERFORMANCE RESULTS
The efficiency of the generator for three values of stator tuning

capacitors is shown in figure 3 plotted against output power. Figure 4
shows the rotor input power for three values of C. Stator tuning capac-
itors significantly improve the performance of the system.
The desirability of rotor tuning is brought out in figure 5. Rotor

tuning reduces the volt-ampere capacity required of the excitation source.
The voltage regulation of the frequency down converter is improved

by the.variation of the effective rotor reactance with output power (see
fig. 6) . Figure 7 shows the variation of the output voltage from rated
output rated voltage conditions as the output power is decreased. It
can be seen that with C=5yF, the output voltage stays essentially con-
stant down to about 40 percent of rated load. Smaller values of C do
not result in this desirable characteristic. It is possible to exploit
this property by properly choosing C to obtain nearly constant output
voltage from low-load to full-load conditions.
CONCLUDING REMARKS
The field modulated generator system (FMGS) described in this paper
116
has several advantages and potential applications. Since the output
frequency is independent of the prime-mover speed, variable speed prime
movers such as the ones available in aircrafts or unregulated high-speed
turbines and wind energy devices can be used to drive the generator.
In addition to being smaller in size and weight, it has inherently

better regulation and independence from rotor inertia effects. The
principle of operation appears to be applicable to systems of any size.
The efficiency of the system is similar to those of conventional systems
of equivalent rating. Whereas a conventional synchronous machine oper-
ating in parallel with bus bars can receive or deliver power (motor or
generator) to the supply, FMGS can operate only as a generator because
the bridge rectifiers will prevent any power flow from the bus bars to
the generator. The behavior of the FMGS operating in parallel with an
existing power system under normal and fault conditions is yet to be
completely explored. The results obtained from the prototype are promis-
ing, and further development, testing, and design optimization are under-
way.
DISCUSSION
Q: Do you have any idea what the dollars per kilowatt would be on, say,
a 1-megawatt unit?
A: Again, you play the game the way everyone else here has played the
game; that is, if you want to build a million of them, the price
will be quite low. We've made a fairly careful analysis of it, and
if you built a lot of them in reasonable sizes, about a hundred
dollars per kilowatt is our projection. We are also standing ready
to sell you some, up to the 60- to 100-kilowatt level, at a very
much inflated price over that unless you want to order four of five
hundred thousand of them.
Q: What is the working rpm of that generator?

A: The generator that we showed operates at 7000 rpm. It would operate
successfully at any speed from, we think, around 1200 to 10,000 rpm.
A 5 or 6 to 1 speed range is not inconceivable. We built it to
operate at 7000 rpm because that was the nature of the prime mover
in the pulley arrangement we had to begin with. We could, of course,
design it to operate from, say, 500 to SOOO rpm. You take your choice
as to what the speed range is. You propeller people tell us what the
speed range should be and what gearing should be, and we will easily
accommodate 5 to 10 to 1 speed variations, we think, with no particu-
lar problem. This particular machine's most spectacular character-
istic is obtained if you run it at its rated load condition, and
then if you simply turn off the power to the prime mover you will
find the 60-hertz sineusoid absolutely stable until there is no prime
mover turning at all. You will find, of course, that its amplitude
varies.
TJ: ~ That -is a brushless generator?

A: It is not a brushless generator at present. We built it to handle
117
brushes because we figured we could make it much lighter that way.
We checked with various people about brushes, and we found that in
terms of trying to get a constant frequency output for variable
frequency input brushes would be the least of your problems. We
think we can make it without brushes, and we are proceeding in that
particular direction now.
/
Q: Can you synchronize this on a single powerline and match your modulate
with a power angle, or with the power output ahead of that?
A: We've done that, yes. You see the thing is that the frequency can be
obtained directly from the powerline, as your particular discussion
indicated, and then the problem is no problem at all. The only thing
that you can't do with this particular machine is pump power from the
powerline into it, which we consider an overwhelming asset, and that
is the reason we designed it that way.
Q: Can you contrast your generator with a Precise Power Corporation's

generator?
A: No, because we don't have complete information on the Precise Power
Corporation's variable speed generator. We are interested in getting
that information, and we will give them free access to all of ours
and hope that they would reciprocate.
Figure 1. Simplified schematic of the field modulated generator system

developed at Oklahoma State University, Stillwater.
118
Figure 2. Photograph of the 10 kw generator prototype and the associated
electronics (prime-mover not shown).
2 4 6 8 10 12
POWER OUTPUT, KW
Figure 3. Efficiency versus output characteristic.
119
1000 10
FIELD WINDING CURRENT
ROTOR TUNED CIRCUIT
INPUT CURRENT
4 6 8 10 12 4 6 8 10 12
POWER OUTPUT, KW POWER OUTPUT, KW
Figure 4. Variation of the rotor Input potter with output. Figure 5. Variation of field winding and rotor circuit Input currents
trlth output.
300
250
200
§
150
100
50
4 6 8 10 12 4 6 8 10 12
POWER OUTPUT, KW POWER OUTPUT, KW
Figure 7. Voltage regulation characteristic of the 10 lew prototype.

Figure 6. Plot of effective rotor winding reactance for different
values of C and output.
120
VOIGHT VARIABLE SPEED DRIVE
Joseph Tompkin
Salem, Oregon
The variable speed drive transmission is mounted within the gondola
and connected with the wind turbine blades and the hub. This unit is
designed for the production of ac power. The turbine turns by means of
the variable speed drive and a set of synchronous three-phase generators.
This motion is controlled automatically by two wind rosettes in such a
way that the wind turbine always opposes the wind direction.
Adjustable speed is frequently a problem. For the majority of
machine applications, mechanical speed changers are eminently suited and
often provide the simplest and most economical answer. Efficiency of
power transmissions employing mechanical principles runs to over 90 per-
cent. Through the elimination of complex hydraulic, pneumatic, and
electrical elements, mechanical speed changers are simple in design,
manufacture, operation, maintenance, and exchange of all structural
parts.
The Voight variable speed drive is a mechanical variable positive
drive gear transmission. It has an unlimited power and torque trans-
mission, a constant ratio with high degree of accuracy, a speed variation
over a wide range, and a nonslip drive. The following are some specific
advantages :
(1) Any desired speed range is available from 1 to infinity.
(2) Smooth acceleration is possible from zero to maximum speed
and deceleration is possible from maximum speed to zero for any character
of load.
(3) Any practical number of positive drive stable speeds are avail-
able for any chosen speed variation, and they are accurate to a split rpm
even with varying loads.
Variable speed drive model 58100 (enclosed) covers three speed
ranges from 0 to 120, 120 to 360, and 360 to 840 rpm and reverse speeds
from 1220 to 2320 rpm. There are 320 nonslip fixed speeds hand adjustable
under load while operating the drive.
(5) Jogging or preset speed is controlled by automatic accelera-
tion or deceleration.
(6) Dynamic braking exists for quick automatic stopping or where
121
controlled deceleration of load is required.
(7) There are multiple driving units with related speeds paralleling
applications controlled from a single control (for applications where two
or more machines must be "link" synchronized) .
j
(8) Operation is by remote control.

(9) There is visual speed indication of operating rate in rpm or
as desired.
(10) The Voigt variable speed drive meets the demands of all three
general types of power transmission: Constant torque, constant horse-
power or variable horsepower, and variable torque except in the very low
speed range where the horsepower is somewhat limited because the torque
tends to infinity, which is impractical.
(11) Ball bearings are used to reduce friction.
(12) All operating parts are splash lubricated.
(13) Applications are possible for powers from 5 to 10,000 horse-
power.
(14) Load shocks are relieved by resiliency of the chaindrive
incorporated in the unit and its autotant tension.
(15) Voight variable speed drive provides a means of controlling
the speed_jxf—ûreôr^ more standard ac induction motors by simply control-
ling the frequency of the power applied, thus making them variable speed
'drives.
If wind power were to be used primarily for pumping water, instead
of generating electrical power, it is most practical to produce compressed
air by means of a radial type compressor built into the wind turbine hub
(see enclosed data sheet and schematic diagram) .
DISCUSSION
Q: Have any of these units ever been_built or are these just designs?
A: The first prototypes have been built, the first one in Los Angeles.
What you just saw was simply patent descriptions. We would like to
get involved in building this variable speed transmission, because
it will control the fluctuation of the wind velocities in our turbines.
This is perhaps a key to the controls so that we can maintain or have
a constant voltage on alternating current generators.
122
ENERGY STORAGE USING HIGH-PRESSURE ELECTROLYSIS AND
METHODS FOR RECONVERSION
William L. Hughes
Oklahoma State University
Stillwater, Oklahoma
About 12 years ago, the School of Electrical Engineering at Oklahoma
State University undertook what became a rather extensive and continuing
study (both theoretical and experimental) on ways to store electrical
energy and thereafter reuse the stored energy in various ways. Initially
(about 1961), theoretical studies were undertaken of various possible
storage methods, which included the following:
(1) Mechanical storage (flywheels and related devices)
(2) Pumped storage (hydroelectric)
(3) Cryogenic magnetic fields
(4) High-pressure electrolysis (producing hydrogen and oxygen)
From the outset, the O.S.U. group was concerned with developing
energy systems which showed promise of being expandable to large-scale
power systems. Thus, systems requiring exotic materials (such as plat-
inum) were rejected from study as having little long-term possibilities
on a commercial scale.
The result of our initial theoretical studies on energy storage
seemed to indicate (to us at least) that high-pressure moderate tempera-
ture electrolysis had the greatest long-term economic promise, and our
rather extensive experimental programs on storage were concentrated in
that area. Over an 8 or 9 year experimental period, we have worked in
the area of electrode design (both solid and porus) , electrode and1 mem-
brane life, and overall electrolysis system efficiency.
The primary results of the energy storage activity can be summarized
as follows:
(1) A number of electrode designs were examined. The most success-
ful was a solid nickel finned type of electrode (fig. 1). Electrolysis
efficiencies in excess of 85 to 90 percent were achieved. Efficiency is
defined as the ratio of the heat content of the gases produced to the
equivalent electricity used to produce the gases. Current densities for
these results ran at 400 to 700 amperes per square foot. Optimum pres-
sures were around 200 atmospheres and optimum temperatures around 350° F
(see figs. 2 and 3).
123
(2) Enough information was obtained so that a practical, efficient
electrolysis system could be designed, built, and operated.
(3) Some studies were made of the projected cost of deep cavern
high-pressure gas storage which indicated this technique could be feasi-
ble in some locations as long as gases were stored below normal hydro-
static pressures.
Simultaneous with the energy storage research, we undertook both
theoretical and experimental studies of ways to reuse the stored hydro-
gen and oxygen as well as certain other energy conversion methods. The
areas of effort in reconversion are as follows:
(1) Moderate temperature, high-pressure hydrogen-oxygen fuel cells
using no noble metal catalysts were studied.
(2) The "aphodid" burner-turbine generator concept was studied (a
method of burning hydrogen and oxygen in a long tube with an injected
moderating water spray such that steam could be generated at any desired
temperature and pressure).
(3) The field modulated generator system was studied (covered in
an earlier paper in this meeting because of its obvious direct applica-
bility to a wide variety of variable speed prime movers such as aero-
turbines and unregulated high-speed gas turbines).
The results of the work on energy reuse were as follows:
(1) Fuel cells operating at pressures up to 200 atmospheres and
300° F were built. The effects of temperature and pressure were experi-
mentally mapped and typical characteristics are shown in figures 4 and 5.
All fuel cell work, as well as all electrolysis work, was done with
nickel electrodes and no special catalysts. Rechargeable hydrogen-
oxygen fuel cells employing a porous membrane (cylindrical geometry,
fig. 6) made of calcia stabilized zirconia and sintered nickel electrodes
with no noble metal catalysts were investigated extensively to study
the effects of pressure, temperature, and membrane porosity.
(2) The aphodid burner (fig. 7) turbine generator system has never
been built, but some years ago Dr. Stanley Brauser (a thermodynamicist
then on our mechanical engineering staff) studied this at our request
and concluded that efficiencies around 40 percent were obtainable (elec-
trical equivalent Btu output over fuel Btu input). The difference
between this and conventional plants is primarily the elimination of
stack losses. Probably another few percent can be picked up by combining
the field modulated generator (discussed earlier) with the aphodid burner.
This would allow turbines to run at much higher speeds and probably at
somewhat higher temperatures. These two factors yield higher turbine
efficiency.
(3) The results on the field modulated generator have been reported;
no other comments are required other than to say that it is fast approach-
ing the stage of direct application to variable speed mechanical inputs.
(4) Finally, it should be noted that high-pressure hydrogen can be
used as a basic ingredient in very efficient conversion of organic materials
to various hydrocarbon fuels, including methane. We have begun to gather
technical material in this area, assisted by Dr. Wm. Crynes of our
Chemical Engineering Department. Dr. Crynes is a recognized authority
in coal gasification, and this area will be pursued as vigorously as re-
sources permit.
The work described herein has,.of course, been spread over several
years. Much of it, initially at least, falls into the hazardous category.
We have a specially built hazardous reaction facility, and, much of the
work has been done there (figs. 8 and 9). That facility essentially pro-
vides an "explosion proof" chamber where reactions involving hydrogen and
oxygen can be safely handled. We believe that we can formulate the rules
for designing high-pressure moderate temperature electrolysis and fuel
cell systems which operate safely. That, however, is a completely sepa-
rate subject.
Finally, a complete wind generator energy storage system built about

5 years .ago just to get an idea of the overall energy availability is
shown in figure 10 on a 20-foot test platform at the airport laboratory
facility. Figure 11 shows a close up, and figure 12 shows the details
of the electrolysis system.
DISCUSSION
Q: How do you foresee the potential eventual use of these types of

systems? Do you ever see the day when they could possibly be
domestically used. There is the safety problem with hydrogen.
A: First, weTre studying ways to use high-pressure hydrogen to hydroge-
nate organic materials to synthesize hydrozenes and methanes, and
this looks very promising. So, one application that we have missed
is a way of making synthetic vehicular fuels. It is not necessary
to convert hydrogen to electricity to get a lot of good use out of
it. Second, I'm a little nervous about using hydrogen as a domestic
fuel. Looking at what we call our "Allison's boom room", the haz-
ardous reaction facility, we have had some explosions there in a
controlled environment. We think we now know how to handle hydrogen
safely. At least, we haven't had any explosions for three years.
Third, there is a lot of use that can be made of the hydrogen in
heaters besides just converting it to electricity.
Q: I wonder if I understood the efficiency figures you had correctly.

The 68 percent you gave was power into the electrolyzer compared to
power out of the fuel cell?
A: When I said 60 percent overall, I meant kilowatt hours out of the
fuel cell divided by kilowatt hours into the electrolyzer.
Q: Any comparable figure for the electrolyzer and gas turbo combination?
A: It's about 40 percent.
Q: You raised the question of using hydrogen in the home. I have an

article that says there are two or three miles of hydrogen pipeline
in Germany that has been used for 20 or 30 years.
125
A: I didn't know they were using it for home use, but I'm intensely
interested. The Germans have, of course, been handling high-pressure
hydrogen in pipelines much longer than we have and are very experi-
enced in it. I'm hopeful it can be used in the homes.
Figure 1. Photograph of six cell electrolysis module.

01 3
> +J
«- U
O
58
<0r-
*o-o
•
0* VI
5-S
0) at
U O
fit
O *->
s!
in
I
a>
§ 2 §
(S.L10A) 30VJ.HOA 818X10810313 I6110A) 39V110A TQO 13Hd
8
ot> x
>>
0.10
a. g •—•5
4
O U-
CO £§
Ul
o a> n
0.0, ai
^
01
tu
K
oc «-t
ts
3 uj o.
2
i,
O 00 ID
(VJ CM CM —
S 6 6 0
(SnOA) 39VJ.10A SISA10U10313
(S110A) 30V110A 1133 130d
127
APHODID
BURNER TURBINE
HIGH PRESSURE I i— T--I

, ELECTROLYSIS j ISTORAGjJ ]
I |
~r^
-~-\—°"2"T-j
Figure 6. Photograph of a five cell battery of rechargeable Figure 7. A simple Aphodld flow diagram.
fuel cells employing cylindrical porous zlrconla
membranes and sintered nickel electrodes.
Figure 8. OSU Hazardous Reaction Facility.
1 Pfi
1
Figure 9. View of the hazardous reaction chamber Figure 10. Photograph of the 500 watt experimental
of the hazardous reaction facility. prototype wind energy storage system
mounted on a twenty foot platform.
Figure 11. Close-up view of the experimental Figure 12. View of six cell electrolysis module
prototype wind energy storage system. with the high-pressure test chamber in
background.
129
USE OF HYDROGEN AND HYDROGEN-RICH COMPONENTS AS A MEANS OF
STORING AND TRANSPORTING ENERGY
Walter Hausz
General Electric Company

Santa Barbara, California
TEMPO has been interested in advanced energy utility systems, which

we call Eco-Energy Systems (fig. 1) where the ecology and economics are
kept in.balance. Our conclusion has been that using hydrogen for trans-
port and storage can lead to very clean, flexible systems that with
improving technology and depletion of natural fuel resources can become
competitive. While the primary source of energy is shown here as nuclear,
the basic concept can apply for any thermal source, or for wind energy,
etc. (fig. 2).
With a nuclear reactor an important reason for using hydrogen

storage is to keep efficient full-load operation of the reactor while
the customer demands fluctuate. With wind, Sun, and tidal energy, both
the supply and the demand fluctuate so some form of storage is even more
important. Hydrogen, or hydrogen plus oxygen, is an important candidate,
because of its easy transport by pipeline and the flexibility it has for
serving all of the energy sectors.
There are many ways of storing energy (fig. 3) . The merits for a
particular application depend on such things as the energy density, the
ease and flexibility of reconversion to a form of energy useful to cus-
tomers, and the cost per unit of delivered energy. Energy density (fig.
4) and flexibility of conversion are, of course, important ultimately
for cost determinations. Hydrogen as a cryogenic liquid and as metal
hydrides is more attractive than hydrogen as a gas. Of course, the
hydrides of nitrogen (NH3) and carbon (gasoline) are even more compact.
There is a rough road ahead to get the costs down. The costs of
storage vessels can be reasonably determined (fig. 5, curve from the
source material of the Synthetic Fuels Panel). But ,to really determine
the cost of storage we have to examine all the energy conversions re-
quired, their efficiency, and their capital cost.
To illustrate (fig. 6), I've assumed a 1-megawatt wind energy source

that operates half the time. Half of its output is used to ser.ye custom-
ers directly as electricity; the other half is to be stored. For a system
of this size, a cost of electricity at the generator of 10 mills (one
cent per kilowatt-hour) may be low, but it is a convenient unit. As
conversions of the storage portion are made to hydrogen, to liquid
130
hydrogen, to stored LH2, and back to electricity, the energy costs and
capital costs of the conversions escalate the unit cost per kilowatt-
hour remaining until electricity at 12.9 cents per kilowatt-hour results.
Figure 6(a) shows the next level of detail of the assumptions.
Many of the high costs result from scale size and would be less for
a 100- or 1000-megawatt system. With adequate research, technology will
improve all of these. Such alternative means of storage as M^NiHq. and
FeTiH2 are being actively explored at Brookhaven National Laboratory
both for mobile and utility applications. And for all but the biggest
systems they may be less costly than LH2.
Another possibility, which in the largest systems may be the best,

is storing in the form of ammonia (fig. 7). While this concept shows
both electricity and fossil fuel used to make the NH^, either could be
used alone.
In conclusion, system configurations that consider the storage

alternatives are an important part of the research needed to achieve
economic viability.
DISCUSSION
Q: One of your figures appeared to show a block with heat energy going
into something and hydrogen and oxygen coming out. Are you really
showing thermal dissociation as a way of making hydrogen and oxygen?
A: I made that block diagram very general so it could cover everything.
But thermal dissociation of hydrogen, particularly things like the
Marketti process, Mark 1 process, is one of the things that on a
very large scale look best. But this, of course, requires a thermal
source and this is a wind energy conference; thus, electrolysis is
the means of preference.
Q: Most people dealing with hydride storage tend to talk about the vol-
ume of hydrogen that can be stored in a given volume of storage
material, but since the installed cost of that system depends on how
much you have to buy, would you care to comment on how many pounds of
hydride or whatever is required to store it?
A: Well, let's deal in terms of per cubic foot, and I said that with
liquid hydrogen you get about 4.4 pounds per cubic foot, whereas with
ease you can get 6 pounds of hydrogen per cubic foot. Now the density
of magnesium - and magnesium nickel is principally magnesium (only
about 6 percent nickel) - is, I think, about 2 or 2%, so you've got
about 150 pounds of magnesium for your 6 pounds of hydrogen. Iron
titanium is, of course, a higher density, about 5 or 6, so you've
got more pounds, but it's a cheaper material. For portable use"like
in a car, you probably want magnesium. For utility use, I think the
iron titanium with its lower cost per pound -is probably superior.
131
EGO-ENERGY
INDUSTRIAL PROCESSES
LAR6 MCLEAI RATER COMPRESSOR LIQUID HYDROGEN & I __.
SPLITTING STATION OXY6EN STORAGE
TRANSPORTATION
Figure 1
FUEL
INDUSTRIAL «-i COMMERCIAL D

THERMAL
SOURCE MUNICIPAL •*• R E S I D E N T I A L I
^
•* S
HE'AT TRANSPORT
t T
R
CONVERSION -k.
TO H Y D R O G E N
^
ELECTRIC
I
A N D OXYGEN B
I U
BYPRODUCT STORAGE BYPRODUCT T
HEAT OXYGEN HEAT
I
0
BYPRODUCT STORAGE BYPRODUCT
WATER HYDROGEN WATER N
Figure 2
ENERGY DENSITY
BTU/FT
PUMPED STORAGE (100 ft head) 14
MEANS OF STORAGE HOT ROCKS/METAL 60-500'F 8,000-12,000
MOLTEN SALTS 60- SOOT 10,000-20,000
MECHANICAL
STEAM 15 pal 212°F 40
HOT ROCKS/METAL
130 347 340
HOT WATER/STEAM
500 467 1,370
MOLTEN SALTS
WATER 15 212 9,000
HYDROGEN 130 347 16,000
• GAS
500 467 21.000
• LIQUID
• HYDRIDES HYDROGEN
• GAS 15 60 280
1,000 60 18,500
AMMONIA
• LIQUID 15 -425 200, 000
METHANOL
• HYDRIDE (MgjNj or F^) 250,000
GASOLINE
AMMONIA 340, 000
BATTERIES METHANOL 430, 000
GASOLINE 830. 000
BATTERIES 10.000-80,000
Figure 3 Figure 4
FUEL STORAGE INVESTMENT (1972 BASIS)
10 102 103 10*

FUEL QUANTITY STORED (108 BTU)
Figure 5
133
WIND ENERGY
WIND ENERGY
1000 MW ELECTRICITY
1 500 kWr I Electrolysis; E££ .75 . 33

1000 MW :TRI
ELECTRICITY 1.0l/kWh V ge
Fixed Charge ..47
3000/d HYDROGEN > 1.8
I Liquefaction; Energy
Fixed Charge
.53
. 87
300J/d HYDROGEN
)RQC 1.8
300«/d LIQUID H2 3. 2
I 1 1200* Storage
300»/d ' luro H,
LIQUID 3.Z I' Fixed Charge 1.74
4 days STORAGE
ORA 4.94
I I | Fuel Cell: Effic. .50 4. 94
4 days ORA
STORAGE 4.9 i" Fixed Charge 3. 00
200 kW ELECTRICITY 12. 9i/kWh
I I
ZOO kW ELECTRICITY 12.
Average Cost
( 2 • 10 6 + .75 . 106 • 12. 9\

2.75-106 ]
Figure 6 Figure 6A
FLOWSHEET OF AMMONIA SYNTHESIS
1100 TONS (733)1 OF BITUMINOUS COAL

OR 1460 TONS OF LIGNITE
660 TONS 0,
(440)
COAL GASIFIER
AND
SHIFT REACTOR
122 TONS H2
(81)
El
HYDROGEN
(136)3
440 TONS O2
BOO TONS -*• WAIrER 55 TONS H2 177 TONSH 2 AMMONIA TONS

H2O ELECTR OLYSIS SYNTH ESI
' NH3
- 100 MW92 "* FACI LITY ^^ FACILITY
830 TONSN 2
1050 TONS AIR

(26 X 106 FT3)
220 TONS O2
A*R OVI•RALL C OAL GASIFICATION
LIQUEFJ ACTION
~ 4 MWe — > FACI LITY RE/ACTION ( THEORETICAL):
/ 3C tO, +4 -UO - SCO, + 4H.
Figure 7
134
STATUS AND APPLICABILITY OF SOLID POLYMER ELECTROLYTE TECHNOLOGY
TO ELECTROLYTIC HYDROGEN AND OXYGEN PRODUCTION
W. A. Titterington

Lynn, Massachusetts
PRESENTATION SUMMARY
The General Electric Company solid polymer electrolyte (SPE) water

electrolysis technology is presented as a potential energy conversion
method for wind-driven generator systems. Under development since 1967,
this technology is relatively new, and further significant improvements
are projected over the next 5 to 15 years. Electrolysis life and per-
formance data are presented from laboratory-sized single cells (7.2 in^
active area) with high cell current density selected (1000 ASF) for nor-
mal operation.
The SPE water electrolysis technology has the following inherent

design capabilities as a candidate energy conversion technique in either
small or large scale wind-driven generator systems:
(1) Long life capability to reduce refurbishment costs, with

increased reliability
(2) High current density capability to reduce cell capital costs
and size
(3) High-pressure capability for hydrogen transmission by pipe
line or storage
(U) Minimum power input requirement to reduce operating costs
and generator electrical capacity
Performance data with demonstrated life to approximately 9000 hours

at current densities between 1000 and 1400 ASF are presented. High-pres-
sure life data up to 3000 psi are also presented at the selected nominal
design condition. Based on current technology, projections of cell life
as a function of operating temperature are made which are supported by
life data to 29,000 hours. For the selected design point of 1000 ASF at
180° F, a 60-cell module is sized to produce 5 pounds of hydrogen per
hour with a power input of 112 kilowatts. The module would weigh 135
pounds and be approximately 16 inches in diameter by 8 inches thick.
Present capital cost for a total water electrolysis system is estimated
to be $3000 per pound per hour of hydrogen capacity.
Based on continued development, projected energy and capital cost

improvements are presented up to the year 2000. Energy requirements
135
of 18 to 20 kilowatt-hours per pound of hydrogen are projected for the
1985 to 1990 period, dropping to as low as 15 kilowatt-hours per pound
hydrogen by the year 2000. A capital cost of $785 per pound hydrogen
per hour capacity is considered obtainable for the 1985 period, with
reduction to $250 to $350 per pound hydrogen per hour by the year 2000/
DISCUSSION
Q: I am very curious of these electrolysis systems. What does the

quality of the water have to be,,and what effect does the quality
of the water have on the expected life of the membrane?
A: As I mentioned before, we do use an iron exchange membrane principle.
This is used in other applications to remove contaminants, for ex-
ample, remove iron from water. So any sort of iron that would tie
up the sites within the membrane would affect the performance. The
performance would decrease, or the voltage go up. Its life would
remain the same as long as that contaminant level doesn't increase
with life.
Q: Is iron the only contaminant? What about salts?

A: We have demonstrated chlorine generators, just putting plain salt
water into the system.
Q: Do you have to use distilled water?

A: Yes, you have to use water maybe of the order of 500,000 iron
centimeters.
Q: This is high quality water. If we're considering a wind generator

in an isolated site, it's like a water treatment factory. It's just
a consideration. I'm not unfamiliar with the system.
A: Yes. That should be considered in any trade-off study, too.
Q: You mentioned you have some modern metals. What is it you have?
A: Our anode catalyst is a proprietary type catalyst. It does have
platinum in it. On the cathode side it's a straight platinum cata-
lyst. The loadings are down around 2 to U milligrams per square
centimeter, very low loadings, so that we are now, rather than trying
to reduce the catalyst loading to get the cost down, attacking the
problem from the high current density point of view to keep the size
of the cell down. We have somewhat reached the limit on the present
catalyst systems.
136
SUPERFLYWHEEL ENERGY STORAGE SYSTEM
David W. Rabenhorst
Johns Hopkins University

Silver Springs, Maryland
Until recently, the use of flywheel storage systems has been limited
to a very few applications. The principal disadvantages of these devices
have been the limited energy storage capability (about one-tenth of that
of a lead-acid battery), the poor energy storage efficiency (short run-
down time) , and the danger of catastrophic failure.
Modern technology has provided a tenfold improvement in flywheel

energy storage capability since 1900. There have also been significant
improvements in rotor drag from bearings, seals, and aerodynamic resist-
ance, resulting in greatly improved energy storage efficiency.
Unfortunately, however, the hazard of catastrophic failure of the

conventional steel flywheel has increased, because of the great increase
in the energy of the failed pieces in the high-performance steel flywheel.
Thus, even these higher performance flywheels have been limited to appli-
cations where either adequate failure protection can be provided or
(usually) where the performance can be derated sufficiently to provide
an adequate margin of safety.
This margin typically increases as the flywheel size increases.

For example, the theoretical maximum performance of an optimized steel
flywheel using the best available material is about 26 watt-hours per
pound. Practical limitations reduce this to about 12 watt-hours per
pound for a small, 30-pound flywheel (ref. 1) . In a current program in-
volving a 1400-pound steel flywheel, the rated performance is 6 watt-hours
per pound (ref. 2), while a third steel flywheel weighing 480,000 pounds
is rated at 0.75 watt-hours per pound (ref. 3) .
For the past 3 years the Applied Physics Laboratory has been stud-
ying a new superflywheel concept. It appears to offer greatly improved
safety, and its performance can be better than that of the best optimized
steel flywheel. Its configuration allows sufficient distribution of
failed particles in size, direction, and total time; thus, effective
failure containment appears to be a practical objective.
The use of superflywheel energy storage will considerably enhance

the performance of future onsite energy systems, such as solar and-wind
energy systems. Its chief advantages will be lower total cost and free-
dom of maintenance of the storage system. It will have several times the
operating life of lead-acid batteries, and it will also readily accept
137
high-power peaks associated with heating and air conditioning equipment
and cooking. This same capability to accommodate high power peak loads
makes the flywheel especially attractive for wind power machines, where
peak power can easily range up to several times average minimum power.
The amount of kinetic energy that can be stored in a rotating fly-

wheel is equal to the specific strength of the material used times some
constant related to the geometry of the flywheel. The basic element of
the superflywheel is the thin rod shown in figure 1. A number of these
rods are assembled in a pregrooved hub lamina (fig. 2) so that they fan
out in radial orientation (fig. 3). Thus, the free ends of all of the
rods are in essentially pure tension when the assembly is rotated.
Adjacent layers of hub laminae are assembled 90° in rotation to each
other so as to form the circular brush configuration (fig. 4) .
The failure of any rod represents but a tiny amount of the total
energy in the rotor, and even if all of the rods failed simultaneously,
the failed pieces would be distributed evenly around the periphery; thus,
the stress concentrations are minimized in the containment structure
from the failed pieces.
In contrast, the stress concentrations in the containment structure

caused by the failure of a conventional solid steel flywheel could be
several thousand times as great, since it would (typically) break into
three large pieces, instead of thousands of tiny pieces.
Another advantage of the superflywheel configuration is that it

allows optimal use of filamentary composite materials. These materials
not only exhibit many times more strength to density (hence energy storage
capability) than steel, but they absorb very large amounts of energy
upon failure, as illustrated in figures 5 to 10. A number of 30-inch
long rods about 1 pound each (fig. 5) were spun to destruction in a special
test setup (fig. 6). In each test a steel ring was used to contain the
fragments at failure. From the destruct sequence shown in figures 7 and
8, it can be seen that the rod is completely destroyed before the steel
ring has begun to move from the impact of the failed pieces. The rod
virtually exploded into the dust-sized particles shown in figure 9.
Also, by comparing the shape of the steel ring after the test (fig. 10)
with its other known characteristics, it was established that only about
1% percent of the kinetic energy in the spinning rod reached the steel
ring as impact energy. It would thus appear that the superflywheel brush
configuration offers the first prospect of realistic failure containment
for a high performance flywheel.
There now appear to be about ten different materials that seem to

offer more economical energy storage (W-hr/$) than the lead-acid battery.
Some of these materials are glass, fiber glass, Dupont Fiber B and PRD-49,
music wire, and some new proprietary materials. Thus, a successful super-
flywheel development would provide an energy storage system with the
economy of the lead-acid battery, but without any of its limitations
(maintenance, depth of discharge, low power peak capability cycles to
failure, emissions, low efficiency, dc to ac conversion, etc.).
138
Also, a wind power system using the superflywheel for energy storage
can be considerably more efficient than systems using any other known en-
ergy storage concept. This stems mainly from the fact that the wind
machine energy can be transmitted directly to the flywheel through gears
and shafting at very high efficiency. The flywheel, in turn, can be con-
nected directly to the ac generator without the need for gearing. A
nominal flywheel speed range of 2:1 can be accommodated by several gen-
erator types capable of producing constant voltage and frequency output
under these conditions.
REFERENCES
1. Lawson, L. J.: Design and Testing of High Energy Density Flywheels

for Application to Flywheel/Heat Engine Hybrid Vehicle Drives.
SAE Paper No. 719150, 1971.
2. Superflywheel Meeting at Airesearch, Division of Garrett Corporation,

Los Angeles, Nov. 2, 1971.
3. Superflywheel Meeting, Inductive Energy Storage Task Group, Naval

Research Laboratory, Mar. 13, 1972.
DISCUSSION
Q: Have you found PRD-!49 is better than carbon for your purposes?
A: The one thing I failed to point out is that the most critical thing
is energy storage per dollar, watt hours per dollar. Never mind
amperes per cubic foot or square foot or watt hours per pound or
anything else, except of course safety, which is on top of the list.
PRD-49, at the present time, is about one-tenth the cost of graphite
fibers and also has about the same performance. Therefore, it's ten
times as good, if all other things are equal, and with PRD-U9 they
essentially are. It just so happens there is one material which is
almost a hundred times better than PRD-149. And that happens to be
wood. The strength of wood is about one-tenth the strength of steel;
the density of wood is about one-tenth the density of steel. So the
strength and density are the same. The energy density is the same
as steel; in fact, it's a little bit better - 20$ per pound.
Q: I understand that you use this material because of the tension. The
problem seems to be two-fold as I understand it. The problem is the
angle of the wire. This angle is not safe. Is that the reason why
you choose the brush type?
A: Are you talking about the Gyroscopic forces?
Q: If you wound the wire, then when the angle comes up a problem arises.
The energy density is high for a wound wheel.
A: That sounds like it's true, but it's not. You get more theoretical
energy per space. No one in the published literature has ever
achieved more than about 30 percent of the theoretical energy in a
wound configuration. The reason is very simple. The only place on
that wound wheel where the stress lines up with the filament is the
. • 139
outer edge. Everywhere else there is a radial component, which is
unfortunately a differential radial component with radius and there-
fore will always break in concentric rings. The only way you can
stop this is to add radial filaments. It does turn out that there
are combinations of orthogonal filament arrangements we have patents
on which can be used to make a solid wheel. It is applicable for
some materials like fiberglass, Scotch ply, and so on. In my opinion,
the reasons for doing this are economics versus safety. If you're
building a million pound wheel, you would never build it this way.
This configuration I'm talking about in a million pound wheel would
have no component in it except the hub that I couldn't carry over my
shoulder in one arm.
Q: It is interesting that about 17 years ago I happened to be with the

General Electric Company in the space power work. When we looked at
flywheels then and with the high strength steels that we had, with-
out the benefit of these composite fibers and fiber technologies, I
just looked at the prediction we made then; it was 26.4 watt-hours
per pound.
A: It's about 26 watt-hours per pound maximum now. The Germans are
building a 480,000 pound flywheel, and it's rated at three-quarters
of a watt-hour per pound.
Q: When we speak of the energy inherent in the rotation of a mass like

the flywheel, we very customarily calculate that energy on the basis
of how much is stored on the basis of full rotational speed minus
the zero energy at standstill. Immediately then, there are two ques-
tions. The first of these is we must recognize that this energy is
in a mechanical and not electrical form. In the second place, just
as we can't expect storage batteries to provide us with the full out-
put, in other words, drain them to zero level of content, we at the
same time can't effectively expect all that energy from the flywheel.
So I would like to ask you to address a few comments to the dual
points.
One is how and with what effectiveness, with what degradation if you
will, do we extract this energy on a repetitive in and out basis.
And secondly, how can some sort of a fairly steady-state extraction
of that energy take place, say from the standpoint of non-fluctuation
of the voltage, rpm frequency, or whatever you intend to do with it.
Could you give us a few comments on these?
A: How much energy is left in the wheel is of no consequence since in
this instance that part of the energy never gets taken out. Even if
it were, if I operated only over a speed range of 4 to 1, I can take
96 or 99 percent of the energy out of the wheel. On the question of
mechanical energy versus electrical energy, we do not start with
electrical energy. We start with mechanical energy; all I need is a
contiguous generator of a variable field pole type, for example,
which can accept the 2 to 3 to 1 input'speed range and hold the out-
put frequency precise and the output voltage within the required
tolerances of approximately the percent. Now, if I go directly from
the wind machine to the flywheel, the transmission energy is 100 percent.
140 —
It is not efficiency that I lose, it is a function of how long it
takes the flywheel to spin down. In a rotor the size that would be
adequate for a home installation, Professor Beams at the University
of Virginia had a magnetically suspended rotor (several hundred pounds)
adequate for a home installation with which he measured the decelera-
tion rate of about 1 percent per week in his vacuum container. Now,
somewhere between what he is doing and what is real, live practical-
ity, we believe there is a realizable goal. We see a number of pro-
grams being initiated for the combination magnetic and mechanical
bearings which can achieve a large measure of that efficiency. Now,
to answer your final question, I've gone through many calculations
and I can't get much below 80-percent efficiency from energy in to
energy on the line as opposed to the 30's, MO's, and 50's that you'll
get with every other system. It's that way. There isn't anything
else in the system, whether you use the generator at 90-percent effi-
ciency, and the electric motor to drive it, or if you connect it
directly.
Q: I think I missed one very important point here. We are dealing with
very high rotational speed disks and very slow speed windmills. How
do you envision this coupling? You are not going to drive one of
these disks directly with a windmill without a fantastic gear. How
do you get this flywheel running at the enormous speed necessary?
A: There are two ways you can get the flywheel speed up. In the smaller
systems, in which the size of the flywheel would be (could be) small
compared to the rpm that you want to operate the wind machine in,
you would have to change the speed mechanically by some means:
timing belt, or gears, or rollers, all three are applicable. In the
larger machines, it's much easier just to make the flywheel diameter
compatible with the speed you want.
Q: How do you get this speed differential? You are operating a windmill
at, say, 30 rpm.
A: Well, if you are operating a windmill at 30 rpm you can gear it up
using pulleys, gears, rollers, and the like.
Q: You need a continuously variable speed transmission in order to
accomplish this.
A: You either need a continuously variable speed transmission or you
need something like a variable speed pole generator.
Q: How do you charge mechanically the flywheel? How do you charge at
the various speeds? How do you build up the speed of the flywheel
unless you have mechanical transmission to accomplish that?
A: I'm saying you can do it either mechanically or electrically. It's
the reverse of driving an automobile, if you will. As a matter of
fact, it's exactly like driving an automobile downbrake with regen-
erative brakes on. It's being done all over the world. And you
can, indeed, either mechanically vary a transmission, which I had
in my efficiency calculations (I had an electric variable field pole
generator), or you could use a variable field pole motor.
Q: You have a windmill converting wind to electricity?
A; Definitely, Oh, yes. I'm sorry, I didn't mean to leave that out.
E/ w - 0 /3p'.lu, 2 /2W
W - WEIGHT OF ROTATING ELEMENT (EXCLUSIVE OF

SHAFT AND HUB) (LB)
E/
W • SPECIFIC ENERGY (IN-LB/LB)
(MULTIPLY BY 0.314 X 10"4 TO CONVERT TOW-H/LBt
u - ROTATIONAL SPEED (RAO/S)
0 - STRESS AT ROTATIONAL SPEED u (PSD
p' - MATERIAL WEIGHT DENSITY (LB/IN3)
1 - MOMENT OF INERTIA ABOUT SPIN AXIS (1N/LB/S2)

(FOR THIN ROD, I - R2W/3g)
0 - GRAVITATIONAL CONSTANT - 386 IN/S2)
Fig. 1 ENERGY STORAGE CAPABILITY OF THE STRAIGHT FILAMENT Fig. 2 TYPICAL HUB LAMINA
Fig. 3 HUB LAM IN A WITH RODS Fig. 4 FANNED CIRCULAR BRUSH CONFIGURATION
Fig. 5a INSIDE VIEW OF SPIN CHAMBER
l«l S-GLASS/EPOXY ROD POTTED INTO HUB
(til GRAPHITE/EPOXY TEST ROD
Fig. 5b TYPICAL SPECIMENS FOR NAPTC TESTS
SOLENOID VALVE
CONTAINMENT RING
SPECIMEN
TIE BAR
MECHANICAL
VACUUM
Fig. 6 GENERAL ARRANGEMENT FOR 1-POUND ROD TESTS AT NAPTC, PHILADELPHIA
143
in
>
o: O
O X
x n.
< I-
0
o
O "-
O oc
O I-
X ";
Q. <
I
o
o
01
i
I
i
UJ
O
<
cc >-
X
s
cc
0
145
BATTERIES FOR STORAGE OF WIND-GENERATED ENERGY
Harvey J. Schwartz
.Cleveland, Ohio
Batteries are the one form of energy storage which is familiar to
everyone. What is often overlooked is that they are generally used for
storing relatively small quantities of energy on a widely distributed
basis, perhaps the best example being the automobile starting battery.
An estimated 50 000 megawatt-hours are currently stored in automobile
batteries alone. Storage of wind-generated energy is similar in concept,
involving fewer but much larger storage units. Batteries will be used
for this purpose if they are cost competitive with other storage systems.
Due to time limitations, I will forego any discussion of how batteries
convert chemical energy into electricity; I will concentrate instead on
why batteries should be considered, what factors influence their costs,
and a brief summary of the state-of-the-art of the most likely candidate
systems.
Figure 1 shows the reasons batteries are considered for energy
storage. Batteries are attractive because they are simple, easy to use
devices which require no complex facilities and little repair or main-
tenance during their operating life. They can be built in convenient
packages and are free of the geographic constraints found in pumped water
or gas storage systems. They produce no harmful emissions and are avail-
able for use on an almost instantaneous basis.
The factors which affect the costs of battery storage systems are
summarized in.figure 2. Costs of batteries are largely determined on
how they are used. One obvious factor is the size or total quantity of
energy which must be stored. This will be fixed by the power to be de-
livered and the maximum length of windless period during which the battery
is expected to operate. The next factor to be determined is life. This
will be affected by the total number of operating cycles, the rate at
which the battery is charged and discharged, and the depth of discharge,
or fraction of the total energy removed in a cycle. Since designs aimed
at maximizing lifetime also result in higher initial costs, it will
probably be necessary to optimize the battery for minimum cost for a par-
ticular installation.
In general, three classes of batteries are considered for bulk energy
storage - conventional types, metal-gas batteries, and high energy density
alkali metal types. Figure 3 summarizes the characteristics of the con-
ventional types most often considered. Three batteries seem suitable.
The lead-acid battery is the standard for comparison. For this service
the energy density, which measures the size of battery required to store
a given quantity of energy, is 10 watt-hours per pound. The power density,
measuring its ability to deliver high current, is 20 to 30 watt-hours per
pound. Batteries of this type are good for about 1500 charge-discharge
cycles and cost about $80 per kilowatt-hour. It does not appear that this
cost will be any lower in the future as this is a mature, cost-conscience
manufacturing industry. An updated version of the nickel-iron battery
is under development; it is expected to deliver 25 watt-hours per pound
and 50 watts per pound. Cycle life is unknown, and a cost close to the
lead-acid battery is projected. Since this battery produces substantial
amounts of hydrogen on charging, reduced current efficiency and the need
for frequent water additions result. The only other current competitor
to lead-acid is the nickel-zinc cell. Substantial performance gains at
comparable costs are expected, but the cycle life is only 200 to 400 cycles.
In summary, at present no conventional battery appears able to compete suc-
cessfully with the lead-acid battery for bulk storage.
Metal-gas batteries, shown in figure 4, have attracted attention

because they promise at least a 4 to 5 improvement in energy density over
the lead-acid battery. Zinc-air and iron-air cells offer the possibility
of one free reactant which should reduce cost. Nickel-hydrogen is of
interest because it makes use of two stable electrodes and should deliver
long cycle life. Since air contains carbon dioxide which can reduce the
life of air batteries, work on oxygen electrodes coupled with zinc or
cadmium has been carried out. Each of these combinations requires air or
oxygen electrodes which use precious metal catalysts. These offset the
economic advantage of using air. Lifetimes measured from hundreds up to
one or two thousand cycles are the best reported, so improvement is needed
in that area. An attractive newcomer is an unusual zinc-chlorine battery
built by Udylite Corporation to power an electric car. Chlorine is stored
as a stable solid compound, chlorine hydrate, at temperatures below 10° C
which eliminates the need to handle and store gaseous chlorine. Raw
materials costs are low (16£/lb for zinc and 3%£/lb for chlorine), and
inexpensive carbon can be used for the chlorine electrode. Life is un-
known, but this system may have the best near term chance to replace
lead-acid.
Exotic alkali metal batteries like those in figure 5 have received

much attention in recent years. Energy densities of 100 watt-hours per
pound and power densities of 100 watts per pound appear reasonable, and
raw materials are plentiful and cheap. The most advanced is the sodium-
sulfur battery which runs at 300° C and uses sodium beta-alumina, a
ceramic-like sodium ion conductor as the solid electrolyte. Life has so
far been limited to 2000 cycles or less. Even with cheap materials,
costs of $10 to $30 per kilowatt-hour are expected. A substantially
lower cost may be possible if a concept under development by Dow Chemical,
which uses fine hollow glass capillaries as the electrolyte, can be
brought to fruition. Argonne National Laboratory has pioneered another
high temperature battery which uses lithium and sulfur. This system has
suffered from severe corrosion problems and apparently will require ex-
pensive materials of construction. A lithium-chlorine battery development
by Sohio has been unsuccessful. Only one large complete battery of this
class has been built, a 30 kilowatt — 30 kilowatt-hour sodium-sulfur
battery to power a van. In general, these advanced systems are expected
to require at least 10 years and $30 to $40 million worth of development
to reach the point where they are ready for large-scale use.
Batteries work. The role they will play in wind power cannot be
determined until a detailed analysis of the storage requirements of wind-
generated energy systems is made.
DISCUSSION
Q: You mentioned $80 per kilowatt-hour for the lead-acid battery cost.
I wonder if you could tell what's involved in that cost estimate? Also,
1 wonder if you have any idea what the efficiency of the lead-acid
battery is?
A: Well-, in answer to your first question, the cost I spoke of is the
cost of the battery alone. That's about what it costs to buy com-
mercial, industrial grade, lead-acid batteries, and it's probably
as low as that cost figure is going to get. In answer to your second
question, the energy in to energy out is a little more difficult,
because you have to look at more than the battery. It depends on
whether your wind system is driving an ac machine. If so, you're
going to have to convert it to dc and use that to charge the battery;
then you will have to take the dc out and convert it back to ac. If
you can use dc power and produce dc power with your windmill, then
your efficiency is going to be better. In that case it's probably
going to be of the order of, oh, I'd guess about 60 to 70 percent.
It depends on how fast you are doing the charging, and what your
inefficiencies are. Without a specific design and a specific rate,
it's a difficult question to answer. It will not be 100 percent.
Q: Which batteries are amenable to scaling to very large sizes?

A: That's a good point; I meant to bring it out and I forgot to.
Q: There is an auxiliary question here: where is the crossover point in

shifting from very large batteries to the hydrogen-oxygen fuel cell
group?
A: Let me first answer the first question. The one characteristic of a
battery that you have to remember is that it does not scale well; a
2 kilowatt-hour battery tends to weigh about twice as much and cost
almost twice as much as 1 kilowatt-hour battery. There is a scaling
factor in practical cell sizes, but it's not like a piece of machinery,
for instance, in which you can double the power by increasing the size
of the wheel a very small amount.
That is the scaling factor for batteries is nearly linear in terms

of the amount of energy stored. This is why you find batteries used
and why they will continue to be used in places where at the present
cost level relatively small quantities of energy are stored.
Now let me answer your second question. Batteries do not scale in the
sense of rotating machinery where the physical size only changes a
small amount for a much larger increase in output. In the past I
worked on the SNAP-2 project where mercury turboalternators and a
SNAP-2 alternator produced a few hundred watts.
If you go up to SNAP-8 and you're talking 30 kilowatts, the system

gets a little bigger, but not 15 times as large. Batteries tend to
scale more linearly.
Concerning the crossover point, I think that's an economic considera-

tion. Most installed costs I've seen projected for fuel cells tend
to be high; for instance, for a 3-megawatt system based on an acid
electrolyte fuel cell, the best figure I've seen is $145 per kilowatt
installed. Now you're talking power in one case and energy in the
other. If you want to compare the fuel cell, then you have to com-
pare the storage tank as well for some given quantity of energy.
Q: I would like to mention two aspects which often get overlooked. First,
I don't think you can say the emissions are zero when the efficiency
that you point out is 50 or 60 percent. There is one heck of a lot of
heat that has to be accommodated, especially in a 1,000,000-pound bat-
tery. Actually, I have seen a 7,000-pound battery in a Mercedes bus,
and it had a complete air conditioning system that goes along with it.
Second, when we think of costs, we must think of costs for the applica-
tion we are considering. I would be very surprised if we're not talk-
ing about 15- to 30-year life systems. If we're talking about a 30-
year system, you had better multiply your cost factor of 6, according
to your own numbers.
A: That's right.
Q: I would like to mention something about the batteries. We have demon-

strated the performance characteristics of at least lead-acid battery
powered systems. Some of the things you have in your slides run into
a very difficult problem, which is creeping up on us very rapidly.
This is the materials availability and cost problem. Of all the ma-
terial you would want to use in your battery, I would say lead, zinc,
and copper are the three most critical raw materials that face us
today in terms of price escalation and availability. And of those
three, copper and lead have an awful lot of recycle potential and
zinc has virtually no recycle. Zinc used in our economy is mostly
for corrosion protection. As such, it is sacrificed, and therefore
not recoverable. While it would appear that zinc air or zinc chlorine
might be a promising candidate for wide-scale use in applications,
there could be a real material problem. I think that's another factor
that we have to look at very carefully. And for that reason I view
with a considerable amount of optimism, if we're going to use batteries,
the sodium approach, which is at least one metal that is very energy
intensive. ,1 think we have to look at the availability of materials
much more with batteries. We ought to also address the question of
material costs. It's one place where it is proportional to the energy
and power usage: twice as much power, twice as much mass. We also
have to worry about the competitive uses of these fairly scarce materials.
If I may make a short answer to your question, I try to stay away
from the subject of electric vehicles although it's near and dear to
my heart. I'm going to give the keynote address at the Electro-
chemical Society's fall meeting on batteries for electric vehicles.
I think it's a tremendous application, but when you begin talking
about power in the megawatt hour scale, I'm not sure our experience
in electric vehicles is really appropriate here. It's a whole new
ballgame. None of us, I really feel, knows a great deal about it.
Your comments on materials availability are well-taken. The cost
estimates on nickel-zinc batteries, for instance, have been done by
battery manufacturers and are based on recycling zinc plates in the
manner in which they recycle lead-acid batteries now. But look at
the vehicle situation, for instance. It would be impossible in this
country to convert all the vehicles we have on the road at the
present time to lead acid because we simply haven't got enough lead.
That's not an answer for hundreds of millions of vehicles, and I sus-
pect it's also not an answer for power in the scale we're talking
about here. It's a good point.
A D V A N T A G E S OF B A T T E R Y
ENERGY STORAGE
COSTS ARE INFLUENCED BY
SIMPLE
- TOTAL E N E R G Y S T O R E D
EASILY MODULARIZED
- O P E R A T I N G LIFE
N0
NO S P E C I A L S I T I N G R E Q U I R E M E N T S ' ' C H A R G E - D I S C H A R G E CYCLES
- RATES OF CHARGE AND DISCHARGE
NO E M I S S I O N S - DEPTH OF D I S C H A R G E
INSTANT S T A R T U P
Figure 1 Figure 2
150
CONVENTIONAL BATTERY PERFORMANCE
PERFORMANCE CYCLE PROJECTED

SYSTEM PROBLEMS
WH/LB W/LB LIFE COST
LEAD-ACID 10 20-30 1500 $80/KWH
NICKEL-IRON 25 50 ? $100/ KWH G A S S I N G , MAINTE-

NANCE, EFFICIENCY
NICKEL-ZINC 30 150 200-400 SAME AS LIFE

LEAD-ACID?
Figure 3
METAL-GAS B A T T E R Y PERFORMANCE
PERFORMANCE
SYSTEM WH/LB W/LB PROBLEMS
IRON-AIR 40-50 10-20 CATHODE C O R R O S I O N , LIFE
ZINC-AIR 40-50 10-20 LIFE, COST
NICKEL-HYDROGEN 30-40 ? VOLUME, LIFE
ZINC-OXYGEN 50-60 10-30 LIFE, COST
CADMIUM-OXYGEN 30-40 ? LIFE, COST
ZINC-CHLORINE 50-75 40-60 LIFE
Figure 4
ALKALI METAL-HIGH T E M P E R A T U R E B A T T E R Y PERFORMANCE
PERFORMANCE CYCLE
SYSTEM PROBLEMS
WH/LB W/LB LIFE
SODIUM-SULFUR 80-100 80-100 200-2000 LIFE, C O S T S

(BETA A L U M I N A )
SODIUM-SULFUR 80-100 80-400 100 + LIFE, MATERIALS

(GLASS) STABILITY
LITHIUM-SULFUR 100 > 100 2000 MATERIALS CORRO-

SION, C O S T S
LITHIUM CHLORINE 50 » 100 100 LIFE

(CARB-TEIW
Figure 5
151
ENERGY STORAGE BY COMPRESSED AIR
George C. Szego
Intertechiiology Corporation
Warrenton, Virginia
I/would like to discuss with you some of the general attributes of

the compression of air as a means of storing what I call off-peak, cen-
tral station, baseload power for peak use. But exactly the same concept
can be used to store wind power.
There are essentially three components to the system: a compressor,

a motor, and a turbine. To store the energy, a motor drives the compres-
sor, which, of course, compresses the air. To extract the energy, the
air is run through the turbine, which drives that same motor, which is
now an alternator. This is the same situation as the pump-turbine and
the motor-alternator in a pumped hydro system.
In this system the compressed air is stored underground in caverns

or aquifers. The use of caverns requires a water piston and a surface
lake to recover flow-work, the PV term, which normally isn't recovered
when pumping into an inflexible tank, for example. It also requires the
use of surface.- area. I like very much better the use of an aquifer to
store the compressed air because the water in the interstitial spaces
would act as the piston and thus no surface area would need to be used.
If you used pumped hydro to store wind energy, you'd get about 0.67
efficiency. If you put in 3 kilowatt-hours, you get put 2 kilowatt-hours.
The use of pumped hydro also entails an almost $200 per kilowatt capital
investment, substantial land use, and the inability to put it where you
want it. Suitable sites are usually far from load centers; therefore,
transmission and the capital costs of transmission are involved.
Now, a normal gas turbine system uses up about three-fourths of the

total output of the turbine in the compressor; therefore, a 1-kilowatt
gas turbine system normally is a 4-kilowatt turbine, a 3-kilowatt compres-
sor, and a 1-kilowatt alternator. (By the way, those systems cost about
$110 or $115 per kilowatt. They are enormous spendthrifts of energy, of
fossil fuels specifically. Their heat rates are near 17 000 Btu per
pound and up. They have the advantage, though, of quick installation,
which is why they are used widely by the utilities.) Obviously, it is to
our advantage to increase the pressure ratio. Presently, we are considering
a 40 to 1 compressed air to atmospheric air pressure ratio for our com-
pressed air storage system.
There is an additional factor. When you operate a separate turbine
152
and compressor for some tank storage system, be it underground cavern,
surface tank, or the aquifer, you have two choices:
(1) You can extract the air from storage and run through the tur-
bine cold, in which case the performance is exactly the same as pumped
hydro, that is, 0.67 efficiency or 3 kilowatts into storage and 2 kilo-
watts out.
(2) Or you can burn a small amount of any fuel and heat the air
before it enters the turbine.
If this is done to about 4000 Btu per kilowatt-hour, roughly 40 percent
of the normal heating rate, the output of the system doubles at a fairly
small cost. In other words, 3 kilowatt-hours in and 4 kilowatt-hours
out. This is an apparent efficiency of 133 percent, but, of course,
you're expending some energy (heat) to get it.
With respect to the combination of windpower and compressed air
storage, I hesitate., without making a detailed technical and economic
analysis, to say why this is an applicable concept. Certainly it is
physically feasible. If there were a battery of wind machines in a
given area with an installed area output of, say, 50 or 100 megawatts, I
believe the underground storage of compressed air would be the most'
attractive concept that you could consider.
Another very important characteristic of the underground storage
of air is its unique flexibility. In pumped hydro, in a battery, or in
a flywheel, when you are up to full storage thatTs all there is - there
isn't any more. And what you have depends on how much money you spent.
Air, on the other hand, being a compressible fluid, is quite flexible.
For instance let's say we've stored 2 or 3 days worth of power, or air,
at 600 pounds per square inch. If we chose to store a week's worth,
which surely you cannot do with pumped hydro (pumped hydro is only
stored overnight because it is so expensive and because it is inflexible)
you can simply continue compression to perhaps, 650 pounds per square
inch. The air will be pushing the aquifer up closer to the dome, and
you will be getting the piston action simply by the air being more
compressed. Because air is a compressible fluid, more energy can be put
into it by increasing the pressure, or by pushing back more of the
interstitial water in the aquifer.
This storage system concept has the reheat flexibility. It has the
lowest capital cost of any storage system of which I'm aware, and certain
beneficial environmental advantages that includes not using surface
area.
DISCUSSION
Q: Why 50 megawatts as the lower limit cutoff?
A: Well, the reason for that is fairly simple. You've got to sink some
wells, for both the cavern and aquifer storage systems. A rough
trade-off analysis indicates, at least for commercial utility use,
153
that the better part of a hundred megawatts is necessary to make the
machinery and the attendant structures pay.
COMMENT: My calculations show that the type of flywheel I am considering

is comparable in cost and size to your system.
Q: Is there any reason why this power storage system could not be used
in off-shore locations several miles off the continental shelf?
A: No, and that opens up an entirely new opportunity not present on land.
The ground under the water can, of course, be used. You can also use
a membrane or bag lying at the bottom of the water or at a suitable
depth. Pump the air into it, and let the water pressure push it back
up to you. The use of the membrane is possible only in deep water.
Q: What if aquifers are needed for other purposes, like furnishing water?
A: I don't think I'm a sufficiently good geologist to answer that. But
I'll try. I think the aquifer itself would resolve that issue. There
are, I believe, few fresh water wells that go down to 2000 feet. So
aquifers that are 175 feet down might be used as wells or for water
storage, and those 2000 feet down for power storage.
EXPERIENCE WITH JACOBS WIND-DRIVEN ELECTRIC
GENERATING PLANT, 1931-1957
Marcellus L. Jacobs
Jacobs Wind Electric Company, Inc.

Fort Meyers, Florida
This report outlines the engineering, construction, performance,

electric output, and different uses of the Jacobs wind electric 2500- to
3000-watt plant, thousands of which were installed in many parts of the
world between 1931 and 1957.
Early engineering started on this wind-operated electric generating

plant in 19251 After several years of testing different types of wind-
mills, the three-blade aeroplane type of propeller was found to be far
superior in power output. By means of a flyball-governor-operated,
variable pitch speed control, the maximum speed of the propeller was
accurately and easily controlled, to prevent excessive speeds in high
winds and storms. The three-blade propeller was found to be necessary
(as compared to the two-blade type) to prevent excessive vibration when-
ever the shift of the wind direction required the plant to change its
facing direction on the tower.
The periods of vibration which occurred on the two-blade propeller,

every time the tail vane shifted, to follow the changes in wind direction,
were found to be caused by the fact that the two-blade propeller, when in
a vertical position, offers no centrifugal force resistance to the hori-
zontal movement of the tail vane in following changes in wind direction.
However, when the two-blade propeller is in the horizontal position, its
maximum centrifugal force is applied to resist horizontal movement of
the tail vane; thus the tail vane is forced to follow wind direction
changes by a series of jerks, causing considerable serious vibration to
the plant.
The three-blade propeller was developed by us in 1927 to correct this

condition. When in operation, the three-blade propeller creates a steady
centrifugal force resistance, against which the tail vane reacts with a
constant pressure and produces a smooth shifting horizontal movement of
the plant facing direction. The centrifugal force generated by the very
light aeroplane spruce-wood blades, when operating at 225 rpm is 550
pounds each, making a force of over 1600 pounds of gyroscopic resistance
force to the horizontal vane movement for the three blades. But this
resistance is in the form of an even pressure or resistance to horizontal
movement, whereas the 1100 pounds of gyroscopic resistance force of the
two-blade propeller to the vane movement is applied and then eliminated
twice during each revolution.
155
A propeller diameter of 15 feet was found to produce ample power for
electric generator operation to develop 400 to 500 kilowatt-hours per
month, based on the available winds in most areas of the states in the
western half of the United States. This required 10 to 20 mph winds for
2 or 3 days per week. A specially designed six-pole battery charging
type shunt generator was developed to operate at a speed range from 125
to 225 rpm for direct connection to the governor hub of the propeller.
It was designed so that its load factor would exactly parallel the power
output curve of the wind-driven propeller when operating in the 7 to 20
mph range that it was felt to produce the most hours of wind per month.
Wind plants that require higher than 20 mph winds to deliver their rated
output will find too many areas where there are too many days with winds
below that speed each month, and thus their effective average monthly
output in many areas is below expectations. The generator weighs 440
pounds with a 9-inch-diameter armature with a 9-inch core length. The
60 pounds of wire on the field poles gave maximum efficiency with a drain
of less than 100 watts for field coil operation. The generator output is
2500 watts at 32 volts, and, for the 110-volt generator, it is rated at
3000 watts.
Our experience with plants installed in many parts of Alaska, Canada,

Finland, northwestern United States, and a number of special installations
such as the plant we have installed for the joint operated United States
and United Kingdom weather station at Eureka, in the Arctic Circle, and
with the Byrd Expedition at Little America has shown that aluminum painted
(copper edged) spruce-wood propellers have considerably less trouble with
frost and ice formation than when they are varnished or when other type
coatings are used.
Generators located on high-steel towers are subject to considerable

static discharge from the armature through the ball or roller bearings,
and excessive charges from nearby lightning will often arc through a
bearing and weld spots on the balls and race, causing it to break up soon.
We found the revolving propellers collected discharges into the direct
connected armature and the lightning pick-up effect of the propellers
was frequent and of considerable intensity. To correct this, we installed
dual sets of heavy grounding brushes on the armature shaft which com-
pletely eliminated any trouble from this cause. With the additional use
of a large capacity oil-filled condenser connected across the generator
brushes and frame, we practically eliminated any damage to the generators
from lightning, so much so that, with high grade ample insulation used
throughout the generator and the grounding brushes and condensers, we
gave an unconditional 5-year guarantee with every generator against burn-
out from any cause and have built many thousands during the past 20 years
using this construction without any replacements ever being required
because of lightning damage or burn-out from any cause.
The price received at the factory for our 2500-watt, 32-volt plant
was $490, less the cost of a suitable tower and batteries, which could
often be secured in the country or area to which the plant was shipped.
We supplied a 21 000-watt-hour glass cell lead-acid type of storage bat-
tery with a 10-year guarantee, for which we received $365. A fifty-foot
156
self-supporting steel tower was supplied for $175, making a total cost
for the plant of $1025. This is about $400 per kilowatt as the manufac-
turing cost of the plant. Shipping and installation costs are additional.
Installation cost requires only the labor of two men for two days and a
small amount of cement to put into the anchor holes when the tower is
built. No special equipment or training is necessary. We have shipped
hundreds of plants to most countries with not a single request for addi-
tional information to enable them to erect the plant. Regular installa-
tion and operating instructions are prepared and sent with each plant.
Operating and maintenance costs of this plant are largely limited

to the replacement of the storage battery which, on a 10-year basis, is
about $36 per year; from records kept of more than 1000 plants over a
10-year period, the maintenance cost of repairs was less than $5 per year.
Some of the owners of our plants bought the Edison type battery and after
20 years are still using the same battery. New batteries of this type
are quite expensive, but these owners bought second-hand batteries which
still gave them 20 years of service.
Special generators designed for the cathodic protection of underground

steel pipelines were developed by us in 1936. These generators were wound
for an external circuit resistance of 1/10 ohm or higher. The generators
produced 10 volts at 100 amperes and were straight shunt wound. When con-
nected to the pipelines in any normal wind, they maintained a pipe-to-soil
potential of 3/10 of a volt pipe negative. Due to the action of the cur-
rent, the pipe maintained a fair degree of protection through calm wind
periods. Hundreds of our plants are protecting many miles of pipelines
in North and South America and in Arabia. Some of these plants have
been in service since 1937.
DISCUSSION
Q: Can you tell me the present state of this design? You say you are
no longer manufacturing wind generators, but are the designs avail-
able?
A: Well, I closed the plant and sold the machinery. I still have the
company, but the engineering I do is a different type of engineering
now.
Q: Are these designs available if another company is interested in

producing it?
A: Frankly, it's been 18 to 20 years since I last produced wind gener-
ators, and I haven't made much effort to keep them. I'm busy with
environmental work, developing a system for cleaning up coastal
canals and waters (I have patented and developed a system for that),
so I have dropped out of the wind electric business. Now, there are
a lot of old plants still running here and there around the country,
but no new ones. I no longer have the plans, blueprints, or infor-
mation on them. I didn't keep them.
Q: Would you have any guess as to what these units would cost today in
per kilowatt?
157
A: They would be about twice what they were when we quit building them.
Q: That's a complete system?

A: That's the plant, tower, and suitable storage battery.
Q: We had earlier a very interesting discussion on the question of

electric plants. It would be of interest if you could comment on
the operation of such gear.
A: Early in the thirties, about 1931 or 1932, I made a series of tests,
and we put a special grounding brush on the generators. We had found
that the airplane spruce propellers with the copper leading ends and
the static pickup, out in wind and sand and from certain atmospheric
conditions, created a static buildup in the armature, which would
jump across to the main frame through the ball bearings and would
wreck and damage the bearings.
And then I discovered in 1932, that by putting a set of heavy
grounding brushes on the big armature shaft, which is 2 inches in
diameter, that eliminated that completely. After that no bearings
ever went bad and there was no more static buildup.
158
REVIEW OF THE WINDPOWER ACTIVITIES AT THE
BRACE RESEARCH INSTITUTE
T. A. Lawand
Brace Research Institute

MacDonald College of McGill University
Quebec, Canada
The Brace Research Institute was set up in 1961 to alleviate the
problems of water scarcity in rural areas, with particular reference to
improving the productivity in these regions. One of the criteria estab-
lished from the outset was that as much emphasis as possible be placed
on the development of natural energy, local material, and human resources
so as to integrate this technology into the indigenous infrastructure.
As power is always needed to satisfy man's requirements for new or addi-
tional supplies of water, it was decided that windpower should be harnessed
as a possible energy source.
Historically, windpower has provided much of man's power needs for
pumping, mechanical power, and navigation. The Institute has maintained
a basic program of information accumulation, contacting as many of the
existing windpower manufacturers as possible. It has invested heavily
in library facilities and possesses an excellent series of cross referenced
reprints in several languages accumulated over the past 12 years. Since
1963 the Institute has'offered courses in wind power technology and
utilization. The initial courses were formulated under the direction of
the late Professor E. W. Golding of the Electrical Research Association,
United Kingdom, under whom this author had the pleasure of studying some
10 years ago.
Initial activities in windpower were undertaken at the Brace Experi-
ment Station in Barbados located in the heart of the Trade Winds. Chrono-
logically, the various phases of windpower work were as follows:
1962-66 Installation and testing of a 1-kilowatt Quirk windmill at
Springhead, Barbados - DC current produced (DT.4).
1963-64 Testing of a 9-kilowatt Andreau windmill. In this novel
propeller design, air was forced out of the periphery of hol-
low blades, operating an electric generator in a central tower
(T. 12) .
1964 Development of a simple, do-it-yourself Savonius rotor windmill
(L.5).
159
1965-67 Development of the 10-horsepower Brace Prototype Windturbine.
This unit is still in use in an irrigation project in Barbados
(MT.7, R.38, CP.19, and CP.20).
Since mid 1967, the principal activities of the Institute have

passed to the current headquarters in Montreal, Canada.
The following activities have continued in Barbados, Haiti, Montreal,

and elsewhere:
1968-69 The 10-horsepower prototype windturbine was evaluated in

Barbados in its role of irrigating land intermittently. This
function was performed quite successfully from a technical,
agronomic, and soils point of view (MT.8).
1970 Savonius rotors were introduced in Columbia.
1970 A 1.5-meter-diameter Lubing Maschinenfabrik. windmill was tested

in Haiti. This mill, still in operation, is used to pump
saline water to a solar distillation plant.
1971-72 Improvements were undertaken to the design of the Brace 10-

kilowatt prototype windturbine.
1971-72 A permanent magnet alternator generator, driven under variable

input power to simulate wind regime was tested. Loads were
induction motors which performed quite successfully (T.68).
1972 A Lubing Maschinenfabrik 400-watt wind electric generator

(blade diameter = 2.2 m) was tested at Montreal. This unit,
currently being examined for performance under winter condi-
tions, powers an experimental solar/wind powered house (T.75).
1973 Designs were undertaken of reinforced concrete towers as well

as reinforced concrete block towers for windmills (EP.2) .
1973 A study has been initiated on developing a mathematical model

which can describe the performance of any known wind electric
generator whose characteristics are known, given a measured
wind regime.
1973 The Brace Windturbine designs were optimized technically and

structurally. Final designs were submitted to manufacturers
for commercial production.
It can be seen that the programs of the Institute are continuing in

an active vein. Future studies are being formulated to set up free
wheeling windmills whose power output will be utilized through improved
electrical and electronic systems.
In addition, aerodynamicists will examine problems relating to the

improved forms of wind energy conversion to mechanical shaft power.
160
There is a significant potential for windpower utilization in the
north of Canada, where the remoteness of the loads favor small, autonomous
installations. A reexamination of equipment destined for use in warmer
climes is planned so that windmills can perform adequately under these
difficult climatic conditions.
The future of windpower both at home, and in the developing coun-

tries, seems brighter and more promising than in recent times.
LIST OF BRACE RESEARCH INSTITUTE PUBLICATIONS ON WINDPOWER
Publica-
tion No.
R.25 Implications of the Utilization of Wind Power for the Develop-

ment of Small Caribbean Communities, by R. E. Chilcott, Pro-
ceedings of the Conference on Technology and Development in
Small Transitional Societies, Trinidad, W. Indies, 12 pp.,
Feb. 1968.
R.31 Putting the Wind to Work, by the Brace Research Institute,

Engineering, Vol. 206, No. 5353, pp. 760-761, Nov. 22, 1968.
R.38 Notes on the Development of the Brace Airscrew Windmill as a

Prime Mover, by R. E. Chilcott, The Aeronautical Journal of the
Royal Aeronautical Society, Vol. 73, No. 700, pp. 333-334,
Apr. 1969.
R.39 The Potential for Medium Power Wind Turbines in Canada and the
Caribbean, by R. E. Chilcott and G. T. Ward, Paper No. D.50.853
AGM, presented to the Engineering Institute of Canada Annual
General Meeting, Vancouver, British Columbia, Sept. 10, 1969.
R.47 Potential for Medium-Power, Fixed-Pitch, Variable-Speed, Air-

screw Wind Machines in Canadian Agriculture, by R. E. Chilcott
and G. T. Ward, presented to the Agricultural Institute of
Canada - Canadian Society of Agricultural Engineers, Hamilton,
Ontario, Jun. 27, 1968.
T.12 Performance Test of an 8-Meter Diameter Andreau Windmill, by

A. Bodek, 24 pp., Feb. 1964.
T.36 The Economics of Wind Powered Desalination Systems, by T. A.

Lawand, 56 pp., Jun. 1967, revised Sept. 1967.
T.37 Notes on the Development of the Brace Airscrew Windmill as a

Prime Mover, by R. E. Chilcott, 7 pp., Sept. 1967 (superseded
by R.38).
T.38 Notes sur 1'Utilisation de 1'Eolienne Rapide Brace comme Source

Motrice, par. R. E. Chilcott, traduit par M. Lantagne, 11 pp.,
161
juillet 1968: French translation of T.37.
T.41 Investigations Concerning Savonius Rotors and Related Turbo-

machines, by G. Bach, Forschung auf dem Gebiete des Ingenieur-
wesens, 2_, No. 6, pp. 218-231, Jun. 1931, translated from
German into English by G. T. Ward, Jun. 1964.
T.42 Practical Experience Gained from the Development of a 100 kw

Wind Power Installation, by U. Hutter, Brennstoff-Warmekraft,
16, No. 7, pp. 333-341, Jul. 1964, translated from German into
English by G. T. Ward, May 1966.
T.43 Specifications of the Brace 10 hp Airscrew Windmill, 15 Assem-

bly and 70 Detail Drawings, by R. E. Chilcott, H. P. Budgen,
0. Goldstein, R. H. Weyts, Feb. 1971.
T.56 Current State of Windpower Research in the Soviet Union, by

N. Levy, edited by G. T. Ward, 10 pp., Sept. 1968.
T.68 A Simple Electric Transmission System for a Free Running Wind-

mill, by T. H. Barton, and K. Repole, Aug. 1970, 29 pp.
T.75 A Report on Preliminary Testing of a Lubing Windmill Generator

(M022-3G 024-400) of the Brace Research Institute by H. L.
Nakra, 5 pp.
L. 5 How to Construct a Cheap Wind Machine for Pumping Water, by

A. Bodek, 12 pp., Feb. 1965. Revised Feb. 1973.
1.26 Proposal Submitted to the Freedom from Hunger Campaign, F.A.O.,

for the Development of a Low Cost Two-Bullock Power Wing-Rotor
Wind Machine for Water Pumping in Underdeveloped Arid Areas,
by G. T. Ward, 10 pp., Nov. 1963.
1.36 Notes on the Selection of a Suitable Water Pumping System for

the Greenland Windmill, by T. A. Lawand, 8 pp., May 1967.
1.40 Report on Visit to St. Kitts-Nevis, Feb. 12-17, 1968, to

Initiate a Windmill Water Pumping Project, by R. E. Chilcott,
7 pp., Feb. 1968.
1.45 Proposal for the Establishment of a 10 hp Windmill Water Pump-

ing Pilot Plant in Nevis, West Indies, by R. E. Chilcott and
E. B. Lake, 8 pp., Jun. 1968.
1.51 L'Utilisation de 1'Energie Eolienne dans la Vallee du Saint-

Laurent: Etude des Donnees Meteorologiques Disponibles, par
M. Lantagne et G. T. Ward, 7 pp., aout 1968.
1.52 Wind Power Study of the St. Lawrence Lowlands: Survey of Avail-
able Wind Data, by M. Lantagne and G. T. Ward, English transla-
tion of 1.51, 7 pp., Oct. 1968.
162
1.76 Comments on Brace Research Institute Windmill by A. Wilson,
Ministry of Overseas Development, Bridgetown, Barbados,
Feb. 1968, 2 pp.
CP.19 Estimation of Maximum Drag and Root Bending Moment on Stationary

Brace Airscrew Windmill Blade and Variation of Airscrew Starting
Torque with Blade Angle, by P. K. Ghosh, 13 pp., May 1969.
CP.20 Low Drag, Laminar Flow Aerofoil Section for Windmill Blades,
by P. K. Ghosh, 8 pp., May 1969.
CP.48 Potential for Wind Power Development in Eastern Canada, by

M. S. Kadivar, 8 pp., Jan. 1970.
CP.59 Correlating the Drainage Needs with the Availability of Wind

Power in Quebec, by Keith Manuel, 13 pp., 1968.
DT.4 Wind Electric Research Report (Evaluation of the Performance of

a Commercial Aerogenerator), by C. Sanchez-Vilar, 30 pp.,
Aug. 1963.
DT.7 Comparacion de Diferentes Teorias para el Calculo de la Per-

formance de Molinos, by M. A. Nevot, 51 pp., Aug. 1966.
DT.8 Potential of Wind Power Utilization in the Arctic Environment,

House Heating, by C. Bettignies, 40 pp., Aug. 1970.
MT.7 The Design Development and Testing of a Low Cost 10 hp Windmill

Prime Mover, by R. E. Chilcott, M.Sc. Thesis, Dept. of Agricul-
tural Engineering, McGill University, 126 pp., Jul. 1969.
MT.8 The Field Performance of a Windmill Powered Sprinkler Irrigation

System, by J. M. lonson, M.Sc. Thesis, Dept. of Agricultural
Engineering, McGill University, 96 pp., Jul. 1969.
MT.9 Dynamic Analysis of High-Speed Wind-Turbine Systems, by J. S.

Duggal, M.Sc. Thesis, Dept. of Agricultural Engineering, McGill
University, 74 pp., Jan. 1971.
MT.10 Potential for Wind Power Development, by M. S. Kadivar, M.Sc.

Thesis, Dept. of Agricultural Engineering, McGill University,
171 pp., Jul. 1970.
EP.2 Windmill Tower Design, by Frank Montesano and Antonio Fernandez,

Project for the Department of Civil Engineering and Applied
Mathematics, Course No. 303-318 B, Apr. 18, 1973, 92 pp.
163
\ ROTATION
WIND
OUTLET
TO SPRINKLERS
VIEW FROM UPWIND SIDE ELEVATION
PROTOTYPE BRACE AIRSCREW WINDMILL PUMPING SYSTEM.
Figure 1
WIND POWER SYSTEMS FOR INDIVIDUAL APPLICATIONS
Henry M. Clews
Solar Wind Company

East Holden, Maine
Mr. Chairman, ladies and gentlemen, I suppose the reason I am here

this morning is to tell you something of my experience in the last year
or so with a modern small wind electric generating system.
Specifically, I have lived for the past year (with my wife and two
children) in a house which is completely electrified by wind power. We
get all of our power for lights, household appliances, shop tools,
etcetera -- even a television set — from wind power. Our wind power
system, which is completely self-contained, consists of a two-kilowatt
wind driven generator, a set of 19 storage batteries (giving us enough
reserve power for 4 days without wind), a small dc to ac inverter, and a
gasoline generator which we use as an emergency backup system in case of
prolonged calm periods.
Now we installed this system ourselves, but we did not design it or

build it. The entire system is based on commercially available produc- .
tion components which, in fact, can be obtained right now by anyone here.
And, I guess my basic message to you today is that there JLS_ equipment in
production right now which has been tested and proven and which can be
put to immediate use in many small-scale applications. And further, and
this may come as a surprise, there are many applications where wind-gen-
erated power is actually cheaper than conventional generating systems.
Let me start by giving you a little background on our installation

to illustrate this point. When we moved to Maine two years ago and
started building our house back in the woods (we're actually located 5
miles from the nearest paved road), we were faced with the problem of
supplying electricity to our home in the wilds of Maine. The local power
company looked at our situation and came up with a quote of 3000 dollars
to bring us in a line. And they wanted a minimum of fifteen dollars per
month for the next 5 years regardless of how much power we used.
Well, this was enough to make us stop and think a bit. But what
really are the alternatives in a situation like this? The only alternative
usually considered, if power lines are not available, is a small diesel or
gasoline generating set, but the economy of such a system is very poor (not
to mention the noise and pollution problems). A rough estimate of the cost
of power generated by such a domestic generating set, diesel or gasoline,
taking into account capital costs, fuel, and maintenance expenses, is 30
cents per kilowatt-hour or about 10 times the power company rate!
165
Now, here is where the wind-generated power comes in. Wind generated
power may not compete at present with mass-produced power from the power
company, but it does compete very favorably with any other type of individ-
ual power plant. Our complete installation, which I will describe in a
minute, cost us $2800. There is, of course, no fuel expense, and the only
maintenance associated with this system consists of changing the oil in
the gearbox (1 quart) once every 5 years. Assuming, conservatively, a 10-
year life for the batteries and a 20-year life for the other components
and adding in maintenance and interest costs on the investment, the total
costs to us of the electricity generated by our windmill comes out to
about 15 cents per kilowatt-hour - or about one half the cost of the gaso-
line or diesel plant. This is based on an average power output of 1500
kilowatt-hours per year (120kW-hr per month) in a location with 8 to 10 mph
average winds.
So, you see, in our case (and in similar cases throughout the country
where power lines are not easily available), wind generated power can
actually represent the cheapest available means of generating power. It
is for this reason that I predict we will witness the reappearance of
modern wind-electric power systems, at least on a small scale, in the
coming years.
Now for a brief description of our set up in Maine. I won't go into

too much detail because the whole system is described quite completely in
a small booklet which we have printed recently entitled "Electric Power
from the Wind." This publication, which we offer through the Solar Wind
Company for one dollar postpaid, describes the operation of modern small
self-contained wind electric power systems and has several tables and
diagrams which allow you to calculate the power output of various size
systems under various wind-speed conditions.
Our basic system consists of a 2-kilowatt Quirk's wind generator

manufactured in Australia. This "windplant" as they call it, uses a 12-
foot diameter propeller with a full-feathering hub controlled by centrif-
ugal weights. The generator is a 2000-watt, three-phase, 115 volt dc,
which we feed through the voltage regulator panel (included with the Quirks
unit) to the batteries. This panel contains large ampere and volt meters
as well as a transisterized voltage control which works by lowering the
voltage to the alternator-field (thus reducing the charge rate) when the
output voltage exceeds a certain value -- which you can set to correspond
to the voltage of the batteries in their fully charge state. Above the
panel you will see the anemometer readout, which reads the windspeed in
mph at the windmill site. For storage during calm periods we use 19 lead-
acid storage batteries, rated at 130 ampere-hours and 6 volts each. These
give us a total of 15 kilowatt-hours of storage at 115 volts. These bat-
teries are especially built for this type of application; they have built-in
charge indicators in each cell. These batteries come from Australia and
seem to be cheaper than American batteries of the same capacity. We sell a
19-battery set of these in New England for $695 or about $35 apiece.
In our installation we use much of our power directly at 115 volt dc.
All our lights, many appliances including the vacuum cleaner, electric
166
drill, skill saw, sewing machine, etc. will run well on do. Our water
pump also has been converted to run on dc. The only appliances which re-
quire ac are the television and the stereo and for these we use a small
surplus rotary inverter which we purchased for under $100. Of course,
for larger loads, there are several types of electronic solid-state
inverters available and we are now selling units up to 8000 watts. As an
example of costs on these, a 2000-watt unit sells for about $1600.
And now let me move on to some more recent developments. In addi-

tion to our arrangement with the Australian company, the Solar Wind Com-
pany has recently contracted for an agency in the U.S. for the wind driven
generating equipment manufactured by Elektro G.m.b.H of Winterthur, Swit-
zerland. It is our feeling that at present this company manufactures one
of the best units available for the price anywhere in the world. (The
Aerowatt unit made in France may be superior in certain respects, but
they are more expensive than comparable Elektro models.)
Elektro makes several different size units. They make two small
vertical axis mills rated at 50 and 250 watts in wind speeds of 40 mph.
Then they make conventional units in sizes from 750 to 6000 watts output.
The 6-kilowatt model delivers its full output at a wind speed of 25 mph.
Typical monthly outputs from this 6-kilowatt generator are: 350 kilowatt-
hours in a 10 mph average wind, 470 kilowatt-hours in a 12 mph wind, and
perhaps 600 kilowatt-hours in an area where the average wind speed is
14 mph.
We have recently installed one of these large Elektro units beside

our Quirks unit for testing and evaluation. So far it has performed very
well, and preliminary tests show that it will produce about three times
the monthly average output of the 2-kilowatt Quirks unit. The output of
the Elektro unit is controlled by a servomotor at the base of the tower
which operates by tensioning cable running up the center of the tower,
and thus regulating generator output by rotating the tail and causing
tîe unit to turn out of the wind. The windmill can thus be operated at
any power setting from 100 to 0 percent by this control. The control can
be actuated manually and remotely by push buttons or automatically by
various factors. These include over-voltaging of the batteries, too high
current in the generator, and excessive wind speed. A small wind paddle
attached to the tower closes a contact in winds over 60 mph putting the
windmill out of operation for a period of 12 hours, after which it will
again start up automatically if the winds have abated. With the auto-
matic control all the Elektro units are capable of completely automatic
and unattended operation in winds as high as 150 mph.
The Elektro Model WVG-5, 6-kilowatt windplant is the largest unit

currently in production. It has a three-bladed propeller of 16% foot
diameter and, like the Quirks unit, uses a centrifugally operated
feathering system to limit propeller and generator rpm to safe maximum
levels. The price of this unit, delivered on the East Coast with auto-
matic controls, is about $3000.
Now I'd like to make a few comments about what I see to be the
16?
immediate future of small scale wind-driven power systems. Besides the
sort of direct residential electrical power systems which I have just de-
scribed, I think the most promising area for small wind generators may
well be in the area of domestic heating. Some preliminary figures show
that wind-driven generators in the IS- to 25-kilowatt output range,
coupled to a direct heat storage system using heated water (no batteries)
would very adequately heat a typical six or eight room New England home.
The cost of this system might easily be made competitive with present oil
or electric heating systems. Right now such a system could be set up
using existing production components for about $7500 total capital cost
with virtually no expenses thereafter for maintenance or fuel for a per-
iod of at least 20 years. And there is no doubt that the price will come
down if any quantity of such installations is contemplated.
This brings me to my final point here today. I think if we are to

get on with the job of developing satisfactory alternatives to our present
fuels, we will have to approach the problem on many levels. Obviously,
one wind powered home in Maine has little significance on the national
energy crisis. But I feel that one operating wind power system, small
though it may be, can demonstrate to many people that the wind is a
viable and even practical source of energy for the future. For many
people this is a more convincing—demonstration than some of the ambitious
proposals and schemes which seem destined to remain in the conceptual
stages for years to come.
If we want to foster the idea of wind power as a viable alternative

to present methods, we must support efforts to harness the wind at all
levels. As I see it, the small and modest projects that we are involved
in right now could be vitally important to the future acceptance of wind
power on a larger scale, and so it is perhaps in this way that our work
is significant on a national level.
DISCUSSION
Q: You said "we." Who does "we" refer to?

A: Well, at the moment the Solar Wind Company consists of myself, my
wife, and secretary. We have another fellow joining up this summer
who will be working with us, and my brother also collaborates. He is
an architect. It is a very small company, we have only been in exist-
ence about 4 months. Our main work is to import these Australian and
Swiss units.
Q: How much would the cost be affected by a loss of skilled manpower to

erect and maintain these units?
A: My wife and I installed both of these units with the help of a pickup
truck, guy wires, and a jin pole, and so on. So they can be put up
in the field with very little skilled labor. They come with a set of
directions and are really not hard to install at all.
168
Figure 1. - Foreground: 2 kW Quirk wind generator
Background: 6 kW Elektro wind generator
169
ECONOMIC CONSIDERATIONS OF UTILIZING SMALL WIND GENERATORS
Robert Dodge !
Pennwatt Corporation
Houston, Texas
Free power from the wind? The wind may be free, but certainly not
the power, as anyone knows who has tried to capture a few of these "free"
kilowatts.
The usual motivation for considering a small wind generator is
economy. That is, the prevailing notion is that wind power can be easily
exploited and, therefore, must be a most economical form of power genera-
tion.
The purpose of this paper is to provide a practise! guide to the
system designer to enable him to make a decision as to whether a wind
generator is a practical solution. Only small generator plants up to
5 kilowatts are considered.
If the object of considering a wind generator is economy, then the
designer should have an appreciation of the alternate power systems that
are available, the costs, and the pros and cons of each.
In our industrialized society, most inhabited areas have commercial
power, therefore, we can eliminate these areas from consideration. Even
those uninhabited areas where commercial power can be brought in reason-
ably are not good candidates for small-scale wind power generation.
These limitations do not mean that wind generators are not practical.
On the contrary, there are numerous applications in remote, isolated sites
where wind-power generation certainly does provide a practical solution.
A good example is the powering of marine aids to navigation signals. In
most instances, these lights and sound signals are situated at remote,
inaccessible locations. Other obvious applications are remote communi-
cation relay stations, weather data gathering stations, including weather
buoys, cathodic protection, and water pumping.
These are five possible solutions available today to generate power
in inaccessible remote locations: solar cells, primary batteries, thermo-
electric generators, wind generators, and engine generators. Figure 1
shows the relative economics of these alternatives plotted as dollars per
kilowatt-hour versus the average electrical load.
Of these five options, the primary battery is probably the least
understood but the most widely used. The primary air cell has been around
170
for some 40 years, ever since it was used to power our first radios in
rural America. The air cell consists of a container with a zinc anode
and an air breathing porous carbon cathode. The cell electrolyte is
usually sodium or potassium hydroxide. These cells are characterized
by a very low self-discharge and can be employed in series and parallel
to provide up to several years of power. The cost of the cells yields
energy at about $12 per kilowatt-hour. The weight is approximately 11
pounds per kilowatt-hour.
The primary air cell can be used to solve almost any remote power
problem and, therefore, can be used as a basis for evaluation of any
other system. For example, a 1-watt load would consume 8.76 kilowatt-
hours per year. The cost of primary cells would be $12 x 8.76 = $105 per
year. The weight transported to the remote site would be 96 pounds.
These figures, although high per kilowatt-hour, are so reasonable that no
serious consideration of a more complicated system should be entertained.
The only possible exception would be where the cost of transporting the
batteries to the remote site is so high as to change the economics dras-
tically. In some cases where environmental considerations prohibit on-
site disposal of spent batteries, the additional cost of disposal would
also affect the economics.
The following assumptions form the basis for determining the costs
per kilowatt-hour shown in figure 1.
Solar cells: Amortization Cells, 5 years

Batteries, 5 years
Housings, 5 years
Primary battery: Amortization Batteries, 1 year

Housings, 5 years
Thermoelectric: Amortization Generator, 3 years

Batteries, 5 years
Fuel 25C per gallon
Wind generator: Amortization Generator, 3 years

Batteries, 5 years
Housings, 5 years
Diesel generator: Amortization Engines and Generators,

2 years
Housings, 5 years
Fuel 25£ per gallon
An annual interest charge of 8 percent is assessed to all systems.

All the systems considered are continuous power systems. Each is capable
of handling intermittent loads equal to several times the average. All
use battery storage to provide continuous power except the diesel gener-
ator. The solar cell economics were generated assuming a 30° latitude
and 10 days storage in the secondary battery. The assumptions for the
thermoelectric system were propane fuel from tanks, with the tanks
171
transported to the site. The battery storage is ten hours. The wind
generator system assumes a 10 mph annual wind velocity. The battery
storage is 10 days. All secondary batteries are industrial type, low-
discharge, lead calcium. The diesel generator consists of two complete
plants, housing, and automatic controls for unattended operation and
The amortization periods are conservative and no doubt proponents

would argue that their machines would or could operate much longer. How-
ever, no attempt has been made to add the cost of transporting equipment,
fuel, and men to the remote site. Obviously these costs could greatly
influence the cost per kilowatt-hour of any system. Since they vary con-
siderably with the location, each installation requires separate evalua-
tion.
The chart (fig. 2) shows clearly that for loads up to 10 watts, the
primary air cell has the advantage. In cases where transported weight is
a problem, the solar cell or small wind generator should be considered.
From 10 to 100 watts, either the thermoelectric or the wind genera-

tor are good selections. The wind generator has the advantage when
transported weight is considered.
The 100 to 1000 watt range certainly favors the wind generator. In
this range, the load is too light for effective use of a diesel genera-
tor, although at about 500 watts the diesel engine begins to look favor-
able. In this range the weight of propane for the thermoelectric
becomes unreasonable. A 1000-watt average load requires 12600 gallons
or 53000 pounds of fuel per year.
From 1000 watts and up, the diesel generator has a definite advan-
tage. True, the systems are complex and, hence, prudence dictates
redundancy. However, the low cost per generated kilowatt-hour makes
their consideration mandatory.
All the systems except the engine suffer from being modular. That
is, they consist of a parallel arrangement of units or cells so that in-
creasing their size by increasing the number of the same size cells af-
fords little saving per kilowatt-hour. It is the battery that is
required with the wind generator that causes the cost to level out at
about $1.50 per kilowatt-hour.
If the storage battery can be eliminated from the wind generator

system for applications such as cathodic protection, water pumping, or
possibly the electrolysis of water, the cost per kilowatt-hour is much
less. This is particularly true for larger machines. For example, a
10-meter machine could easily generate 12000 kilowatt -hours per year.
The cost of the machine and tower would be about $12000. Therefore,
using the same 3-year amortization and interest, the cost is only U3C
per kilowatt-hour, a figure very comparable with small diesel plants.
In summary, small wind generator plants offer an attractive

172
alternative to primary battery systems and constantly running engines to
generate power in remote areas. They are particularly advantageous where
the costs of transporting fuel or batteries are high. The limitation is
an annual average wind velocity of at least 9 to 10 mph. The presently
available units are most useful in the average load range of 10 to 1000
watts.-
DISCUSSION
COMMENT: I object to your making these look so favorable in areas where

the cost of diesel fuel may run from SO cents to a dollar a gallon. I
don't think there is this large a difference in costs. I think Clews'
calculations did not agree with yours.
A: I used 25 cents. Actually, the cost of fuel is quite small compared

with the cost of the amortization of the machine.
* 30.00
AT UNATTENDED REMOTE SITES

AUTOMATIC POWER INC., HOUSTON, TEXAS JUNE 3,1973
*30-
».20-
(WEIGHT TRANSPORTED TO REMOTE 8ITE/KWHR
I WATT 10 wans 100 WATTS 1000 WATTS 10 KW

AVERAGE LOAD
Figure 1
173
WIND UTILIZATION IN REMOTE REGIONS -
AN ECONOMIC STUDY
James H. VanSant
Hydro-Quebec Institute for Research

Varermes, Quebec, Canada
There are presently many diesel generating stations being used in

Quebec to supply electrical power to small remote communities. The fuel
costs for these stations are very high (5 to 16^/kW-hr) and will probably
become even higher. If a cheaper source of power could be used, sub-
stantial savings would result. Wind energy is a good candidate for a
new source of power in many remote regions because it occurs in sufficient
quantities to drive wind machines. Consequently, an economic comparison
of wind generated power to Diesel power for a small northern village was
made to determine if installation of wind machines would be feasible.
First, we should realize the nature of the remote communities. By

definition, they are isolated and undeveloped. Power must be produced
locally for short-distance transmission. The communities are generally
small with peak demand loads of 30 to 5500 kilowatt-amperes. Power com-
mitments require at least one full capacity standby system. The availa-
bility of "natural energy", such as wind, sun, hydro, etc., will vary
greatly from one region to another; although sufficient wind energy is
available in most regions.
Any power generating system that is selected must be reliable, have

long life, easy startup, long unattended operation and be compatible
with local means of supply and/or storage. A most important requirement
is that the system must be economically competitive.
A wind-driven generator was considered as a supplement to a diesel

group, for the purpose of economizing fuel when wind power is available.
A specific location on Hudson's Bay, Povognituk, was selected. Technical
and economic data available for a wind machine of 10-kilowatt nominal
capacity (developed by the Brace Research Institute of McGill Univ.) and
available wind data for that region were used for the study. Referring
to table I, after subtracting the yearly wind machine costs from savings
in fuel costs, a net savings of $1400 per year is realized. These values
are approximate, but are thought to be highly conservative.
A very important factor in determining the worthiness of a wind

driven system is its duration of utilization. Frequently, there is not
sufficient wind when power is needed. Consequently, an energy storage
system that would provide power on demand is advisable. Also, a wind
174
machine and storage system that could be "tuned" to the wind velocity
for maximum efficiency would be very advantageous. Pneumatic storage
with air-motor driven generators seem to be a good candidate for this
type of system. However, some research and development are needed before
putting a system into service.
DISCUSSION
COMMENT: I was interested to see that for the first time the economic
value of windpower was compared with the actual fuel saving. When we
did our arithmetic on our program, we were only looking for wind power
costs. These were equivalent to the fuel costs.
The second point is you had a figure of 58000-kilowatt yearly out-

put from a 10-kilowatt machine. This is 5800 kilowatt-hours per kilowatt,
which is about the highest value I have seen for any particular site.
A: Yes. As I said, it's a fairly windy place, but we used real wind
data to try to get that estimate. It should be considered fairly
accurate.
175
TABLE I. - EXAMPLE OF WIND MACHINE ECONOMICS (ESTIMATES ONLY)
Location, Povungnituk (pop. 2000).
Diesel Data Capital Costs

Installed 250 kW Wind machine $ 5,000
Maximum demand 90 kW Generator 1,000
Annual output 290 000 kW-hr Transport and foundations 2,000
Fuel cost 6C/kW-hr Regulation 5,000
$13,000
Wind Machine Data Yearly Costs
Installed 10 kW Amortization (20 yr) $ 1,400

Annual output 58 000 kW-hr Lubrication and Main-
tenance 400
Labor 300
$ 2,100
Yearly savings in diesel fuel $ 3,500

Yearly cost of wind machine 2,100
Net savings per year $ 1,400
176
TECHNICAL FEASIBILITY STUDY FOR THE DEVELOPMENT OF A
LARGE CAPACITY WIND POWERED ELECTRICAL GENERATING SYSTEM
Ralph E. Powe
Montana State University
Rozeman, Montana
This report describes a study to be undertaken at Montana State
University with the support of the National Science Foundation. The
major objective of this research effort is to investigate the engineer-
ing feasibility of developing a basic mechanical system necessary for
extracting large amounts of power (on the order of 10 to 20 MW) from the
wind using the concept of vertical airfoils moving along a closed hori-
zontal track system. The research plan shows that this effort can be
divided into four distinct phases, each with its own specific objectives.
The accomplishment of these specific objectives will be major indicators
of progress toward completion of the overall project objective. During
this preliminary study, attention will be focused on those components
necessary for the conversion of wind energy to mechanical energy, al-
though the general characteristics and critical aspects of other com-
ponents will also be considered. The four phases of this program may
be briefly described in the following manner:
(1) the establishment of component specifications and interface
requirements for major system components;
(2) the formulation of alternative sets of conceptual designs
for major system components;
(3) the engineering analysis of various components and systems;
and
(4) the re-examination of basic concept and identification of
any desirable follow-up work.
DISCUSSION
Q: What power level are you talking about? What efficiencies are you
talking about?
A: We really haven't looked at it in detail enough yet to come up with
good numbers at all for this. We're shooting for a system though
on the order of 10 megawatt system or so, and we estimate that we
will be talking about a 5-mile-long track, 5- to 10-mile-long track,
or so.
Q: At what efficiency?
A: Your guess is probably as good as mine on that. That was figured
at about 30 or 40 percent or so, and that's a wild guess, really.
177
Q: I think you have given us a good run-down on potential problems and
liabilities of this system, but I didn't see why you were considering
this over the conventional rotor system.
A: We feel that the advantage in this scheme is that you can get a very
large output from a single unit, whereas to get 10 to 20 megawatt
output from the rotor type system requires a very large number of
units from sizes that are available or conceivable now.
Q: Each one of your rolling stock pieces might be considered a separate

unit?
A: Yes, it could be, thatTs right. But on this, to add capacity you
simply increase the length of the track. When you add capacity, you
donTt increase the structural problems, some supporting base for the
entire system, like you would with the rotor.
Q: I would like to ask you if you are familiar with the Madaras experi-
ments, which were conducted by the Public Service Commission,
Burlington, New Jersey in 1933 on this same type of scheme?
A: No, I'm not.
COMMENT: I think it will save you a lot of time if you become familiar
with the Madaras experiments.
Q: I think that some of the questions and comments that you have just
made save me repeating them. But I still have one nagging problem
with what you have here in the slides. And that is I see a very
substantial program of technological analysis which is devoted to,
let's say, shedding light on a series of connected questions. But
in each one of your approaches I sense what I consider the primary
key question is always placed at the end as something of an appendage,
and quite frankly that is economics. Why go through all this detailed,
complex analysis if you can have established at orice at least an esti-
mated economics which show the scheme to be favorable. .Would you com-
ment on that, please?
A: Well, we felt that we should show it could technically be done first,
and then, like most of the other units, cost is something you can
sort of affect if it proves to be technically feasible and you went
into production on this type of unit. So this was the logic in put-
ting the cost at the end.
\
Q: In other words, you're going to undertake a very expensive technologi-
cal feasibility evaluation program and then if it looks good, then you
are going to look at economics. Why not look at economics first,
because manpower is a pretty scarce resource?
A: First of all, it's not a very involved and expensive study to start
with, and I didn't mean to imply that we are going to completely ignore
economics during this first phase by any means. It's going to be looked
at, but the economists that have been involved in the group that's been
working on this are not going to be involved to a large extent during
this preliminary study, although they will be used as consultants.
This would come in after this one year period.
178
J
&
a
I
I
V
I
179
THE OREGON STATE UNIVERSITY WIND STUDIES
Robert E. Wilson

Corvallis, Oregon
This discussion concerns the work in progress done on engineering

analysis of wind operated power for Oregon State University's Wind Power
Project. The objective of the project is to assess the economic feasi-
bility of commercial use of wind-generated power in selected areas of
Oregon. Wind data collection and analysis is being accomplished by the
Department of Atmospheric Sciences; the engineering work is centered in
the university's Department of Mechanical and Metallurgical Engineering.
A number of machines for generating power have been examined. These

include the Savonius rotor, translators, conventional wind turbines, the
circulation controlled rotor and the vertical-axis winged turbine. Of
these machines, the conventional wind turbine and the vertical-axis
winged turbine show the greatest promise on the basis of the power devel-
oped per unit of rotor blade area.
The estimated cost of Palmer Putnam's 1500 kilowatt .preproduction

unit was updated from 1945 to 1971 using only the effects of inflation.
Without taking into account the effects of 26 years advance in tech-
nology, the 1971 inflated cost was estimated to be $700 per installed
kilowatt. The major cost component in Putnam's design was the rotor,
which accounted for 43 percent of the total cost. As a result attention
has been focused on the structural and fatigue analysis of rotors since
the economics of rotary-winged, wind generated power depends upon low
cost, long lifetime rotors.
Analysis of energy storage systems and tower design has also been
undertaken. An economic means of energy storage has not been found to
date. Tower design studies have produced cost estimates that are in
general agreement with the cost of the updated Putnam 110-foot tower.
180
OREGON STATE UNIVERSITY WIND POWER PROJECT
sponsored by
OREGON P.U.D. DIRECTOR'S ASSOCIATION

*
*
*
PROJECT STUDY AREAS
1. Wind data collection and analysis
2. Wind tunnel investigation of terrain modification
3. Engineering analysis of wind power systems.

*
SCOPE OF ENGINEERING ANALYSIS
1. Energy storage schemes

^
2. Update of Putnam cost estimate
3. Wind generator performance analysis
4. Tower weight and loads
5. Identification of problem areas
181
TYPICAL P.U.D. LOAD CHARACTERISTICS
Load is split between residential and industrial
High incidence of electric heating
200 MW peak load
Peak load occurs in winter months
ECONOMICS
STRUCTURES -* DESIGN
FATIGUE
UTILITY LOAD BONNEVILLE PRICING WIND POWER

CHARACTERISTICS POLICY CHARACTERISTICS
POWER USE STRATEGY
182
WIND POWER MACHINES
Standard
Savonius
Translators
Propeller
Novel
Vertical axis winged turbine
Controlled circulation rotor
MACHINE SURFACE AREA REQUIRED
TO PRODUCE 1 MW IN A 30 MPH WIND
Machine Area, ft2 Tip Speed Ratio
Drag Translator 180,000 1/3
Lift Translator 600* 10
Savonius 60,000 1
Smith-Putnam 1,700 6
Vertical Axis Winged Turbine 1,700 6 •..•:'
*C- -'- 1, L/D = 15
183
COST SUMMARY OF PUTNAM DESIGN FOR 1945 AND FOR 1971
1945 Cost Inflation 1971 Cost
Per Unit Factor Per Unit
ENGINEERING $ 10,000 ? 20,000
MANUFACTURING
1) Standard Equipment
Generators 8,870 3.5 31,045
Main gears 20,344 3.72 75,680
Electric coupling 4,612 2.36 10,884
Governor 2,508 3.50 8,778
Bearings 16,282 3.5 56,987
Switch gear 5,125 2.55 13,069
Coupling,flexible 1,970 3.5 6,895
Elevator 2,665 3.5 9,327
Service hoist 1,680 3.5 5,880
Miscellaneous electrical 2,100 2.3 4,955
Tower (includes erection) 21,395 3.6 77,022
Paint 691 3.5 2,418
Subtotal $ 88,242 $ 302,940
2) Rotor Components
Blades 29,480 3.72 109,666
Hub assembly '42,935 3.72 159,718
Pintle assembly 48,600 3.72 180,792
Patterns, tools, jigs 800 : 3.72 2,976
, Subtotal $121,815 $ 453,152
INSTALLATION
Freight 2,054 3.0 6,162
Land 0 0 0
Roads 7,460 3.43 25,587
Erection 30,000 3.43 102,900
Total Installed Cost $260,071 $ 910,741
CONNECTION
Transformers 3,600 1.92 6,912
High line 15,000 2.55 338,250
Unit cost $278,671 $ 955,903
Contingency 10% 27,867
$306,538 $1,051,493
184
TOWER ANALYSIS
Design based on
1. Pinned joint truss
2. a nn ,, = 17,000 psf
e
allowable '
3. 60 mph wind loads
4. Dead weight at top = 500,000 Ibs
5. Wind loads from "Smeaton table"
6. Design factor of 2
*
*
TOWER WEIGHT
Tower Height Weight
70 ft 13,000 Ibs
120 ft 38,000 Ibs
Sensitivity at h = 120 feet
- 1000-Us/ft
d Weight _
u u:i
a Topload '
*
*
FURTHER STUDY
Prediction of rotor life
Energy storage system technology
Wind power regulating agencies
Aerodynamic analysis
Vibration analysis
185
FRENCH WIND GENERATOR SYSTEMS
John M. Noel
Aerowatt Corporation
Paris, France
HISTORY OF THE LARGE FRENCH 800 KILOVOLT-AMPERE WIND GENERATOR
B.E.S.T. (Bureau d'Etudes Scientifiques et Techniques) was a con-

sultant engineering firm in the field of applied aerodynamics. This com-
pany worked for 18 years on contract for the French Electricity Authority
(E.D.F.) in the field of wind power. B.E.S.T. was dissolved in 1966, and
the newborn Aerowatt Company acquired the know-how, the patents, and the
staff of B.E.S.T. that were involved in wind power. The Chief Engineer
of the 800 kilovolt-ampere wind generator project is now the Technical
Director of Aerowatt.
The Project
The objective was to design a wind driven generator with a rated

power of 800 kilovoIt-amperes, capable of being connected to the main net-
work.. It was an experimental machine and not a prototype.
The rotor was a three-bladed propeller. Each blade was twisted.

The fixed pitch of the blades was, however, adjustable. Other aerodynam-
ical features are
Diameter of propeller, m (ft) 31 (105)

Starting wind speed, m/s (mph) . . . 7 (15.6)
Rated wind speed, m/s (mph) 22 (49)
Survival wind speed, m/s (mph) 56 (130)
Nominal rotation speed, rpm 47
The asynchronous, 800-kilovolt-ampere generator was driven by the

propeller through a gearbox. In the event of a no-load condition in the
network, the machine was connected to a dissipating resistor made of iron
wire located on poles (to prevent overheating) .
The hub was supported by a vertical, pivotable tower having the length
of a blade. The pivot rested on the top of a three-legged pedestal at a
height of 33 meters. The entire system could be tilted about a horizontal
axis going through the feet to two legs. So with the aid of winches, it
was possible to lower the hub for maintenance and repair.
186
Tests
The machine was located near Nogent Le Roi, a small country village
120 kilometers west-southwest of Paris, in a wide flat plain. The site
was reasonably gust-free.
The first propeller on the machine lasted 18 months without any
trouble, and the generator was connected many times to the network. Some-
times the delivered power reached 1.2 megawatts. We do not know the total
energy delivered by the machine during the period.
To improve the characteristic, the rigid propeller was replaced with
a flexible one. Tests in a wind tunnel had shown a flutter effect on.
such blades. Nevertheless, the new propeller was installed. Experience
demonstrated the wind tunnel tests to be correct: one blade was broken;
the hub was then destroyed by the unbalanced torque. It was the end of
the E.D.F. experiment on large-scale wind machines. It ran for 18 months
from 1958 to 1960.
Results
The results of this experience were the following:
The breakdown of the second propeller due to a flutter effect led
B.E.S.T. to make a thorough study of this point and to find ways to avoid
such a phenomenon on new propellers. We have since designed long, slender
propellers without having any trouble.
We believe we now have the knowledge necessary to build a large-scale
wind driven generator.
TODAY'S USES OF WIND POWER

Wind energy is free but it has two drawbacks for a normal use: its
short term, random character and its power varies as the cube of the wind
speed.
In the recent past interest in the wind energy has declined because
of the developments of networks and heat engines. At the same time the
need for small, remote power sources has grown. The development of solid-
state electronics has enabled stations of any type to be unattended as
long as the life of the equipment. For instance, 20 years ago a microwave
relay station needed kilowatts of power; nowadays only a few dozen watts
are needed. Then, a major lighthouse was fitted with a 6-kilowatt lamp;
now, with a 1-kilowatt halogenous lamp.
The Wind Motor
We can split a wind generator into two parts: the wind motor (i.e.,
the propeller, the hub, and the rudder-fin) and the generator.
The problem with the wind motor is that it must meet two opposite
requirements: It must deliver full rated power at low wind speeds to
overcome the need for energy storage; and it must be automatically trans-
parent to the high wind speeds.
187
We have designed autoregulated machines which meet the second require-
ment. Six major lighthouses have been fed by such machines for 15 years.
We have seldom stopped the machines with the centrifugal brake, but these
machines do not meet the first requirement. They are only useful in high
wind areas (Around Brittany the average wind speed is about 10 m/s, i.e.,
25 mph.). So we think the solution to the problem can be found only in
the variable pitch propeller.
The characteristics of all present Aerowatt wind generators are the
following:
Average starting wind speed, knots 3 to 6
Average wind speed necessary for power delivery, knots 6
Nominal wind speed at full power, knots 14
Rotation governor efficiency (wind speed, over 14- knots;
load-no-load speed ratio) . 5
Survival wind speed, knots (m/s) 120 (60)
The Generator
All the Aerowatt generators are permanent magnet ones. Hence, no
excitation power is required. The generator output is ac current, which
is easier to handle than dc. In addition no maintenance is required. For
medium range machines the delivered current is of the industrial type:
three phases ac, 50 c/s.
Survival in Extreme Conditions
In sandy areas the problem is the blocking of ball bearings or slip
rings with sand. This problem is overcome with sealed ball bearings and
tight bodies. Propellers are protected against erosion with a Neoprene
coating.
In icy areas the most dangerous enemy for the wind generator is
sleet. Two ways of beating this enemy are
(1) oversized machines, able to withstand, at nominal rotation
speed, the unbalanced torque given by a 1-inch-thick layer of ice on only
one blade, and
(2) Teflon coated hub, blades, and rotor to reduce the adhesion of
ice.
Several years experience with these machines in the Alps and in Norway
has shown that these methods of reducing the chances of damage caused by
sleet are successful.
Storing the Energy
For small machines or small installations, the random wind power
production is smoothed by storing electricity in a battery bank. Aerowatt
has combined the off-load working ability of its machines with Llle~~a(fde-
livered current in a solid-state control device, which stops the rectifier
188
when the gassing voltage of the battery is reached. So, in temperate
climates, there is no need to add water to the battery within a year.
For medium machines it is possible to install a diesel generator in

parallel with the wind-driven one. Energy is stored in a negative way,
that is, by saving diesel oil when the wind blows. .
Ultimate Power Source
Primary Cells. - The industrial uses of energy require a no-break

power supply, so Aerowatt delivers with the small machines a bank of pri-
mary cells that are able to deliver a steady current of 5 amperes, with a
capacity of 1000 ampere-hours. These primary .cells have a shelf life of
3 years.
Diesel Engine. - The wind-driven generator control boxes are fitted

with special circuits enablingjjt^te-start a diesel engine when the main
battery bank is dischargetTandto stop it when full charge is reached.
Conclusion
Aerowatt has developed a comprehensive system (see table I and figs.

1 to 3) based on wind power able to meet industrial requirements for power
in remote areas for unattended stations.
NOTE ON THE UTILIZATION OF WIND POWER FOR PROVIDING INDUSTRY

ELECTRIC POWER IN THE FUTURE
Wind power has always been used to meet the human needs, either to
provide driving power to windmills, or once transformed into mechanical
power to propel ships. Nowadays one is interested in transforming this
potential power into electric power, the easiest form of power to use.
Studies and tests have been carried out during the last decades to apply
this transformation industrially, that is, to provide electric power to a
distribution network.
Machines with a 1200-kilowatt nominal power rating have been built.

And the Aerowatt Company has the technical data of the B.E.S.T. 800-kilo-
volt-ampere wind generator. However, none of the machines that we know
of ever made use of standardized applications. We believe that these
large machines failed because they were too complicated, and therefore,
too expensive. The price of the supplied kilowatt-hour was too high as
compared with that provided by other sources, for example, fossil fuel
plants or water power. This was due to the basic characteristics of wind
power, namely, it is impossible to store, and it is impossible to fore-
cast when the wind will blow.
Let us compare wind power with water power. We notice that, at

least originally, they are both random. It is difficult to forecast
whether it will rain next week, and in the same way it is impossible to
forecast whether the wind will blow. Water power can be stored in the form
of potential power, on the one hand thanks to mountainous areas and to
189
water mean flow speed, and on the other hand by the creation of dams
(artificial storage). As far as wind is concerned storage is not so easy.
However, the "Comite Technique pour 1TEtude du Vent" has made ob-
servations over a few years, and these observations have shown that the
average annual wind velocity on a given site does not vary much, about
20 percent.
Existing Possibilities
We have stated here that previous efforts to master wind power failed
because of the high price of power supplied to the network. Actually, if
we look closer at the machine we know best, the BEST-ROMANI 800 kilovolt-
ampere machine, we note that it was equipped with a three-blade 30-meter-
diameter propeller and that it supplied 800 kilovolt-amperes for a wind
velocity of 22 meters per second. Moreover, the aerodynamical regulation
did not allow the machine to operate off the network, and a brake had to
be provided so that the machine would not race if the connections with the
network were interrupted. The twisted-wing propeller was difficult to
build. Finally, the machine, a purely experimental one, was too expensive
for the relatively low power supply.
We believe that the problem has been solved since the Aerowatt Com-
pany designed and built a variable pitch propeller which operates far
better than the other existing variable pitch propellers. Thanks to this
propeller, wind-driven generators have been built that can provide their
rated power as soon as the wind speed has reached 7-meters-per-second
(14 kts). For example if we consider the chart of wind velocity charac-
teristics on a site off the French coast, Sept lies, we notice that this
wind velocity (7 m/s) is reached or exceeded 72 percent of the time,
whereas the 22-meter-per-second wind velocity is reached or exceeded only
1.5 percent of the time. As soon as the nominal wind velocity is reached
or exceeded, the machine rotational speed is constant to t 1 percent,
whatever the wind variations are, of course, within the machine-power
limits. The variable pitch allows the machine to keep its normal charac-
teristics even for high wind speeds (up to 60 m/s). In short the Aerowatt
variable pitch propeller allows us to build machines that, over a year,
operate at their nominal power for a greater number of hours than the
machine built heretofore. The result is due to the simplicity of the
design. Under given conditions electric power could be supplied at a
price that is more competitive with steam-plant produced -power.
Machine Structure
The Aerowatt high power wind-driven generator has the same structure
as the low and average power wind-driven generators built up to now with,
however, some operation improvements in consideration of the machine size.
The Propeller. - As for the other Aerowatt wind-driven generators it

is a two-wing propeller regulated by pitch variation and maintained at the
stalling limit. The propeller wings are made of extruded aluminum alloy,
and have a constant section. They are guyed with compensated stretch guys.
190
If necessary, they can easily be coated to protect them from ice deposit
or sand erosion. The wing setting is calculated so that the propeller
reaches its maximum efficiency within the more common wind velocity range,
usually 3.5 to 7 meters per second.
The design propeller nominal rotation speed is limited to a 90 meter

per second relative wind speed (vector sum of the wind speed and rotation
wind speed) at wing tip. Thus the aerodynamical flows always remain with-
in the low subsonic range.
The hub has two kinds of springs: the starter springs and the regu-
lation springs. The starter springs set the rotor blades, when at rest,
to a value high enough, in relation to the wind, to get a starting torque
larger than that for V > 3 meters per second. The regulation springs
compensate for the centrifugal force on the blades by controlling the set-
ting so that the rotation speed remains constant, independent of the wind,
once it has reached its rated value.
The Mount. - The mount holds a spindle, the step-up gearbox, and the
pivot. The shaft of the spindle is connected on one end to the propeller
hub and on the other .to the step-up gear train by means of a coupling.
The step-up gearbox brings the propeller rotation speed to such a value
that the coupled electric generator supplies an industrial-frequency
electric current. The pivot allows the machine to swivel windwards under
the action of the vane moment. The pivot also holds the sliprings, which
transfer the electric power supplied by the generator towards the distri-
bution network.
The lower part of the pivot is fitted on the driven generator sup-
port. The upper part is connected to the mount by a spindle assembly.
The pivot is protected by a flange of the same length. The flange is
connected to the mount and carries a ladder which provides access to the
mount.
Of course the center of gravity of the machine moving parts is ad-

justed so as to lie on the pivot axis.
The Vane. - As for all Aerowatt wind-driven generators, the propeller

rotation plane is located up wind of the pivot. So located, the propeller
operates in a stream which is not disturbed by the pivot wake. The main
purpose of the vane is to keep the propeller rotation plane perpendicular
to the wind. The vane essentially consists of a surface fitted at the
end of a long shaped support linked to the mount rear part. This surface,
the tail fin, tends to stay in the wind stream. If it deviates from this
position or if the wind direction changes, the thrust on the tail fin no
longer goes through the pivot, and a return moment places the propeller
rotation plane perpendicular to the wind direction.
The high power wind-driven generators are equipped with a new device
designed by Aerowatt. This device limits the yaw rotational speed to a
preset value and, in consequence, limits the stresses due to gyroscopic
effects on the propeller. This is very important for the propeller
mechanical strength.
191
Limits for use. - We shall consider the following type of operation:
supply power to a small distribution network, not connected to a national
network, using a heat engine plant with "n" generators of "p" kilovolt-
ampere unit power driven by diesel engines as the power source. If the
network under consideration is in a very windy area, or is so far from the
usual sources of petrol products that the latter once delivered are very
expensive, then the price of power supplied by the high-power wind-driven
generators would then be lower than that of the fuel necessary for diesel
engines to supply the same power. .
Operation conditions. - The wind-driven generator has,a quasi-

constant power. Yet, taking into account the characteristics of the
Aerowatt wind-driven generators, the power that is supplied to the network
when the wind speed ranges between the starting speed and the nominal
speed is negligible. , -
On the other hand, the wind-driven generator has a low operating

cost, in that the primary source of power is free, and the maintenance
expenditures are very little. So, once money has been invested in a wind-
driven generator, it is desirable to use all electric power it supplies.
Under the limits of use already specified, the costs of the diesel-
driven generator includes the price of the fuel used and the price of the
maintenance, as diesel engines require periodic maintenance which is a
function of the operating hours. It is thus desirable to use the diesel
engines as little as possible.
The normal operation of the electric network, however, adds new

restrictions. The power required by this network must be supplied at any
time, independent of the wind conditions. The frequency of the supplied
electric current must remain within narrow limits. And the required
power varies over a short period of time according to human needs. Clearly,
it is impracticable to have alternately periods when power is supplied ex-
clusively by diesel-driven generators and periods when power is supplied
exclusively by wind-driven generators, because there are transition periods
when the wind speed comes to the starting threshold and then to the rated
productivity threshold, either increasing or decreasing, and periods when
the network requirements increase. It is our belief that the best solu-
tion consists in having a diesel-driven generator in continuous operation.
This generator would have two main functions: It would define the fre-
quency of the electric current supplied to the network, and it would
absorb the rapid variations either of the power required by the network
or of the power supplied by the wind-driven generators when the wind speed
ranges between the starting threshold and the rated power threshold. Under
these conditions the wind-driven generators will be coupled to the network
most of the time. We are sure that this will be easier if the electric
generators coupled with the wind-driven generators are asynchronous, as
they are strong electric machines that are perfectly suited to wind-driven
generators.
The idle power necessary for network operation can be supplied either
by the power generator or generators still working and at least partly by
the batteries of the condensers.
192
Adapting the Wind Driven Generator to the Site
Figure 4, curve A, shows the theoretical maximum specific power that

can be taken from the wind according to BETZ theory; that is,
« - if x I pV3
Yet to take off the whole wind power would mean that the wind speed
downwind of the propeller would be zero, which of course, is impossible.
Curves B]_, 82, and BS of figure M- show the variation of the specific power
output actually supplied by the propeller with wind speed for three dif-
ferent values of rated wind speed, 7, 9, and 11 meters per second.
Figure 5 shows the probabilities 6 for having a given specific power
(kW/m ) (Power duration curve) taking the three characteristics of figure
1 as parameters and taking into account the data of the V(t) function in
two places:
(1) Sept lies lighthouse, northern coast of Brittany, France, where
4> = 48° 53T N, G = 3° 10T E, and the mean wind speed is about 8.5
meters per second
(2) Johannesburg, South Africa, where $ = 26° 18' S, G = 27° 10T E,
and where the mean wind speed is about 4.8 meters per second.
Aerowatt systematically chose to adapt its low- and average-power
wind-driven generators according to curve Bj_; that is, the machines start
between 3 and 3.5 meters per second and supply their maximum power around
7 meters per second. This allows a greater use range of wind-driven gen-
erators to areas where it was not possible to use them hitherto; and at
the same time in the other areas it facilitates the storage of electric
power as the number of productive hours is increased.
However, in the areas defined in the section Limits for use where
high-power wind-driven generators are to be used and according to the
regularity wanted, it can be economical to modify the wind-driven genera-
tor decreases with the propeller diameter. If at a given site, the per-
centage of time during which the wind speed exceeds 9 meters per second
for example is close to that during which the wind speed usually exceeds
7 meters per second, it is desirable to choose a nominal speed of 9 meters
per second for the propeller diameter will be reduced by (9/7) • • = li4-6.
Moreover the. reduction.of the mechanical multiplier gearbox, and of course
its price, is linked to the'dimension of the propeller, for a smaller
propeller can rotate more rapidly and the multiplier will thus have a
lower torque to transfer and a smaller velocity ratio to provide.
Figure 5 gives,an example of the diameter evolution of a wind-driven
generator supplying an average:power of 50 kilowatts at nominal speed if
it were installed in Sept lies.
19-3
Power of Wind-Driven Generators to be Built
Table II gives the approximate dimension of wind-driven generator
propellers according to their power, for a 7-meter-per-second nominal
speed.
Building a wind-driven generator with a propeller diameter of 45

meters does not require expensive technologic methods. With a propeller
of the same size and a nominal speed of 11 instead of 7 meters per second,
the wind-driven generator could supply a power of about 400 kilowatts;
and with a nominal speed of 22 meters per second, like the machine built
in 1958, the power supplied would be 3000 kilowatts.
Table III gives a different view of this question. It shows how the
choice of different adaptation speeds affects the main parameters of a
wind-driven generator, this in the site of Sept lies, the power supplied
over a year being maintained constant.
DISCUSSION
Q: What is the cost of these units? Do you have any estimates of either
the smaller ones or the larger ones of those units?
A: Yes. I can give you some ballpark price for f.o.b. machine in the
U.S.A. The costs start from $1800 for the little one, the 24-watt
machine, up to $10 000 for the 4 kilowatt machine including the
machine itself, the rectifier, the controls, and so forth.
Q: Will you describe the 800-kilowatt machine; its size, the number of
blades, and other characteristics? I might add that I have had a
great deal of difficulty finding literature on the efforts in France,
so anything you could tell us would really be helpful.
A: This big machine was a three bladed one. It was running on the prin-
ciple of autoregulation. That is to say it was fixed pitch, but the
speed of rotation limitation was done by the difference of the slope
of the generator itself and the curve of the propeller. The propel-
ler was located downwind of the pivot.
The structure was able to swing around the wind rotor axis so that
the machine could be built on the ground and not on the top to save
some money. .
' The span was 31 meters, roughly 110 feet. The rotation speed was ;
47 rpm, the cut-in wind speed, as I already said, was 14 knots, and
the rated wind speed was 45 knots. That is the reason why we think
this machine was only experimental. Such.windspeeds are only avail-
able about 5 percent of the time.1
It ran very properly for 18 months. The problem was when we installed
the flexible blades, we had trouble with flutter and no more blades.
Q: What year was this?

A: It ran from 1958 to 1961.
194
Q: Why did you switch from rigid to flexible?
A: To try to improve the wind speed and to reduce the rated wind speed.
Q: Do you remember what the cut-in speed was with the flexible blade?
A: It was the same cut-in speed, 14 knots.
TABLE I. - SOME FIGURES ON AEROWATT BIG WIND GENERATORS
Model Span, Rotational Total

No. ft speed, blade
rpm weight,
Ib
4100FP7 31 150 130

15KFP7 59 75 1 000
50KFP7 107 45 6 000
100KFP7 153 30 16 000
TABLE II. - RELATION OF PROPELLER DIAMETER TO OUTPUT POWER
Rated Propeller
power diameter,
kVA m
4.1 9.2
15 18
50 32
100 45
TABLE HI. - EFFECT OF ADAPTATION SPEED ON MAIN PARAMETERS OF A

WIND-DRIVEN GENERATOR
Adaptation Generator Propeller Torque on

speed, rated diameter, propeller
V, m/s power, 0 shaft,
P, kW meters C, rn-kg
7 60.6 32.5 1552

9 68.5 25.7 999.5
11 82 28.8 790
195
.5
Maximum power output

of ideal wind turbine
.4
14 28
Wind speed, mph .3
Figure 1. - 800 KVA B.E. S.T. machine.
.2
14
Wind speed, mph
Figure 2. - AW FP7 machines.
.1
5 10 5 10 15
Wind speed, mph Wind speed, m/sec
Figures. -AWFP5machines. Figure 4. - Power output characteristics of wind
generators having various rated wind speeds.
Johannesburg Sept lies

Rated wind speed, VR, m/s 7i 9 11 7 9 11
Average spec, power, P, kW/m2 2 0.036 0.052 0.060 0.66 0.124 0.19
Rated spec._ power, Pm, kW/m 0.08 0.17 0.31 0.08 0.17 0.31
Ratio Pm/P 2.22 3.27 5.5 1.21 1.37 1.64
Equivalent percent time, t, at 45.5 30.3 20 82.5 73 61
rated power operation
Rotor area diam., if, m 1.0 0.84 0.77 1.0 0.73 0.59
For same energy output

at same location and two
different rated wind speeds.
40
E
E
30
20
5 10 15
Time, 6, percent Rated wind speed, Vm, m/sec
Figure 6. - Propeller diameter as
Figure 5. - Power output duration (cumulative frequency) curves for wind a function of rated wind speed
generators operating at Johannesburg, South Africa (mean wind speed, for a wind generator having 50
4.8m/s), and at Sept lies lighthouse on North Coast of France (mean kilowatt rated output and sited
wind speed, 8.5 m/s) and for rated wind speeds VR of 7, 9, and 11 m/s. at Sept-lies Island.
196
A PROPOSED NATIONAL WIND POWER R&D PROGRAM
William Heronemus
University of Massachusetts
Amherst, Massachusetts
My interest in wind power began back in 1970 when a group at the

University of Massachusetts was studying a proposed increase in the elec-
trical generation system for the Connecticut River Valley. In other
parts of New England others were concerned about the problem of how to
meet the increasing energy demand without causing serious ecological
damage to the New England environment. These studies led to a University
of Massachusetts proposal to investigate a national network of pollution-
free energy resources - the wind being one of them.
As a member on the NSF/NASA Solar Energy Panel, my involvement with

wind power increased. It was clear that a comparison of the energy poten-
tial of the different methods was needed. So, I decided to evaluate the
potential of wind power using one totally wind-driven system. The objec-
tive was to calculate the cost of energy delivered to the customer on
demand in New England and to compare this with the cost of energy from
systems that were being planned at the time. Meanwhile, the cost of
electricity was increasing from 28.2 mils per kilowatt-hour in 1968 to
32 mils per kilowatt-hour.
The system I chose to analyze was the "off-shore wind-power system,"

which consisted of wind-driven electrical dc generators mounted on float-
ing towers in the waters off the coast of New England where the winds are
high. The output from the generators supplied underwater electrolyzer
stations in which water was converted to hydrogen and oxygen. The hydrogen
would be piped to shore where it would be converted to electricity in fuel
cell stations.
With this system, it was estimated that 159x10^ kilowatt-hours per

year could be produced at an average annual revenue somewhat higher than
the 28.2 mils per kilowatt-hour of 1968. Even though the estimated price
was higher, the absence of the pollution that would result from fusion-
combustion processes was a benefit that was worth the added cost. This
fact - the reduced pollution resulting from wind power - and some deeper
thinking led me to the conclusion that solar energy (includes wind energy)
is the only way out of the nation's energy (and pollution) dilemma in the
not too distant future. The sooner the nation starts using solar and wind
energy, the better off everyone will be. There are several reasons for
this:
(1) We have been converting matter into heat at an ever increasing
197
rate and we may be close to the time when the temperature of the earth is
going to start increasing.
(2) The coal, gas, and oil that are left stored in the Earth are
among the most valuable fossil resources that nature provided for man.
It is a crime against mankind to burn any of it.
(3) I don't believe that the U.S. chosen alternative to fossil fuel,
namely, nuclear power, is proper. The nation is headed for very grave
trouble with that alternative, a major part of it being economic. The
costs of energy have risen dramatically in three years and will continue
to rise. These then are the reasons I believe the nation needs a national
wind-power program.
The major conclusions of my wind energy study are
(1) The wind is an enormous source of power, but we must go where
it is, namely, high up, over 500 feet above the ground, to get it.
(2) It is incorrect to view and to cost out wind-powered systems
solely as fuel savers. A self-contained storage system is needed, and
power should be offered for sale on demand. The costs should be made on
that basis.
The national wind-energy program I envision should consist of three
general programs:
(1) A technology program whose objectives would be to improve the
performance of components, reduce their costs, increase their life, and
develop new concepts.
(2) A resource assessment program aimed at determining the energy
potential of the nation's wind, the most suitable sites for wind power
systems, and the effects of the wind systems on the local weather.
(3) A production plan program with the objective of using the
nation's tremendous productive capacity to produce at reasonable costs
the large number of wind power plants that are going to be needed.
In addition, an assessment should be made of the social, political,
and legal problems associated with wind power.
The implementation of the wind energy program is also an important
part of the overall program. One possibility, for example, is to suspend
large numbers of small wind-generator units on cables like suspension
bridges, or on other types of very tall towers.
In conclusion, my studies suggest that wind-powered energy systems
have the greatest chance of being used to supply an important portion of
the U.S. energy need in the near future at costs that are competitive
with other available systems.
198
DISCUSSION
COMMENT: I don't think the professor has left us any questions really to
ask. .1 am absolutely with him, really. If we look at the world as a
whole and at the future of mankind, we must look more than 50 years ahead.
(And mankind usually doesn't look more than 5 ahead.) I see a future in
which our reserves are so low and our pollution levels are so high that
the sort of life that is left is not the sort of thing that I want for
my children, grandchildren, or greatgrandchildren. If we don't want that
to happen, obviously we've got to do something now. Whether windpower is
really the answer I don't know. The professor is a slightly more imagina-
tive man than I, and I see perhaps more barriers than he does. But I am
entirely with him, that wind power is one of the energy methods that we
should develop now and as hard as we can.
Q: What is the relative height above ground level that you are contem-
plating these suspended windmills, and what logistics problems you
might envision in a reasonably congested land area like the North-
eastern corner of the United States? I could easily foresee a
machine like this being used in Canada and I can see it in some of
our uninhabited islands, but I really wonder, with people's addiction
to small airplanes, glider clubs, et cetera, what legal problems you
might see, in your immediate vicinity?
A: Well, let me take that last thing first. If we do have problems with
small aircraft and gliders, then perhaps it's time that we make that
choice. We shouldn't ignore these choices any more, and just because
the FAA has allowed the small plane pilot certain privileges over the
years is no reason why that has to continue. Now, that is the way I
feel.
As for the sizes of mills or machines I studied, the system I put
together for the off-shore wind-power system was based on two sizes
of machines, because they were available to me: a 200-foot-diameter
machine and a 60-foot-diameter machine. I copied the New York Uni-
versity report and was very happy to have their results. As I said,
though, I have since come to the conclusion that in much of the
United States smaller windmills will do. As a matter of fact, I
promised a Senator from Wisconsin a study that is now a couple months
late (but I'll get to him). In that study I've been looking at wind
power supply in Wisconsin. Perhaps I can answer your question in
that context.
On the north-south running highways, which run about 20 miles apart
in Northern Wisconsin from Green Bay north on through upper Michigan,
significant wind energy could be extracted using wind generators sus-
pended over the highways. It is just as straightforward to talk
about north-south wind barrages located over the woodlands between
the highways. In this kind of wind barrage the axis height of the
lowest' machine would be 100 feet. We could suspend large numbers of
wind generators in a suspension system whose towers would rise to
some 600 feet. They would be, perhaps, cage-mast type towers with a
199
bottom so configured that they could straddle the highways. The span
would be half a mile. The upper wires would drop from 600 to 300
feet. Now, on those wires we would suspend groups of vertical axes.
Each axis, capable of turning to face the wind, would in turn carry
the structural framework on which many machines, about 32 feet in
diameter, would be installed - 20 kilowatt units. (Mr. Noel probably
feels that he has made a convert. And I think he has!)
Q: Obviously you studied the Russian work that was done many, many years
ago on that same thing. What is your opinion of that?
A: I think the Russians have done some very excellent work in wind power,
but they aren't doing much of anything now. At least the last two to
whom I spoke simply ignored it or wouldn't discuss it with us. They
said it was entirely too old-fashioned for the Russians.
Q: They had looked at the grid system approach, hadn't they?

A: Yes, they had. In fact we could use structural grids. A very inter-
esting study to make would be to go back to Grandpa's Knob and that
design of a 175-foot-diameter machine (which I thought was magnifi-
cent) , and substitute for it an array of 100 32-foot-diameter wind
generators, each of 20-kilowatt capacity for a total of 2 megawatts
installed capacity. This array would look somewhat like a big bill-
board, or radar mattress. They could be spaced far enough apart so
that each machine could develop its wake fully, so we wouldn't suffer
an efficiency loss there. How would this different 2-megawatt wind
generator compete? I think it would do a lot better than the 175-foot-
diameter machine did.
COMMENT: May I make a brief remark about the Russian effort? Very re-
cently I did some intensive digging, talking to people in Washington and
elsewhere about this alleged big wind power effort in Russia. Most of
the Russians just smiled and thought I was rather primitive. They pointed
out that they had built a few, but that they had so much hydroelectric
power that they could run wires anywhere in the country. They don't need
windmills.
COMMENT: I will just make an observation on the basis of the WKY towers.
There is about a M-0-percent higher wind velocity at 600 feet and about an
additional 60 percent up to about 1000 feet.
COMMENT: Putnam, I think, came out with those figures quite nicely. He
showed that you are really in the realm of diminishing returns by the
time you go much above 150 feet. But I feel that you have to say, "Okay,
so I'm in the realm of diminishing returns, but that's where the wind is.
If I want to stay below it, I'm not going to get much wind." It's like
those systems studies of fishing in the Gulf of Maine that said so con-
clusively that the most economic place to fish is in the street in front
of the fish pier! The only problem is that the fish don't know that.
COMMENT: There is one point we must remember, unless you've thought of a

new way of keeping your machines up there: The higher the tower, the more
steel or some other metal you have to use. It takes some 13 200 kilowatt-
200
hours to produce a ton of steel and 50 000 kilowatt-hours to produce a
ton of aluminum.. We shouldn't waste energy in the structure before it's
even built, there's no energy balance in that.
Q: How sensitive are your power costs in the final step of converting
hydrogen to electricity if it's done by fuel cells. What lifetime
energy have you amortized in the case of fuel cells?
A: It's quite sensitive in the case of the hydrogen fuel-cell link,
though the most expensive portion of that system is the blades of
the wind generators themselves. The second most expensive portion
was the complete system from the electrolyzer through the fuel cell.
Now, the fuel cell life I used was 15 years, and you go ahead and
whistle! I know that some of you people keep saying that Pratt &
Whitney really don't know what they're talking about. But I happen
to have been in their lab off and on for many years, and I'm quite
confident that the 15-year life is going to be achieved. In fact,
this is just one aspect of the wind power system using a hydrogen
link that I feel has a real future - all right, go ahead, shake your
head! I'm sorry!
COMMENT: I think I have to answer that. I worked in the last 15 years

in the field and I can't share their optimism that a 15-year fuel cell is
around the corner. I've seen no test data from anyone that exceeds 2 years.
A: Someone thought enough of it about 4 months ago to assign them a $90

million R & D program.
COMMENT: I understand that some of our people are going to work on it.
That's fine.
COMMENT: I would like to join him. Oklahoma State has had 10 or 12 years
experience. You must know something we do not see in the published lit-
erature. I don't think a long-life fuel cell will come around in the
next 5 to 8 years.
A: I know of one New England utility who is purchasing from Pratt &
Whitney right now a considerable number of kilowatts of hydrogen-air
fuel cells at a total cost of $185 a kilowatt, which includes re-
former. Now, I don't know the life specs, but I'm sure that it's at
least 15 years. Now that's as much as I know; you've got me in over
my eyebrows!
The most significant work in fuel cells perhaps has not been - and this
is really going to hurt - the NASA work of the last few years, but the
U.S. Navy work which created the fuel-cell power system for the deep
ocean search vehicle. Those results are not available to all of us.
COMMENT: Bill, I think you have released a flood of philosophy here,

which I think is not unwelcome. I am frequently called on to address
groups on the subject of the energy crisis and the like, and I have been
using a subtitle to my talk. Usually it's something having to do with
the resource crisis. I think it's appropriate this afternoon to mention
201
it and perhaps coin a new phrase.
What I am referring to is that in any resource pinch such as that in
which we find ourselves now, our criteria have got to change very rapidly.
The escalation of the energy crisis really took off essentially like a
rocket in about 1969. Many of us take a good deal of comfort in looking
at the average cost of energy curves. Though the average costs are up
from 1969, they are still artifically depressed by the fact that some of
the energy contracts in force today were negotiated several years ago.
If you look very carefully at the current contracts, you will see that a
cost of about 90 cents per Btu is fairly common. Those are the numbers
we should look at, not the averages.
Now, when Bill speaks of suspension via hydrogen balloon or cables or
towers, or what have you, I don't think itTs exactly pertinent to point
out that aluminum takes so many kilowatt hours per ton and steel so many.
These are one-time costs, and they can be energetically amortized.
Now, the subtitle to which I refer and which I think is appropriate today
is, "Don't worry about Mars and Venus. The question is 'Is there intel-
ligent life on Earth?'"
I would also like to add that, if we assume that our present point in
history is perhaps 5000 years removed from the first time man recorded
things by infusing them on stones, perhaps this generation has another
five thousand years. The fossil fuel era, through which we are presently
half way through and passing very rapidly, I very inelegantly refer to as
a small pimple on the rear end of history.
I think it would also be appropriate for us to ask ourselves something
to the effect that (I think the word "war criminal" was coined during the
latter stages of World War II) "Are we behaving as "resource criminals,"
because we're using it up just as fast as we please without a thought to
where it ends and what our children and their children will have to do with
when their time comes?"
COMMENT: If you are brainstorming, you can come up with all kinds of
designs for windpower plants. I could surprise you by some of the ideas.
For example, if you put a heavy rotor on a tower and put a rope on it and
drive it, it will go higher up in the air. Then, if you fix the next one
then you can put both rotors up in the air to the stratosphere.
This is just one of the hundreds of ideas that could be mentioned here.
The solution to accumulate many rotors on a system of steel ropes or some-
thing like that has been brought up by Russia and by Holland in the middle
of the twenties. Going in this direction is a market question.
Let me mention one thing, which is the best size for wind electric power.
This is an important question. If you make small ones, you can produce
them in quantity and get the advantage of a lower price by "their small
size.
202
If you make big ones, you get advantages of the fewer parts and there is
somewhere a minimum of the cost, and we have to be very careful to do
this. We can calculate the minimum.
A: Doctor, I think some of your fellow engineers have designed very

distinctive wind generators. I am not sure I could contribute any-
thing.
203
A COMMENT ON TOWERS FOR WINDMILLS
H. P. Budgen
Budgen and Associates

Pointe Claire, Quebec, Canada
The earliest windmills appeared almost simultaneously in France and

England towards the end of the twelfth century. In England windmills
were built in grain growing areas, where there was insufficient water-
power, and were to be seen in a line east of a line connecting Newcastle
with Portsmouth.
No early designs have been found, and it would appear that the parts
were set out full size on the workshop floor and made to templates. The
earliest accurate drawings of windmills are perspective views shown in
Ramelli's "Le Diverse et Artificiose Machine" of 1588. The descriptions
given would enable a millwright to make a complete windmill. The earli-
est published working drawings are those of the fine Dutch mill books,
starting with that of Pieter Limperch, a millwright from Stockholm.
These drawings were published in Amsterdam in 1785. The earliest mills
were very simple, called post mills, and consisted of a box shaped body,
supported on a vertical post. The earliest known illustration of the
post mill is in the "Windmill Psalter", made in England about 1270.
Later the tower mill was developed, and Dutch mill books show this prac-
tice before 1700. Thus, towers were incorporated in the main structure.
In the fabrication, or design, if any, of these earliest windmills, the
main considerations were to accommodate machinery and to hold the wind-
shaft, which was usually inclined 5° to 15° to the horizontal. Generally,
there were four sails, but five were used in Leeds, England, by John
Smeaton in 1774. Also a few with five and .eight sails were tried. Mat-
ters of wind obstruction, or 'drag', on towers were not ever seriously
considered. Later towers were made of masonry and tapered to the top
cap, which was rotated manually at first and later by a small fantail.
Towers built in the Caribbean about 1750 were of masonry from local quar-
ries. These towers were profiled in side elevation, much in the form of
a parabola, and not as cones, so common in later Dutch and English con-
struction.
Of later years the concept of using windpower more efficiently has

lead to compact grouping of machinery, whether for pumping or, more re-
cently, electrical generation, and at the same time to increased speeds
and reduced overall swept area of the vanes. Because of this regard to
minimize size of the rotor blades and to attain high speeds, much consid-
eration has been given to the economics of tower designs. This has been
possible because of the near universal grouping of all mechanicals and
electrics in one housing, so that it operates concentrically with the
rotor shaft. In such a manner only the conveyance of mechanical, or as
may be, electrical energies has to be considered. This can be via a
204
designed structural frame, a structural steel tube, a concrete or a
masonry tower. The requirements for any of these alternatives is dictated
by the weight and size of the complete operating system. For small wind
turbines manufacturers mount the complete power assemblies on a tube,
the height of which is determined according to wind obstructions. Such
tubular towers are usually guyed, but the system is arranged so that the
whole can be lowered, using a winch, about the fixed baseplate. For the
larger power units made, tubular towers are popular, but access has to
be provided into or outside of the overhead power system.
For a similar power output consideration, Brace Research Institute

designed and built a windmill for pumping in Barbados, which, with a
three-vane rotor and a 32-foot swept area, develops a pump water horse-
power of 1.16 and 31.3 at wind speeds of 10 and 30 mph respectively.
This had a structural tower that could be raised or lowered as required.
More recently, Brace Research Institute has designed an improved struc-
tural tower to be raised or lowered, a 3-foot diameter standard drawn
telescopic tube tower, a parallel concrete circular tower, with a 6.5-
foot outside diameter, and one built of reinforced standard concrete
blocks with a 6.5-foot outside diameter. All these towers are suitable
for the designed mechanical pumping, or electrical generation components
designed by the Institute.
In the design of any tower, consideration had to be given to the

effect of normal wind forces on the rotor and the tower and to the drag
effect of the rotor vanes, and the tower, all create an overturning moment.
Circular tubular or masonry towers present a relatively simple aerodynamic
solution. This is not the case with lattice structural shapes. Because
of this easier approach and because of lower manufacturing costs, the
tubular tower now takes precedence everywhere for small and medium sized
windmills.
The Brace Research Institute examined concrete, and standard concrete

block designs for towers, especially for construction on sites by local
labour in developing countries. These designs were found to be cheaper
than shop prefabricated steel structures which have then to be freighted
to the areas of usage.
Fabricated steel towers cost $4750 (made in Canada) to which freight

must be added. A circular reinforced concrete tower made in Canada costs
$6787, but $1984 if made in a developing area with lower labour costs
such as the Caribbean. A reinforced concrete block tower made in Canada
would cost $5000, but, if made in a developing area it would cost about
$1600. A telescopic 3-foot diameter, reducing to 2-foot diameter, tube
which can be extended by two small winches incorporated within the tube,
costs in Canada about $3000.
Thus, on-site construction of towers for medium sized wind turbines,

is cheaper when made of locally available materials—largely because of
labour costs. For similar constructions in Canada the steel tubular tower
is probably the cheapest.
205
SOME EXTEMPORANEOUS COMMENTS ON OUR EXPERIENCES
WITH TOWERS FOR WIND GENERATORS
Ulrich Hutter
University of Stuttgart
Stuttgart, Germany
The supporting tower of a wind generator is subjected to various

forces, the dominant ones being the horizontal component of the rotating
machine (horizontal or vertical axes) plus the horizontal force of the
wind. Also, because tne wind is usually fluctuating in time, the tower
is subjected to vibratory forces and hence fatigue forces. Hence, a
tower must be designed to withstand the fatigue forces imposed on it.
Another important design consideration is that caused by high winds

and gusting. Heavy gusts cause little trouble with fast rotating rotors
that have high tip to wind speed ratios because the rotor blades are
stalled and the forces on the blades do not increase. The more serious
danger is that of overspeeding of the rotor which could destroy the
machine. Hence, the structural integrity of the tower is dependent not
only on the strength of the tower, but also on the regulating systems.
If the regulating systems are not foolproof, the tower will be destroyed
even if it is 10 times stronger than it should be. Finally, the regu-
lating system has a great influence on the cost of the plant because of
the influence on the cost of the tower. Thus, it is essential that all
possible accidents be considered and that fail-safe controls be incor-
porated into the plant.
Another important force is imposed on the tower by the rotor. If

the rotor is located upwind of the tower, the wake of the passing rotor
blades causes a change in the wind forces on the tower. Thus, the tower
is subjected to pulsating forces caused by the passing blades. If the
rotor is situated downwind of the tower, the vibrations on the tower are
drastically reduced, but the blades are subjected to vibratory forces
because they pass through the tower wake. Given a choice of failures,
it would be less catastrophic for a blade to fail than for the tower to
fail.
The last important question -considered concerns the height of the 1

tower. The wind velocity increases with height according to a power law
where the velocity is proportional to the height raised to a power, say
1/7 (the exponent depends on the ground roughness, among other factors).
The optimum tower height depends on the cost of the energy to the customer
because an increase in height results in an increase in the cost of the
plantj which, in turn, offsets the increased energy output. My studies
206
of this question suggest that the costs are minimum for the shortest
tower and that, since the energy extracted varies as the square of the
rotor diameter, the rotor should be as large as possible.
207
WIND MACHINES
P. B. S. Lissaman
AeroVironment, Inc.
Pasadena, California
Starting in 1969 the American Institute of Aeronautics and Astro-

nautics (AIAA) has sponsored an annual technical symposium on the aero-
hydrodynamics of sailboats. As AIAA Distinguished Lecturer (1972-1973)
the speaker was asked to prepare a semi-technical lecture describing
the history and contents of these four symposia under the title The
Ancient Interface, Blackboard to Bluewater. A version of this talk was
given, with the emphasis on wind-driven vehicles.
The basic elements of the air/water momentum exchange were

described: the environment, the potential, the air and water subsystems,
the total system, and the rule. Many of these topics have direct ana-
logues in aerogenerator design. Aspects of optimal sail design and of
waveless hulls were briefly outlined. A wind-driven vehicle, designed
by Andrew Bauer and capable of moving directly downwind faster than the
wind, was described.
The lecture was illustrated with slides and movie clips showing
surfing catamarans (Arnold), land and water versions of the Bauer vehi-
cle, hang gliding (Kilbourne), land sailing (Ripinsky), and wind surfing
(Schweitzer).
208
REPORT OF THE COMMITTEE* ON
WIND CHARACTERISTICS AND SITING
Reliable data for wind power installations are not always readily
obtainable from existing records. Wind stations have often been located
at airports in order to meet the requirements of aviation.
Wind power needs are best served by choosing sites where the winds
are higher than those representative of a broad area. Unfortunately,
there are few wind records for such high wind speed sites. Having in
mind the desirability of several established proof-of-concept units in
the near future, it is recommended that three areas be chosen in which
such units will be located.
On the basis of existing meteorological data, three recommended high
wind areas are the Pacific Coast, the Great Plains, and the Atlantic Coast.
A variety of nonmeteorological as well as meteorological criteria should
be employed in pinpointing exact sites.
Relevant meteorological data are wind speed, wind direction, wind
turbulence, and the variation of these within the lowest hundred meters.
A priority listing of research and development requirements for an area
is given below.
1. Basic wind information, existing data: A search should be made
for all existing wind data for the area. These data should be assembled,
their relevance assessed, and then analyzed if the data appear to be
relevant and reliable. A summary of existing relevant wind information
can then be prepared.
2. Basic wind information, new data: These are hourly averages of
wind speed and direction at two heights, 10 meters and 30 meters, along
with peak gust speeds at both heights with the frequency of occurrence of
gusts in the high range specified.
A minimum of 12 months of data at each site is required, overlapping
the long term record at a nearby station to determine if the winds for the
12-month period are reasonably representative of climatic normals.
Devices for recording directly the standard deviation of wind speed
are commercially available and are recommended for the 30 meter height.
E. Hewson, chairman; W. Barnes; D. Beattie; K. Bergey; R. Cohen;

V. Nelson; R. Rotty; A. Stodhart; T. Wentink; and J. Wharton.
209
Standardization of units and of methods of making and analyzing measure-
ments should be adopted.
3. Basic wind information, turbulence structure: A detailed study

of turbulence structure in the lower levels should be undertaken, using
existing wind data from one of the Great Plains' instrumented TV towers.
Such a structure may be taken as reasonably representative, except over
very rough terrain.
The extensive literature on the dynamic wind loading of structures

should be examined as being highly relevant. Discussions should be held
with the leading authorities in this area for the purpose of determining
the extent to which recent research may be applicable to the design of
equipment for generating power from the wind.
4. Weather modification: The possibility of significant weather

modifications being caused by single or clustered wind turbines should
be examined.
5. Public policy: The content of environmental impact statements

should be set forth for the guidance of those who are to prepare and
those who are to evaluate such statements. Possible legal restraints
should be analyzed in detail. Sites should be selected so as to minimize
both audible and visual pollution.
6. Dissemination of information; A comprehensive, annotated bibli-

ography should be prepared, kept up to date, and widely distributed.
Translations of significant results of research in other languages should
be made and distributed. Some appropriate agency should be encouraged to
collect and reproduce the documents that are fundamental to wind power
studies. Many of these are generally unobtainable at the present time.
Explorations should be commenced with the Solar Energy Society and

its Journal concerning the possibility of changing names to the Solar and
Wind Energy Society and Journal. Sponsoring agencies should support such
publication by authorizing substantial page charges.
7. Size of proof-of-concept units: Since ten 100-kilowatt wind

turbine units appear to have substantial advantages over one 1000-kilowatt
unit at this time, sites chosen for proof-of-concept units should be
suitable for accommodating ten such units even if all are not installed
at one time.
DISCUSSION
Q: Why do you recommend that the heights 10 and 30 meters be established

as standard for measurements of hourly average wind speeds and direc-
tions, along with peak gust speeds?
A: Thirty meters was chosen as being approximately the height of the hub
of a large wind turbine. For a smaller wind turbine the hub would be
below 30 meters. Thus the winds at 30 meters could be taken as giving
roughly those which specify anticipated wind power production and
210
associated with gust loading on the system of rotating blades. A
second set of wind measurements at 10 meters offers two primary ad-
vantages. First, since 10 meters has been adopted internationally
as the height at which surface wind observations should be taken if
at all possible, winds at this height at proposed wind power sites
permit ready comparison with long term winds measured elsewhere.
Second, wind measurements at two such heights permit meaningful ver-
tical extrapolations of wind speed, direction, and peak gusts beyond
30 meters to provide valuable preliminary wind power design data.
Q: Why not take measurements up to 500 or 1000 feet to obtain wind infor-
mation at heights which were of interest to Percy Thomas of the
Federal Power Commission?
A: The group's recommendations are based on the premise that the first
larger wind turbines to be built in the United States, such as the
proof-of-concept units mentioned above, will have a rated capacity
of 100 or perhaps 200 kilowatts. Wind measurements at 10 and 30
meters, along with the upward extrapolations that such measurements
permit, are entirely adequate for the preliminary wind surveys de-
signed to locate possible sites for wind power installations. If
much larger units are contemplated, wind measurements up to 500 or
1000 feet require expensive high towers.
Q: Use a balloon.
A: A balloon will not give the required long term data. Do you mean a
pilot balloon?
Q: A tethered balloon.
Al: Tethered balloons are both expensive and difficult to use, and es-
pecially so for measurements for a full year. When high winds -
those of great interest for wind power - occur it would be necessary
to reel in the balloon to prevent it from being blown away or driven
to the ground. Attempts have been made to measure higher level winds
by the use of tethered balloons but very limited success has been
achieved.
A2: If we are concerned with winds at high levels above ground, measure-
ments are not needed because synoptic data for gradient winds 2000
feet above the ground can be obtained from the pressure pattern
charts.
Q: If I understand you correctly, you mentioned 10 units of 1 megawatt

each. What were the size of these?
A: No. I spoke of 10 units of 100 kilowatts each for a total of 1
megawatt.
Q: Why did you choose those numbers and sizes?

A: This choice represents our group's distillation of the discussion of
the previous two days. This was, in our opinion, the consensus of
the workshop. We also mentioned five units of 200 kilowatts each
for a total of 1 megawatt. The total capacity of a group of proof-
211
of-concept units should be no greater than 1 megawatt in the present
stage of development.
What is the present status of the wind measuring network in the
United States? How adequate is it?
For the flat areas of the Great Plains and the land to the east, the
network is adequate for first rough estimates of wind power potential.
Over both coastal waters and the mountainous regions the existing in-
formation is completely inadequate. For wind power estimates for such
areas, we need information not on representative winds but on ones
that are not representative because they are stronger than character-
istic regional winds. For example, over coastal waters the wind cur-
rents and water currents are a coupled system with feedback from each
component to the other, and the whole must be considered as a unit.
Thus the location of maximum coastal winds may be expected to shift
somewhat with the season in a manner which may become predictable as
the dynamics of this coupled system is better understood through
research. Similarly, research into the kinematics and dynamics of
high-speed air flows in mountainous terrain will assist in locating
favorable wind power sites. For certain selected areas over both
coastal waters and mountainous terrain there is already sufficient
wind information available to permit us to proceed with proof-of-
concept experiments of the type discussed in this workshop.
212
ROTOR CHARACTERISTICS
I would like to preface my report on the Rotor Committee's work by

two pieces of information that have come to my attention since we left
our meeting last-night. I think these points help us understand better
the challenge facing wind power. One of them is an observation, by one
of the more active participants in this symposium, that every time he
sees the pictures of Professor Hutter's hundred kilowatt machine which
is truly an engineering achievement of some significance, he is reminded
of the fact that it provides approximately half the horsepower that is
under the hoods of each of the United States' 50 million automobiles.
The other piece of information that I think is worth remembering comes
under the general heading of "the black goo that comes out of the ground
is tough to beat." Perhaps some of you saw on the TV news this morning
that the AEC has announced that 150 000 gallong of radioactive waste
leaked into the soil in Richland, Washington, recently. But they assure
us there is no need for concern for the population, and in 150 years the
soil will be usable again.
To summarize the conclusions of the Rotor Committee, the quest for

improvement in performance of rotors should not look for improved aero-
dynamic performance of the rotor blades. The current state of the art
is getting all but about 7 percent of the theoretical limit of the energy
that is in a square foot of wind. The magic 0.593 number is there, and
we are not going to beat it by improving blades. The concentration of
effort in aerodynamic performance, per se, should direct itself either to
capturing a greater fraction of the wind stream tube by use of novel con-
cepts, or to reducing the cost of energy available by conventional means.
It is in that context^that the rest of the Committee's information is to
be interpreted.
1. The moi t significant technical problems that limit improvement

of rotors are in the dynamics of the blades. Blade dynamics becomes an
increasingly important factor in escalating costs as one attempts to go
to larger machir es in order to realize the economies of scale,
2. Today's commercial state of the art deliberately sacrifices a

little aerodynamic performance in the interest of producing a blade of
a particular size and load capacity more cheaply. This makes perfectly
good sense, because in losing a couple of points in efficiency you can
change the cost by a very large factor, and therefore get more power per
dollar. Rather than look for improved aerodynamic performance, we need
*R. Oman, chairman; H. Chang; U. Hutter; T. Lawand; P. Lissaman;

H. Meyer; W. Nixon; J. Noel; R. Ormiston; R. Puthoff; W. Vance; W. Wiesner;
C. Wilcox; and R. Wilson.
213
better aerodynamic performance for a given investment.
3. One of the problems that is significant in terms of the relia-

bility and long life of rotors is the general problem of fatigue - and I
deliberately avoided calJLing this metal fatigue because there is good
indication that fiber-reinforced composites will be used extensively in
wind turbine blades. Fasteners are a particularly bad cause of local
fatigue problems in the vibratory environment of a turbine. The aero-
space industry has made recent progress in improving fatigue life of
fastened joints by the use of interference fasteners, but their costs
are still quite high.
!4. The definition of the failure modes of windmills is an area

that needs specific study. What are the ways in which windmills can go
bad? What are the pathological conditions of the loading of a windmill
rotor?
5. Although materials and manufacturing improvements can be taken

from other technologies, there is still a great need for improvement in
both of these areas, especially with the direct result of reducing cost.
Costs of advanced composites are dropping rapidly with increased usage,
and some of the unique stiffness properties of advanced fibers make them
very attractive candidates for blending into a fiberglass layup to in-
crease stiffness. High stiffness and low weight are very important in
raising the vibration-limited speed of large blades.
6. One point was made after we adjourned, but was discussed in an

impromptu appendix to our session because it is considered to be rather
important. The use of strictly disciplined dynamic modeling - careful
observance of similitude laws in small-scale experiments - is a very
powerful technique that should be given considerable emphasis before
commitment to any large-scale expensive installation. The emphasis
here cannot be made too strongly on the need for strict similitude
discipline in that work.
7. A lot of people have worried about the wakes from towers and
the passing of the tower disturbance in the wind field as the blade goes
by. The consensus of those who have experience in this area is that the
turbulence, gust loads, tower vibrations, and so forth, present far more
significant dynamic disturbances to the rotor than do the wakes of the
tower.
8. There is a very strong need - and perhaps this should not be so

far down on the list - for better, and totally reliable, control tech-
niques to match rotor speed and pitch characteristics to changing wind
factors. This is particularly important in terms of ensuring that any
failure of the wind turbine will occur in a safe manner.
9. Under the requested general heading of environmental problems,

it seems to be unequivocal at this point that wind turbines are not
noisy. Professor Hutter cited an example of a large installation on top
of a hotel in the Schwarzwald where the wind turbine could not be heard
214
anywhere around the hotel, but the diesel generator that it replaced was
audible to everybody in the hotel at all times that it was running.
There is the general aesthetic problem of the visible impact - vis-
ible pollution, if you wish - of large numbers of large wind towers, but
that must be left to somebody other than rotor designers to solve. The
increase in surface friction factor of the wind and its potential effect
on climatology has been a concern that many people have voiced. The con-
sensus of this Committee is that it is more likely to be a favorable than
an unfavorable effect in most of the places in which wind turbines would
be installed, most notably the Midwest plains, where what the farmers
want most is something to "soak up" the wind. If you have ever sailed
around Nantucket Shoals, you would agree with that conclusion.
There were concerns for such things as vandalism and public safety,
the need to fence and to protect wind turbine sites, air traffic inter-
ference and direct impact by airplanes, the wakes of wind turbines as a
disturbance factor to airplanes, and the legal questions of wind rights.
No one felt these problems were basic limitations, and we felt they were
outside of our charter to deal with rotors.
To summarize the development tasks to be done and their relative
priorities, Professor Hutter nominated the following list, and it was
enthusiastically received by the balance of the Committee.
The first priority is selecting and trading off configuration candi-
dates; that is, deciding the best way to get the most power per dollar
out of the wind. I would personally add that we probably should start
with the European systems that have been developed most recently and use
them as baselines against which to compare candidate designs and concepts.
The second priority is the question of dynamics, particularly dy-
namic problems associated with removing those limitations to the increase
in rotor diameter that are the main factor responsible for escalating the
costs of large designs.
In a normal engineering development program, the place we start is
in the definition of the requirements; the determination of what the
thing we are going to make is supposed to do. In this list of priorities,
we place the definition of the requirements for wind turbines third
because wind turbine requirements are so heavily dependent on particular
site and demand characteristics. However, requirements definition must
receive some emphasis, in particular, the identification of failure mecha-
nisms and acceptable failure modes.
Fourth on the list would be a better understanding of the control
problem, both from the standpoint of the computational end of it - the
autopilot, to use an aerospace analogy - and of how one actually brings
the intelligence of the control logic into the control surfaces - the
hydraulics, the muscles, mechanisms, whatever. Both of these areas need
further emphasis.
215
Materials, fatigue resistance, and aerodynamic improvements of
blades were not called out for special mention, because wind turbine
technology can be copied directly from applicable portions of current
practice in aircraft technology. The rest of the questions of wind tur-
bine design we feel were best left to the provinces of other committees.
DISCUSSION
Q: What should the size of the first prototypes be and how much power
should they deliver? Should you go for large or intermediate size
windmills?
A: Well, the comment on the need for dynamic modeling is relevant to
that question. If we are going to be choosing configurations, it
would be my personal opinion that we would want a small one if we
are talking about something which is not pretty close to the types
that have already been pioneered. Scaling laws that have been care-
fully formulated can be used to generate larger prototypes from the
performance of the smaller machines. If, on the other hand, we
decide on a machine that looks very much like Professor Butter's or
the one that Mr. Noel is selling, where considerable amount of
development work is done, then I think we can go pretty big right
away with good design.
Q: What do you mean by pretty big?
A: I think everybody has his own idea what big is, but a 100-kilowatt
machine is state of the art. Beyond that there have been problems
with blade failure. The blade failure is largely attributed to
vibration problems and metal fatigue. So, if we feel we have a good
handle on the vibration and fatigue problems, we can go bigger than
that. That's one answer.
Q: Any suggestions or recommendations as to what type of rotor has been
recommended as the most economically feasible? Maybe Professor
Hutter may have a comment on that.
A: If you want me to comment on this, I should say the plant in 1942
had a diameter of about 53 meters. That was the state of the art in
1942. Presently, there are some installations between 30 and 34
meters diameter. Eventual problems occurred in lesser rotor diameters
so we came from this point to increase the diameters as necessary.
The next aim should be a plant of 130-meter-diameter swept area.
This should be a step not to get into too serious risks. But the
aim could be to make even bigger ones and find the solutions to do
this.
As we mentioned yesterday, there are additional problems - especially
dynamics - of erecting such a plant. A very special problem that
could occur would be that the formation of the blades, due to the
gravity field, could cause a permanent unbalance of the rotor system.
We should aim towards 10 000 square meters. This should be feasible
in the next few years.
216
We should be able to develop this and put such plants into operation
and put them in many climate conditions from Alaska to Florida.
I have just a short remark. The question has been how much power
should be installed in such a plant. This is a question which has
been a topic of some of the organizational studies.
I plan to install not too much kilowatts. If you install less kilo-
watts per square meter, less than 300 watts per square meter, it
should be an average of 200 watts per square meter. Plants of this
installation size have been operated with many years successful
running.
If you install more - if you have a 5000 kilowatt plant, it looks

good, but it doesnTt give any more kilowatt hours.
I might addj as an aside, that when I attended this session last

evening, one of the men commented that a wind turbine system is
really a fatigue machine. I thought that was a nice description.
217
ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS
This summary has been prepared from the notes of Dr. George Szego
of InterTechnology Corporation who served as Chairman for this session.
The group agreed to limit our deliberations to that area of the wind en-
ergy system between the rotating shaft and the end load. We also decided
to consider all approaches in terms of a 30-year service life requirement.
Let me first discuss energy storage systems, starting with the elec-
trolysis of water which stores energy in the form of hydrogen. It was gen-
erally concluded that for a 30-year system, costs on the order of $200 per
installed kilowatt capacity were approximately the present state-of-the-
art. This is reported to be a reasonably mature technology, although
problems such as hydrogen embrittlement are going to require future study.
The availability of suitable water is also something of a problem, with
purification being required before electrolysis. The end product is hydro-
gen, and there was some discussion of the feasibility of storing hydrogen
in the gaseous state. We feel this is primarily an economic rather than
a technology problem.
If the hydrogen is transported by pipeline, there may be problems

related to leakage rates and, therefore, safety, and there is also a ques-
tion of the economic feasibility of doing this because of the higher
pumping power losses involved in compressing hydrogen gas compared with
natural gas. Liquefaction is practical and is used today. However, it
is an extremely low temperature process which would require special pipe-
lines. A question was raised as to the loss rates from large storage
tanks, and whether these are economically acceptable? A general feeling
toward hydrogen is that it must be regarded as a potential hazard, and
the safety aspects will have to be explored carefully. There is also a
psychology problem — that is, a "Hindenburg Syndrome" — involved in get-
ting the public to accept large scale use of hydrogen.
For secondary batteries, the performance characteristics can be

reasonably well identified. Presently, energy densities tend to fall in
the range of 10 to 100 watt-hours per pound and power densities at 30 to
100 watts per pound for lifetimes of a maximum of 5 years. The cost asso-
ciated with 5-year lead-acid batteries is about $80 per kilowatt-hour.
Lead-acid batteries like water electrolysis is a reasonably mature tech-
nology, and it is felt that there is only a modest opportunity for
*G. Szego, chairman; H. Allison; N. Beard; W. Carl; R. Dodge;

W. Hughes; E. Lutzy; D. Rabenhorst; R. Ramakumar; D. Reitan; G. Rinard;
L. Robertson; T. Rowe; H. Schwartz; G. Sheperd; R. Thomas; and J. Tompkin.
218
performance and cost improvements. The most promising research and de-
velopment opportunities are in other advanced battery systems. Regarding
policies, there appear to be possible critical materials shortages, par-
ticularly if lead and zinc are going to be required on a very large scale.
There is also a question of whether bulk energy systems based on batteries
will have heat dissipation problems and therefore result in thermal pol-
lution. Batteries, in general, were considered to be nonhazardous, with
the possible exception of large, high-temperature, alkali-metal batteries.
Compressed air storage was discussed at some length. The general

cost figure arrived at was $80 per kilowatt with a land area requirement
of about 6 acres per megawatt.
Efficiencies of cold air storage systems were reported as 67 per-

cent — that is, three kilowatts put into this system would yield two
kilowatts delivered later on.
(Note: Regarding the question of efficiency, Dr. Szego asked that it be

entered in the record that by adding 4500 Btu's per kilowatt-hour to the
stored gas before it enters the turbine, the overall system efficiency is
raised to 130 percent. Based on the information available, the committee
was unable to assess the validity of this claim. Research and development
is required, particularly in the area of turbine technology since the 600
psi systems will require a 40-to-l pressure ratio turbine, which is not
state-of-the-art. This approach does not appear to be applicable for
very small installations. It appears that the approach would be environ-
mentally satisfactory.)
The flywheel is reported to be capable of storing 30-watt-hours per

pound, and delivering extremely high powers for short periods of time;
the figure of 1000 kilowatts per pound for 2/10ths of a second being one
example given. Costs are expected to be in the order of $50 to $75 per
kilowatt-hour for a 30-year lifetime. Research and development are needed
in the areas of economic analysis, construction of prototypes, and on
technical problems related to bearings and dynamic resonances. In the
policy area, the major questions seemed to be, first, safety and, second,
the public's lack of familiarity with this concept.
Pumped water storage systems are fairly straightforward. Effi-

ciencies are about 67 percent. Costs are reasonably well-defined. A
figure of $180 per kilowatt-hour electric was quoted. They suffer mainly
from the limited number of acceptable sites available which are deter-
mined by the climate, geology, and geomorphology of the area. Frequently,
the acceptable areas are far from load centers, and they are environmental-
ly undesirable because they occupy large areas of land. No R & D seems to
be required in this area since this is an established procedure.
In the energy conversion area, it was noted that there are some
approaches to wind energy utilization that require no energy storage.
On-line generation of wind power, for example, was felt by a number of
members of the panel to be feasible without storage by simply feeding
the power into the grid as it's produced. This would require frequency-
219
controlled alternators, as one approach. Research and development is
needed for such alternators to accomplish reductions in weight, size,
and costs. Direct, nonsynchronous machines were also advocated. Here ac
is converted to dc and back to ac again, using batteries as the inter-
mediate dc step. This has the advantage of decoupling the variable
frequency source, the windmill, from the fixed-frequency load. This pro-
cedure has been used for large scale plants. Some development would be
needed for small scale applications.
Because of the limitations of time and the broad scope of this topic,
the Committee was concerned that these recommendations could gain unwar-
ranted authenticity and credibility by virtue of their appearance in the
proceedings of the conference. The Committee wishes to note that these
are the opinions of a heterogeneous group, and that they should be re-
viewed by competent authorities to assure that the recommendations are,
in fact, credible.
DISCUSSION
Q: I'm an electrical engineer and couldn't be expected to know anything

about thermodynamics and I really don't. But I am having trouble
comparing compressed air and pumped water storage, for example. Ob-
viously, if our experience tells us we can get two for three with a
noncompressible fluid, how do we get two for three with a compres-
sible one? What happens to dv/t? I'm not criticizing, I'm just
asking somebody to explain it.
A: I don't feel I know enough about Dr. Szego's concept to comment on
it. I have not read an analysis of the pumped storage versus gaseous
storage approach. Is there anyone else who does feel comfortable in
discussing this point?
COMMENT: It seems to me you have an upper limit in a noncompressible

one, but I could be wrong.
Q: Did you mention our discussion of energy storage with cryogenic con-
ductors and our decision that it is of little promise?
A: Thank you for reminding me. I accidentally skipped it, as a matter
of fact. Cryogenic conductors were discussed. They are too expen-
sive, too large, and too hazardous. And they did not represent any
appreciable R & D opportunity. The adjectives used to describe the
approach ranged from "impractical" to "absurd."
COMMENT: There are two things I'd like to comment on. I didn't quite
catch the point about the- synchronous principle. There doesn't seem to
be any problem about which type of machine one uses on a network. This
seems to be a common principle.
The point I would really like to make has nothing to do with that.
It's a pity in some ways that this energy storage problem has been dis-
cussed with and associated with wind power at this meeting. Because if
anybody ties these figures for energy storage costs to wind power costs,
then you have completely jammed the thing before you start. I think
220
energy storage is a problem and it probably associates with all sorts of
things. But not particularly with wind power.
A: I agree.
Q: There is another aspect of energy storage that we may not have dis-
cussed in the session last night. Thinking in terms of a large power
generating unit that would supply utilities, how we would cope with
the peak demands or base loads when the winds are below the optimum
range? Perhaps to alleviate the economic factor of storage, I think
the storage facility should be located near the wind-generating plant.
I learned that little lesson 4 or 5 years ago where we had to set up
our wind facilities to generate power and even produce hydrogen gas
in, letTs say, West Palm Springs to supply some of the needs, let's
say, somewhere in Oklahoma, because they may have tornadoes. We can-
not have units there. I'm speaking of a very large system, a grid
system. So this storage problem and the economic costs could be min-
imized if we could study meteorological data to determine where we
can install wind power plants and have storage facilities nearby.
Hopefully, there is a transmission system in that proximity which
the meteorological data could justify as an appropriate location.
COMMENT: I would like to add one comment to what Mr. Stoddard said. I
think perhaps energy storage was discussed in the wrong context at this
meeting. I think the question of energy storage has to be raised, and
should be raised, but until you can define fairly clearly what the energy
storage requirements are and whether they exist at all, it's very diffi-
cult to discuss methods for providing storage intelligently.
221
APPLICATIONS
A three pronged attack should be made to convince the people of the

United States, applicable windpower manufacturing industry, and eventu-
ally the electric utility industry that wind power should be applied
within the United States:
1. A rather short-term program that would provide financial sup-

port toward the demonstration of existing or improved windpower hardware
for either heating or the provision of electricity at a number of demon-
stration residences within the next 2 years.
2. A relatively short-term program of about 3 years duration which

would result in operating wind power plants of 25 to 100 kilowatts capac-
ity by 1976.
3. A more deliberate 5- to 6-year technology improvement program

which would culminate in the selection, construction, and operation of a
number of wind power systems, totally self-contained with storage subsys-
tems, of the 500- to 25 000-kilowatt size.
The suggested locations for the operating systems are
1. At institutions of higher learning which show a genuine interest

in using the product and in using the installations as instructional, re-
search, and public service facilities.
2. At one or more National Laboratories who want windpower systems

and are willing to operate them as research and public demonstration
facilities and who want to use their product.
3. At one or more remote U.S. Air Force Bases and at a number of

other U.S. military bases where the wind resource is good and where fuel
logistics are burdensome. Cooperation with the DOD should be investigated
here.
4. At a very large number of new construction private homes.
5. At one or more New Town or institutional sites in conjunction

with the MIUS program administered by HUD.
W. Heronemus, chairman; E. Barnhart; H; Clews; W. Hausz; W. Hughes;

M. Jacobs; B. Jessop; R. Madey; J. Mockovciak; R. Powe; F. Rom; M. Sherman;
and J. VanSant.
222
The geographical location for any one of the above should be such
that the windpower resource is at least modest, and preferably large.
Operating plant in regions of good winds near large population centers
would be of greater value than remote plant. The New York City - Long
Island, the Boston, and the Gape Cod regions are suggested as starters.
Buffalo and Rochester, N.Y., and Cleveland, Ohio, are other suggestions.
We wondered if somehow a significant wind-power demonstration plant

could be added to any 1976 Bicentennial parks or cities that might yet
be considered.
A coordinated effort should be made to obtain the cooperation of

the rural electric cooperatives and other state or regional associations
of publicly owned utilities in the advancement of wind-power systems.
We agreed that in the next year and perhaps during the next 2 years,
the major introduction of windpower systems to the U.S. public as reli-
able, operating alternatives would be done by individuals or very small
companies like Henry Clew's "Solarwind" and Hans Meyer's efforts. Those
efforts warrant support.
The impact of large numbers of single-dwelling units, perhaps a 25-

kilowatt generator on a UO-foot pole with a 0.20 plant factor and a 300-
kilowatt hour lead-acid storage battery, in any one market, operating
successfully and economically for their owners, might be the best impetus
toward adoption of wind power by a utility.
The federal government should be asked to consider direct grants to

utilities to encourage their early adoption of significant windpower
plant, following the pattern used by the government to encourage construc-
tion of nuclear plants.
It was asked that NSF/ttASA sponsor the publication of a monthly

newsletter to be mailed to anyone interested in windpower, this to be
above and beyond any quarterly R&D project reporting system. Many
younger people are seriously interested in this energy alternative, and
they would like to be kept informed. "Communitas" of Washington, B.C.,
has offered to start such a newsletter based on the attendee list of this
workshop. It was suggested that the newsletter be called "The Zephyr".
It was suggested that other deliberate continuing action be taken

to encourage the interest of younger people in this program. Perhaps
here we have something started which is "technology patterned for humane
living". The country must convince many young people that technology
can not only create problems, it can also be used to get us back on the
track of a more ecologically sound way of life. How many excellent
brains would again be turned toward hard science and engineering if the
chance of making a contribution toward windpower systems and other
pollution-free energy systems were the goal of that kind of education?
Perhaps here we have a good opportunity to put at least some of the old
and the new back into genuine harmony with each other.
223
As many people as possible should be shown that windpower can con-
tribute. As many enterprises as possible must be excited into producing
windpower hardware. Then, through large-scale defection from the in-
dividual residence utility customer and through an indoctrinated or
propagandized portion of utility management the program must bore in. And
if that by itself doesnft bring us ever-increasing amounts of wind-gen-
erated electricity, then direct grants should be attempted. In those areas
where the federal government is already in the role of electricity pro-
ducer and/or marketing agent, the job of conviction and conversion may be
easier. In the East, particularly, where those roles are predominantly
private-enterprise roles, the job will be tougher.
The debate about visual pollution associated with large wind plants
should be started at once. It should be a moderated debate, kept both
lively and honest An educated public, if given full particulars, will
and should settle this. We must learn to face issues like this democrati-
cally.
DISCUSSION
Q: How about the identification of agencies which would come under

regulation and control of wind generated power?
A: We did not talk about that last night. This is certainly something
that has to be looked at as part of any significant deliberate pro-
gram. The political institutional problems associated with ever
getting something off the ground are just as important and perhaps
more important than the technology. Yes?
COMMENT: I went to the Utility Commission in Oregon to find out their

views or opinions on wind generating plants. First of all, they have no
objection, provided it does not have any adverse effect on the community
in which these wind generating plants would be located.
I asked if they would object to my getting into the business of

producing electric power. "That's your prerogative", they said. That's
the Utility Commissioner's answer in the state of Oregon, This will
probably be the comment in many of the other 49 states.
COMMENT: I have an idea we would find 50 different approaches to this,

and it's very important that we find out what they are.
224
PANEL DISCUSSION
Moderator:
Dr. Frederick H. Morse

Members:
Mr. Everett Lutzy

Manager, Town of Hull Municipal Light Plant
Dr. Robert L. Loftness

Manager, Washington Office
Electric Power Research Institute
Mr. Lawrence M. Robertson

The Western Electric Industry
Mr. James Wharton

Tillamook P. U. D.
Mr. Charles W. Lines

Assistant to the Chief
Division of Power Surveys and Analyses
Bureau of Power
Mr. Brian R. Jessop

Rural Electrification Administration
U. S. Department of Agriculture
Mr. Lee L. Douglas

Boeing Vertol Company
Mr. John Mockovciak, Jr.

Manager, Energy Systems
Grumman Aerospace Corporation
225
Moderator, Dr. Morse:
One of the elements of any program which is introducing a new energy

source, or perhaps not a new, but in the picture today is relatively new,
it is quite important to get the point of view of other government agen-
cies, of utilities, of the user, and of the industry or industries that
might be involved in manufacturing the system.
So we have asked several people from other government agencies, from
some utilities, and from industry if they would briefly say what the
present or future role of their agency, company, or utility might be in a
wind energy program and what the key problem or problems from their point
of view might be. Then, as a group we will ask them what they feel the
Federal government role should be in a national wind energy program.
Later we have scheduled a discussion of the NSF wind energy program
and the NASA involvement in that program. Although this might be a minor
handicap for our present discussion, let us remember that the objective
is to develop wind energy systems which are reliable and economic and to
prove the concept of large-scale wind energy extraction.
Mr. Lutzy:
In our nation, it is important how a company stays competitive in

serving the public. Basically, our plan is to provide electric rates
that are at least equal to, if not lower than, the rates of our competi-
tion. Secondly, the reliability of our service must be as great as, if
not greater than, our competitors. And the service we offer must be
equal to or more advanced than our competitors.
Now, since we are a municipal light plant, we are part of a town

government. Therefore, we feel that the life of the utility depends on
the life and the success of the economic conditions of the community we
serve. Since the public owns us, we answer to it.
We meet our annual costs from revenue. We have a substantial invest-

ment in any generation and transmission of a distribution facility. Since
we generally have to get bonds issued, this requires that we go before
the public and get the vote of the people to approve any project we have.
We must, therefore, show that our suggested project is good for the com-
munity, particularly since in Hull we have one of the highest tax rates
in the nation. So naturally we don't propose anything that is not eco-
nomically sound.
We are always watching technological developments for ways to im-

prove the efficiency of operation of our electric utility. An example is
automatic load control where you can see in the system what's exception-
ally out of line on an on-line, real time basis. This automatic load
control can improve the efficiency of operation, and it is variable,
based on the way the customer uses it. We are considering telling the cus-
tomer that if we use load control it will benefit him as well as us. Since
a load control of his different appliances reduces our power costs, that
part of the saving goes to him.
226
The highest peak demand in our electric system is at a time when the
wind blows the hardest and that's generally when the temperature is the
lowest. Therefore, the chill factor is such that tremendous heat loss
occurs at the time the wind is blowing the strongest. So, the question
is - is there any way to economically convert this wind power to a mean-
ingful use? For us in Hull, its major use would be for heating in the
winter.
If the total use was 20 000 kilowatt-hours a year, practically 60 to

70 percent or more would be used for winter heating, and this would result
in saving fossil fuel energy. Also, when the wind blows, you use that
energy and you don't do any polluting.
Dr. Loftness:
The Electric Power Research Institute (EPRI) has inherited the on-
going R and D activities of both the Electric Research Council and the
Edison Electric Institute. Dr. Starr and members of the technical commit-
tees have reviewed those research and development programs still under
negotiation at the time of the transfer and most of them have been approved.
We are now in the process of looking at new activities, including new

energy sources. The role of EPRI will be to provide the utilities with
options for the future and to fund research and development that is hot
being funded elsewhere to the extent our revenues from the utilities
permit.
As far as I know, in reading the documents of EPRI and its predecessor

organizations, I have not seen much reference to wind energy. This does
not mean that the Institute will not be interested in wind energy. In fact,
a symposium or workshop such as this is extremely valuable in pointing out
the technical as well as economic status of wind energy. As with other
alternate energy sources, the problem wind energy will have in the future
will be the question of competing energy systems.
As you all are aware, the problem the country faces is a growing
shortage of petroleum. In terms of national policy, it is possible that
the Federal government, if it so desires, could require that wind energy
machines be installed, even if the cost is higher, in order to reduce the
importation of oil. The government could also direct that we gasify or
liquify coal to meet our oil and gas needs or that we adopt other options,
for example, the construction of large solar stations.
Decisions on these options by the government will require a factual

base of information. On its part, EPRI would also like to have as much
data as possible on the technical merits, the history, and the economics
of alternate energy sources as well as information on new programs and
proposed ideas. I am not saying that EPRI will fund every new idea, but
we will certainly include them in the broad assessment we will have to
make in judging which technical activities the Institute will fund.
227
Mr. Robertson:
I am here in a dual role: One as retired Vice President of Engineer-

ing of the Public Service Company of Colorado which is a utility which
serves much of the State of Colorado and have a load of about two million
kilowatts; the other role as the representative of the Western Energy Sup-
ply and Transmission Associates (WEST Associates) which is an association
of about 23 organizations and utilities (public, private, and municipal)
in the western one-third of the U.S.A. and Canada with a load of about
25 000 kilowatts. These utilities are all interconnected into a large
network covering this area. Since this is a large area, there is quite
a diversity in atmospheric conditions and wind velocities in different
locations so that there would probably never be total calm nor severe
storm conditions over the entire area at any one time. Therefore, I would
assume that the total generation of power by wind, could be somewhat of an
average if the units were located at suitable places over the area.
It appears that the wind units would not be large in relation to the
capacity of the system and with the wide diversity possible, that the units
could generate and feed power into the system in an amount and at such times
as wind would be available. This should not upset the system and should not
require any special system load control, dispatching, or scheduling.
With this sort of operation, it would not be necessary to provide

energy storage equipment which some presentations at this workshop indicate
would be quite costly and might even be hazardous and high maintenance.
Thus, the kwh fed into the system could be metered and the value might be
the incremental fuel cost at the location. This would dispense with in-
volved and expensive metering and controls and this could be the best
procedure at first. There seems to be some foregone conclusions that
storage was needed but it would appear that this would not be necessary
nor economical.
The fact that the units are relatively small compared to the load
of the system means that they could be connected into the low voltage dis-
tribution systems at small cost and avoid requirement of expensive trans-
mission and substation installations. The distribution lines cover most
of the rural, suburban, and urban areas and little or no investment in
such facilities would be required.
The utilities are interested in the wind energy conversion and would
cooperate in obtaining data and in studying the proposals that might be
made for installations and to assist in the plans and developments.
The utilities are interested in conserving energy and resources and

in the public welfare and providing energy at the lowest cost and maximum
reliability.
The units should be dispersed to get diversity of wind conditions,

appearance, environmental satisfaction, security of service and equipment
from storms and vandalism.--Towers over 100 feet tall might produce severe
objections.
228
Cost figures over the units which were presented at the workshop may
be in Line, and it would be necessary to determine the wind data, environ-
mental data, costs of installation, fixed charges, operation and maintenance
costs, and data, and revenue to be expected to establish the feasibility.
It appears, off hand, that the idea might be feasible and economic in cer-
tain locations and under certain conditions.
The utilities are regulated by the Federal Power Commission and by

State Utilities Commissions and whatever is done would have to be worked
out with these bodies.
Mr. Wharton:
Since our utility in Tillamook, Oregon, is publicly owned, I was

elected by our consumers; thus, it follows that I have to use economics as
a guideline. In Tillamook, we are interested in windpower, because of our
windy coast features and the Columbia Gorge wind exchange.
I would like to suggest some ideas that I feel would be of use to

the utilities in the near future concerning windpower. In the Northwest
only, where our power is hydrogenerated and it costs 3 to 3% wholesale
mil rate to our consumers, I think the 1 000 kilowatt size, or clustered
even to produce 1 megawatt, could be feasibly run into the grid. I
believe 5 or 6 mil wholesale rates should be the financial guideline.
Now, if the Northwest is going to use wind machinery, you're going to have
to realize that at present-day you're going to have a 5 or 6 mil economic
guideline, unfortunately. And, at a later date, possibly when they run
the nuclear plants that are now in construction into the grid at Bonne-
ville, you then will be probably be talking about 8 mils. Now, please
understand that these are just my opinions.
I would like to see a workable unit constructed and in use so that

the utilities could look at the costs and esthetic ecology. I would ask
you to consider a low profile. I realize this is not popular either, but
you are not going to get the ecology movement to hold still for a gang of
windmills hanging from a balloon or up on a wire. Anything that sticks
up and attracts attention, is going to be attacked by the ecology people.
I think 30 meters should be your extreme height. Now, this is going to
put you in a pigeon hole again. It may be that I am wrong; this could
easily be. But I think esthetics and public relations with ecologists
will demand this, not only in the Northwest, but probably nationwide.
Since the Northwest is not wind machine oriented, we must do a public
relations job. The wind machine will definitely be in a fish bowl, so it
should be engineered with an ecology and an economic basis in mind. The
energy squeeze is on us, and utilities will want to make commitment deci-
sions very soon. Therefore, I would urge you to use some haste on a
prototype.
I would like to suggest the diversity of location concept, so that a

wind machine would be run at varying times. I believe the Northwest would
not need any storage facilities, because of the inter-tieline to Southern
California, which we can use as an energy reservoir.
229
it will be extremely hard to sell the Northwest utilities on an
experimental involvement basis. They are not used to this. They are
used to an industry coming to them with a working model with the cost
available and an environmental impact study already made.
Mr. Lines:
I believe the matter of policy, with regard to not only Federal

agencies but the utility industry as a whole, has been mentioned. The
Federal Power Commission, as a regulatory agency, establishes its policy,
in general, by means of hearings, rule-making procedures, and the like.
Therefore, nothing I can say, as a staff member, can predetermine any
policy that the Commission, through its actions, might establish.
It may be of interest to review some of the responsibilities of the

Federal Power Commission. In the area of licensing generating facilities
and attendant bulk-power transmission systems, it is the licensing author-
ity for non-Federal hydroelectric projects on navigable streams or on
streams that affect interstate commerce. It has no licensing authority
over thermal plants or any other kind of plants. It has no siting respon-
sibilities or authority, other than those directly associated with the
hydroelectric plants previously mentioned. For me to guess how a type of
generation that has already been used in the industry would be considered
in the future as to Commission policy would not be warranted. I do re-
emphasize that a wind-electric generator has already operated in synchro-
nism with the interconnected network of the utility industry.
The Federal Power Commission's responsibilities, with regard to the

bulk-power systems of the United States, embraces, in the area of rates,
the approval of wholesale rates. These are the rates between contracting
utilities engaged in interstate commerce and not rates applicable to
utility retail consumers. The quantity of energy subject to Federal Power
Commission rate jurisdiction is relatively small when compared with the
energy subject to retail rate approvals. The states generally, through
state commissions or other agencies, exercise retail rate-making authority
as it affects the ultimate consumer.
In other activities, as detailed in the Federal Power Act, the

Commission collects data from all segments of the industry and dissem-
inates these data in many forms for public use. Many of these data serve
to keep the Commission staff current on trends and costs that affect the
consumer. The cost of delivered electric energy as affected by any means,
including the actions of the Federal Power Commission, is of interest to
the Commission and its staff. The electric utility industry in the United
States is the most capital intensive industry. Its capital requirements
are tremendous and constitute quite a problem. In advocating the exten-
sive use of wind powered generation, this capital intensiveness must be
kept in mind. Wind power offers a costless source of fuel, but from what
I have heard here, its use requires a very costly capital investment,
firm capacity-wise, fer an already highly capital intensive structure.
It is a combination of these fuel and capacity costs that will have to be
sold to the industry. I would not take exception to the range of costs
230
that the preceding gentleman has mentioned, but wind powered generation
does not appear to be currently competitive. One must also consider
logistically the magnitude of the industry and the size of generating
units employed in relation to the size and cost of the wind powered units
discussed here for integration into the industry.
Mr. Jessbp:
I am with the Power Plants Branch of the Rural Electrification Ad-

ministration. We are probably responsible for putting more windmills out
of action than any other agency. Economics is the name of the game from
where I sit, I am not a policymaker; I deal with power plants and make
recommendations on their cost of installation and their cost of -operation.
Sometimes I overestimate, and sometimes I underestimate. Nine times out
of ten we've got enough funds to build the things, but that one time out
of ten we run short of money. So costs are pretty important to us. I
want to make clear that REA-financed powerplants are still only about 1%
percent of the total installed U.S. capacity. Even to our borrowers, we
are in a minority. I should explain that our borrowers are mostly rural
electrification cooperatives who have taken on utility responsibility in
sparsely settled parts of the country. These cooperatives have brought
up the number of connected households from 10 percent of the total in
rural areas, in its inception in 1935, to about 98 percent recently^
So there is very little left that hasn't been covered. We have

spread the tentacles of central station power service throughout the high-
ways and byways of America until we have pretty well completed the job.
But we are still working in a very, very remote areas. We are presently
connecting plants in Alaskan villages and small settlements in the United
States which are still without central power service.
As a consequence, the spread in the cost of power to these facilities

is quite great. We have many borrowers who are able to generate power at
around something like 1 cent, not 1 mil. And then we go up, in the remote
Alaskan villages, to power costs which are more like 8 cents, 5 to 8 cents.
But, if you are looking toward large applications, then I have here the
Thirty-Second Annual Report of Energy Purchased by REA Borrowers. The
costs have drifted down from about an average of 1.1 cent, when these
figures were first compiled in 1940, to about 0.65 of a cent in 1965.
After a flattening out process, they have again turned up; they are now
going up quite rapidly.
But, so the past is prologue, and it isn't necessarily a good indi-

cation of what the future will be. Costwise our plants are going to cost
more because of the environmental features. Large central station power
generators, which we have built at a cost of about $120 to $170 per kilo-
watt, depending on the type of fuel we burn, are the cheapest because
there are fewer environmental considerations and less fuel processing.
Coal-fired plants are more expensive; they range up to about $200 a

kilowatt if we're in lignite fields. Possibly because of the things which
must be added due to environmental considerations, we might have to go up
231
to $4-00 per kilowatt for these central power stations. Costs for plants
in remote areas, for example, Alaskan villages, have ranged from about
$225 to $300 per kilowatt. These plants are small diesel powered ones.
All of these plants have one common characteristic: they can come
on-line when the demand is there. They do have surplus capacity. In the
case of the Alaskan plants, if one unit fails, or is taken out for main-
tenance, there is another unit standing by.
So these are reliable peak-supplying systems. I might make a comment

on storage. Time is of the essence, and if you are dealing with a system
with no storage, then you've got to compete against incremental fuel costs
as far as the marketplace is concerned; this can be very, very low, even
as low as 2 mils. But, again, fuel costs are going up, so here again
you're aiming at a moving target. So you've got the choice, possibly,
between storage competing with 2 to 4 mil power, without storage competing
with 2 to 4 mil fuel costs, or with storage competing with 2 to 4 cent
power costs. This is on peak power.
It seems to me, from what I've heard as an individual, that heat has
the best application, because thermal transience is the longest. Not
maybe as long as the transience you get with wind velocity, but they may
be catching up on the deal, and thermal storage is usable feasible. , Rocks
are cheap, water is cheap. Put together a pool in the basement with rocks
and water in it, and you've got a real good thermal tank.
Mr. Douglas:
The Boeing diversification program includes consideration of various

forms of energy. One source of energy that we have worked at in 'some
depth is that available from municipal, commercial, and industrial waste.
As a systems problem, its economics depends on a balance of refuse collec-
tion, processing, and marketing of the products. Similarly, wind energy
is a systems problem involving technologies, siting, and customers.
We have heard that rotor technology, suitable for wind energy plants,
is available today. The real question is whether we can produce electric-
ity at a price of 3% to 7% mils per kilowatt with the technology that is
available and for the conditions prescribed by the power companies. If
not electricity, can we pump water, run miils, or convert wind energy
into useful work competitively in any market for energy?
I believe that our demonstration programs should be aimed at the

systems problem rather than proof of concept of a component in isolation
from its matching elements.
Mr. Mockovciak:
Our activities in the Grumman Aerospace Corporation's Energy Systems

Group are twofold: one, energy conservation, and two, solar energy appli-
cations (which include the use of wind energy).
232
As regards our potential role in this area, we see ourselves as
possible manufacturers of wind machines. From our standpoint, therefore,
the existence of a market for wind energy machines is a major concern.
We think that it exists or could be created, but advocates of wind energy
must go out and "beat-the-bushes" to find people who really want to use
it. A user motivation, I think, is a key to future wind energy usage.
For example, one of the suggestions that was made in a committee

meeting last night is that perhaps there are many universities that would
like to use wind energy, both as an actual energy plant supplying elec-
trical power to its facilities and as a relevant engineering project in-
volving both students and faculty in its design, construction, and opera-
tion. This example is illustrative of who could be real users of wind
power. To reiterate, I believe that advocates of wind energy must identify
the people who want to use it.
There are two ends to the wind generator spectrum: the small local
wind generator application and the larger, utility-type wind generating
machine. I personally feel that the utilities (since they are generally
conservative) are going to take a "wait and see" attitude. Therefore, I
believe that the place to begin is in the smaller, localized wind genera-
tor applications. When the utilities begin to see that wind machines do,
in fact, provide electrical power, that people are satisfied with their
performance, arid that a base of operating experience is being developed
across a reasonable spectrum of wind generator sizes, then I think the
utilities will sit up and take notice.
Most important, in my view, are (1) acquiring operating experience

and (2) establishing realistic costs. These two factors, if anything,
are going to convince the utilities that wind energy can be a useful elec-
trical power source.
One of the interesting aspects of the wind energy business is that

there is an existing and adequate technology base. By that I mean that
there is an adequate base to begin to engineer wind machines for power
production. It disturbs me, however, that many advocates of wind gen-
erating machines call for more research and development. I would almost
call this an "R & D syndrome." I frankly think there's an overemphasis
on the amount of research and development that has to be done, and too
little emphasis on finding ways te make it happen. I personally feel
that we have an adequate technical base and that we should start thinking
about building these machines and looking for people who want to use this
energy.
Obviously, the people who may be interested in using wind energy are
also going to be concerned with its cost. In this regard, I think that
the government could play a key role. In the near term the government
could offer direct subsidies in order to get the machines out where people
could see them operating and performing useful functions. In other words,
what I am suggesting is that the public (through the Federal government)
could make wind energy "happen." If the public wants to use wind energy,
the government can make it happen; the same way that we've made a
233
tremendous highway system happen in this country, and the same way that
we've provided direct subsidies for public housing where we want housing
to happen! The government mechanisms are there, if the public motivation
is also.
Moderator, Dr. Morse:
We will open the floor to questions. We are interested, from the

point of view of industry, other government agencies, and utilities, in
what role the Federal government might play in this wind energy program.
What would it like to see done so that it could make the kind of deci-
sions that it will have to make.
Are there any questions to the panelists?
Mr. Cohen:
It occurs to me that perhaps it might be worth looking into the

possibility of coupling the high quality mechanical energy generated by
windmills to heat pumps and storing that heat rather than degrading the
energy and the heat to begin with.
Dr. Morse:
It's a thought.
Mr. Lutzy:
It's the basic role I am attempting to do. In other words, don't go

through intermediate investment steps for high cost investments and
gadgetry. Get that mechanical heat directly to the heat pump to improve
its efficiency.
Dr. Morse:
By the way, NOAA is an agency that certainly has a role to play in

that new wind energy program, and I didn't mean to exclude you.
A Voice:
I would like to ask what the proper mechanism of our existing govern-
mental structure is for the government to decide to make it happen. They
decided to make nuclear power happen, and it happened. We all have our
own opinions on the way it happened, what it cost us to have it happen,
but where do we start? What can we as private citizens do to make it
happen?
Mr. Lines;
If I may, I will respond in general with more of a personal observa-

tion than a recount of any staff position of a Federal agency. I call to
234
your attention two related items that might be of interest. One is in
the June 1973 issue of Spectrum, the I.E.E.E. publication. It lists two
pages of Federal agencies involved in energy policy and problems to some
extent. The other item includes the President's energy messages of about
18 months ago and just recently. Over the past weekend, the President
announced a proposed reorganization which included the establishment of a
Cabinet-level agency in which energy and resources considerations would
be centered and which would affect to some extent, among others, the
Department of the Interior organization. There were other proposed changes.
This proposal will be sent to Congress for its approval or action. The
atomic energy program was established in a similar manner by the Act which
established the AEC. The Joint Atomic Energy Committee is the congressional
body that is actively interested in that program.
To answer your question directly, we have an outlet, through our con-

gressional representation, to establish and make known our personal view-
points and what we as individuals think of both the energy policy of this
country and its many, many ramifications.
Dr. Savino:
As I listen, I get the feeling that among certain government agencies,

such as the Rural Electrification, the FPC, and the power utilities, the
attitude appears to be one of business as usual. We talk about things
having to be cost effective as though we can continue supplying the public
with all the power it can consume, as though we had plenty of it. Every-
one here is aware of the fact that we are on more or less a collision
course over the next 20 or 30 years. And there seems to be a reluctance
to take the action that is necessary before we get into a real bind. I
believe it was Dr. Hutter last evening who mentioned that we as individ-
uals accept many dislocations in our lives. We have automobile accidents
that cause a dislocation, or we lose our job. There are a number of such
things that happen, yet when we push the light switch or decide to drive
somewhere we want power available to us at that instant. We don't want
any dislocations in our energy supply. Isn't it time the public utilities,
as well as the agencies, tell it to the public like it is? We are going
to have to start paying higher prices. We must stop chasing the demand
curve. Shouldn't your companies and agencies also start getting involved
in supporting alternative systems? Many of us believe we can no longer
look to nuclear or fossil or any one or two systems to provide all the
energy. There is going to have to be a mix, and we cannot have this mix
unless we have the involvement of the utilities, the agencies, as well as
the people who are proponents of this system. We, the proponents of wind
power, can never create the environment necessary to move forward until
you people who have the influence get involved instead of sitting back and
waiting for someone to come forward and say here's a package that works —
would you like to use it?
Mr. Loftness:
As a matter of fact, that is precisely the reason EPRI was estab-

lished by the utilities. They recognized that business as usual - the way
235
it had been going on for many, many years - was no longer the case.
Obviously, EPRI will have activities that relate to conventional

technologies - improvements in transmission and distribution, in nuclear
power generation and related safety questions, and in fossil fuel gener-
ation - all to improve the ability to generate power with the equipment
we now have. At the same time, EPRI will have a group on advanced tech-
nologies that will be concerned with evaluations, by the in-house staff,
of the status of solar, geothermal, wind, ocean thermal gradient, and
other alternate energy sources. Hopefully, there will be major inputs of
information from individuals who are experts and who have been working in
each field for a long time. These inputs will be judged, in terms of
future research and development support, by the contribution each alterna-
tive might make to the total energy mix.
There is nobody I know in the utilities who feels it is possible to

go on forever using oil or coal as they have in the past. There is a lot
of current interest in fission power and fusion power, but there is also
interest in looking for alternate sources of oil - from the liquefaction
of coal, for instance - and in alternate sources of gas through coal gasi-
fication. These coal conversion processes are now more expensive than
natural oil and gas, but, nevertheless, they are processes the utilities
feel must be developed if they are going to continue to generate power to
meet the needs of the country.
I think the utilities have given evidence of their interest in pur-

suing alternate energy technologies. I don't think people interested in,
say, wind energy are any longer voices in the wilderness.
As I mentioned, the Institute is very much interested in having all

the factual information it can possibly obtain. We can't possibly develop
it all ourselves. We should proceed from the basis of information that
already exists, and I would encourage all of you to submit the information
you have to the Institute.
EPRI will not be alone in developing new technologies. For example,

both NASA and the National Science Foundation have activities in solar
energy and we will be working with them in this area. In any particular
technology, we will be judging what needs to be done based on an assess-
ment of the relative contribution of that technology. Where wind energy
would come out in such an assessment, I don't know.
The assessment process would involve all the individuals who are
interested in making a contribution to the argument on what should be done
or what should not be done. Out of this dialogue, I believe there will be
a decision among interested organizations, including the Federal government,
that certain technologies should be funded as a national effort - as is the
case now for solar energy. As you know, such an assessment has resulted in
a tripling of the solar energy budget of the National Science Foundation
in the past year.
In summary, I don't think there is a lack of recognition for the

236
importance of new energy sources either in the government or in industry.
Mr. Lutzy:
It appears that we have come from a nation of abundance to a nation

of scarcity, and that involved in this is basically the national security
and.the question of the balance of payments problems in regard to the
value of the dollar on the world market, in regards to our being able to
compete in the world that exists and to maintain our standard of living.
I would think that, based on our feelings in the past, we did not want to
be substantially committed to surviving based on foreign energy supply,
as an example; this would demand a commitment in the very near future by
the national government to solve these problems and to dp so by the capa-
bilities that exist in this country. The other thought is that, in regards
to capital investment, we all are schooled that we must be economically
competitive and must maximize the use of an investment, good or not. In
our personal lives the percent use of any investment that we put our money
in doesn't have to be justified. As an example, don't buy a generator to
generate a little power, but buy a boat for $1400 or a snowmobile and put
them in the corner. What's the percent use of a snowmobile, plus the con-
sumption of energy in this? There is no comparison between this, so there
is a question that maybe there ought to be a play opera between personal
investment and business investment whereby some of this personal invest-
ment is put to a more meaningful use.
Mr. Mockovciak:
I'd like to address myself to EPRI's future role, but I must qualify
this as a personal observation. Coming from an industrial organization
and being heavily involved in past research and development activities, I
have observed that the basic function of (what are called) new businesses,
advanced programs, or research and development organizations is to promul-
gate the current business line. In other words, the "new business" aspect
is a misnomer. It's called new business, but it really means keep the old
business going.
In this regard, I can't help but observe, when exposed to the kinds
of things that EPRI is proposing to do, that they are planning to do just
that - keep the old business going. EPRI appears to be largely interested
in improving the operations and performance of existing electrical power
systems or those that already have extensive research and development
bases. Thus, I can't help but feel that there really is no motivation
there to make anything new happen.
I would, therefore, suggest that EPRI strongly consider sponsoring

the engineering design and development of actual wind generating machines
for a number of regional locations. Since there is no new technology
needed, the machines can readily be engineered. There is nothing to prove
by research and development studies, but much to prove by operating the
machines: namely, prove that wind generators can supply electricity reli-
ably, can be operated for long durations, and that they are or are not
economically practical.
237
I am very much concerned, as I mentioned earlier, about this "R & D
syndrome." We always seem to approach a problem saying that we need to
institute a research and development program. I don't think that's the
case with regard to wind energy. I think there is an adequate technology
base that can be used to build wind power machines. Once we get some
operating experience under our belt across a spectrum of sizes, that ex-
perience should point the way for new research and development directions.
Furthermore, if the initial wind generating plants-indicate that they can
be made economically competitive, the research and development would likely
become more economic as opposed to technically oriented.
In the nuclear field the utilities have worked out the economics of
the atomic business. Right now it's costing about $550 per installed
kilowatt. This does not include the decommissioning, and it has recently
come out in hearings that the decommissioning of one of these plants, so
the land could be reused, would cost much more than it did to build it.
In addition to that, you have each of these thousand megawatt plants
producing over a million pounds of radioactive waste that has to be taken
care of. l! think all of these costs have to be added in when we start to
compare windpower with the other forms of power.
Mr. Schwartz:
A very important point was made; that is, it is difficult to see how
all this can be made to work without the government interacting with the
utilities. Dr. Starr testified recently before Congress that he didn't
feel the Federal government had a role in deciding what kind of energy
research should be done. That's best left to the utilities and suppliers.
I wonder if there is; does Mr. Loftness have a comment on that?
Mr. Loftness:
I don't recall that particular statement. Was he stating that the

government should not decide what the utilities should do, or what EPRI
should do, or what the nation should do?
Mr. Schwartz:
His comment was he didn't feel the Federal government should be in-
volved in energy research related to utilities and their suppliers; they
can decide better what they could carry for policy.
Dr. Morse:
I think that in fact the increased funding in solar energy would

counter that statement. I think there are Federal funds going into the
development of new energy sources. In MHD there are significant funds,
and in geothermal the same. I think there is that indication. I think
that Joe Savino's comment as to how to get the utilities to take an active
role, or to get the user to take a more active role, is a relevant point.
238
Mr. Schwartz:
The words "significant funds" have been used several times in connec-
tion with the forthcoming budgets in certain areas. I hope we will know
about that this afternoon. But I wish to ask Mr. Loftness, or anybody
else who has data on the subject, if there is anyone who can characterize
the funding that is going into all forms of energy sources in terms of
the annual capital outlay of the utilities?
Mr. Loftness:
The figure that Dr. Stever used in testimony before Congress was a
Federal budget for next year of $772 million for energy research and
development: about 65 percent was for nuclear research and development
and the rest was spread across all other technologies. He used a figure
of $1.2 billion being spent on energy research and development in the pri-
vate sector - by industry, by the utilities, and by other organizations
in non-Federal, funded activities.
Mr. Schwartz:
Is that engineering of new power plants using existing concepts?
Mr. Loftness:
He didn't break this down, so I don't know how he arrived at the

figure of $1.2 billion for the non-government activities. EPRI itself
will have a budget next year, supported by both the private and the public
utilities, of a bit over $100 million for research and development. This
compares to the $772 million for government-supported research and develop-
ment work. I really can't imagine that Doctor Starr said that the govern-
ment should not decide what research and development is important. I am
sure, however, he would say that EPRI is not looking to the government
for all of its direction on what should be done, and I don't think it will.
I would expect, however, that many of the programs that EPRI will have will
be cooperative programs with the U.S. Government. We already have a coop-
erative program in coal gasification, for example, and I would guess that
programs in geothermal development, if we have them, will be in cooperation
with government programs. There will be a lot of joint planning of activ-
ities, even if there aren't joint programs in the sense of being jointly
funded. We are talking to the AEC about several joint programs; we feel
the programs are impprtant and they feel the programs are important. No
one organization has either all the wisdom to decide what needs doing -
or all the money to support every program.
Dr. Morse:
I think we could go on talking for quite a while on this topic, but

we do have a session coming up in which the NSF and NASA programs will be
discussed. Therefore, I think we will adjourn at this point.
239
NSF PRESENTATION
Frederick H. Morse*

Washington, D. C.
The following is a brief summary of the National Science Founda-

tion^ energy conversion program, which is one part of the NSF solar
energy program. Also included is a comment on NASA's involvement in
the wind energy program.
Perhaps the best way to begin is to look at NSF, then the RANN
(Research Applied to National Needs) program, the solar energy pro-
gram, and, finally, within the solar energy program, the wind energy
conversion program. For the purposes of managing research programs
addressed to national needs, the National Science Foundation has
organized components of its coordinated and problem-focused research
into the Directorate of Research Applications. A major activity in
this Directorate are the RANN programs. The Director of the RANN
programs is Dr. Eggers. The authority of the NSF to become involved
in research that is directly related to the problems of society and
the environment was enhanced by the provisions of ammendments to the
NSF Act in 1968. The NSF is therefore directly engaged in research
programs that are related to social and environmental problems as well
as the potential impact of future technological development. Solar
energy is just one example of the new technology in which NSF, through
RANN, is deeply involved.
While the emphasis of the RANN programs is on problem orientation,

NSF supports fundamental and applied research through the divisions of
the NSF Research Directorate. In many cases, programs within the
Research Directorate are related and complimentary to certain RANN pro-
grams. For example, within the Research Directorate there is a pro-
gram of global atmospheric research; also, there is a wind engineering
program which is concerned with the interaction of wind with structures.
Within RANN there are four divisions: Environmental Systems and

Resources, Social Systems and Human Resources, Exploratory Research and
Problem Assessment, and Advanced Technology Applications. The Advanced
Technology Applications Division is concerned with new or improved
technologies that can enhance economic productivity, exploit the
potential contributions of advances in science and technology, or
stimulate those applications that will contribute to the solution -of
*Present Address: University of Maryland, College Park, Maryland
240
some major national problem, such as the energy "problem. The solar
energy program is centered within the ATA Division. The NSF solar
energy program was initiated in fiscal 71. .In July 1971, the President,
in a special energy message to Congress, called for programs to provide
the nation with adequate sources of clean energy. Shortly after that
message, the Office'of Science and Technology, through the Federal
Council of Science and Technology formed eleven panels, in the various
energy areas, to establish Research and Development goals for those
energy technologies. The National Science Foundation and NASA were
asked to jointly organize a solar energy panel. This panel was estab-
lished in January of 1972.
The Solar Energy Panel consisted of M-0 individuals from universities,
industry, and government, with backgrounds in electrical engineering,
mechanical engineering, solid-state physics, chemistry, biology, and
architecture. Also included were a sociologist, an environmentalist,
and an economist. The panel assessed the potential of solar energy as
a national energy resource. The scope of the Panel included direct
solar energy applications as well as the indirect applications - wind
and ocean thermal energy and renewable organic fuels. In December 1972
the report of the Solar Energy Panel was released. This report is avail-
able from the Solar Energy Panel, Department of Mechanical Engineering,
University of Maryland, College Park, Md. 2090M-.
The panel's key recommendations are that the Federal Government take
the lead in developing research and development program for the practical
application of solar energy as an alternative energy supply to meet the
heat and power needs of the United States and that this program be a
simultaneous effort in three areas - economical systems for heating and
cooling buildings, economical systems for reducing and converting
organic materials into solid, liquid and gaseous fuels, and economical
systems for generating electricity. The Solar Energy Panel identified
seven areas, as most promising from technical, economic, and energy
standpoints. These are the following:
(1) Heating and cooling of buildings
(2) Photovoltaic energy conversion
(3) Solar thermal energy conversion
CO Wind energy conversion
(5) Ocean thermal energy conversion
(6) Photosynthetic production of organic matter
(7) Conversion of organic matter into fuels
Sometime after the panel's release of the report, the National
Science Foundation presented a 5-year program to the government, and it
was given the responsibility for the terrestrial solar energy program.
The objective of this program is to develop, at the earliest feasible
time, the many applications of solar energy as alternative energy
sources. An interagency panel was recently convened to inform and
coordinate the activities of other agencies such as NASA, NBS, NOAA, DOD,
AEC, and others in terrestrial solar energy applications. This inter-
agency panel presently meets on a monthly basis.
A brief comment on the funding. In fiscal 71, $1.1 million was
spent on terrestrial solar energy projects. In fiscal 72, that funding
was $1.6 million; in fiscal 73, it was $3.8 million; and in fiscal 74-,
the estimated budget is $12.2 million.
There are many ways of collecting and converting solar energy into
electrical energy. Solar energy is collected naturally in the EarthTs
atmosphere, which gives rise to the wind. It also warms the surface of
the ocean, thereby establishing the temperature gradients therein. And
it is collected on the surface of the Earth, a fraction of which is cap-
tured by the photosynthesis process.
In addition, man can construct collectors, such as solar cells, to
convert solar radiation directly into electrical energy, or concentra-
tors to convert solar radiation into electrical energy by means of heat
engines such as those operating on the Rankine cycle.
Let us now turn our attention to the wind energy conversion program.
The objective of this program is to develop reliable and cost competitive
wind energy conversion systems that are capable of rapid commercial ex-
pansion to produce significant quantities of energy on a national scale.
There are many technical challenges to face in meeting this objective,
such as performance predictions, configuration tradeoffs, failure mode
analysis, development of low-cost structures, etc. There are also many
environmental, social, and economic programs involved in the large-scale
extraction of energy from the wind; for example, the environmental impact
at these systems on the marine or plains ecology, or the institutional
constraints on these systems, and so on. Time does not permit a full
discussion of these types of problems.
In fiscal 73, the NSF wind energy conversion program initiated three
projects. One project was a grant to the NASA-Lewis Research Center to
organize and conduct this workshop. The second project, at Montana State
University, will identify the major technical problems of the tracked air-
foil system that were previously described at this meeting. A grant to
Oklahoma State University was also awarded for the development of a
variable input-constant output generator and an electrolysis units in
the 10-kilowatt size, suitable for integration in a wind conversion sys-
tem. In fiscal 73 these three projects totaled about $300,000. In fiscal
74-, a funding level of the order of a million dollars is anticipated.
NSF will use the phase-project-planning approach. This approach
consists of an orderly progression from Phase Zero, in which the con-
ceptual design and performance requirements are specified, into Phase
One, where the preliminary design is made and the critical subsystems
are researched, designed, and tested, and finally into Phase Two, where
the proof of concept experiments are conducted. In the NSF wind energy
conversion program we expect to be through Phase Two within our 5-year
program. The remaining phases, Phase Three (demonstration system design,
construction, and testing) and Phase Four (commercial system design con-
structing and testing) are left to the user. In parallel with Phases
One and Two, research on components and advanced concepts will be con-
ducted on a continuing basis.
242
In carrying out its responsibility for the solar energy program, the
National Science Foundation will involve universities, industries, and
government agencies on the best-performer basis. As you know, NSF can
and does award grants to universities. In addition, the NSF/RANN program
can award contracts to profit-making industry. It is no longer necessary
for industry to join with a university, as a subcontractor, in order to
receive support; however, in many cases the resulting joint effort is
stronger than either alone. In the wind energy program, as well as other
areas of the solar energy program, NSF/RANN will continue to encourage
and accept unsolicited proposals which represent the ideas generated by
people in universities and industry. In addition, NSF/RANN forsees the
release of program announcements and RFP's as the program develops.
Unsolicited proposals should continue to be sent to NSF. NASA Lewis
Research Center has reviewed all proposals in solar energy and will con-
tinue to do so.
In the area of wind energy conversion, the NASA Lewis Research
Center has expressed a strong interest in the program and has been asked
to prepare a plan indicating how they would support the wind energy con-
version program. While the details haven*t been finalized, it seems
quite certain that the NASA Lewis Research Center will play a very major
role in the implementation and execution of this national wind energy
program. In the following presentation, NASA will describe some of, the
details of their proposed plan.
243
NASA PRESENTATION
Ronald L. Thomas

Cleveland, Ohio
Good afternoon, ladies and gentlemen. I will briefly go through a

program plan here that we've worked up. We've been taking a hard look
at this area for the past year, and also other solar energy areas; and
we have put together a preliminary program plan in wind energy which we
have discussed with the NSF.
Obviously, as a result of our workshop, there are going to be a lot

more inputs into this plan than you're going to see now. This plan, I
want to stress, is tentative. But it will give you an idea of this five-
year program. First I will briefly review the objectives, the approach
and the planned accomplishments for this program.
The objective is to develop a wind energy system that supplies

reliable energy at a cost that is competitive with other energy systems.
This is a government-directed industry program with strong university
support. We plan to set up a review board which would consist of utili-
ties, manufacturers, consumers, and the appropriate municipal, state,
and federal agencies. This board would be like an advisory board. It
would meet and review thevprogram from time to time, and offer valuable
inputs as to the way we are proceeding and so forth, which areas should
go forth faster, which slower, that type of thing.
We feel very strongly that there would be an effort, within our na-
tional laboratories, to provide the basis for program direction.
We would study, build, and test wind energy conversion systems and
components without storage. We would also study, build, and test energy
storage systems. And I will go into the reason here: We feel that we
should get on quickly with the job of the wind energy conversion system,
and not necessarily tie it directly in with the storage right at the
beginning.
A very important facet is that we will conduct meteorological studies

in cooperation with other agencies, particularly with NOAA, to estimate
the wind energy potential and determine favorable regions and sites for
wind power. Another important part of this overall program is to study
and identify the suitable applications for a demonstration test — just
what is the overall potential and what are these applications? At the
end of 5 years we hope to have identified cost-effective wind energy
244
conversion systems, not necessarily with storage tied to them. And we
would have these systems in operation and have some data to back this up.
Not only will we have some of these prototype subsystems in operation,
but we will have had bench tests and subsystem tests of the key components
going on within industry and within universities or the laboratories.
Demonstration systems for selected applications and some of these,
hopefully, with storage would be at the point where we would be ready to
begin tests at the end of 5 years. It is unlikely that in 5 years, we
would actually have systems complete with storage and for actual applica-
tions ready to go.
At the end of five years we hope to have determined the potential
for wind energy in this country. We would also plan to have under develop-
ment analytical techniques for selecting sites for wind conversion plants.
Next, I will discuss several action diagrams to point out the key
phases of the overall program. The key phases are the wind energy con-
version system, the meteorological studies, the energy storage system,
and environmental impact studies.
In the overall program (fig. 1) you can see we would be carrying
on a number of steps in parallel, starting in the first year. This
would be the study, design, build and test of wind energy conversion
systems, without storage. We gather and assess wind data to come up
with the favorable site selections and what is the potential of wind
energy. We would study, design and build and test energy storage
systems. We would identify and study the suitable applications for
wind energy, concentrating on the most favorable ones in the beginning
and coming up with what are the practical applications for wind energy.
Then, we need to determine what the requirements are for major

facilities. We are encouraged by what came outôf the workshop here. It
looks like a lot of the components and sub-systems can be adequately de-
signed with simulations and modeling.
All the above phases are parallel and will focus on the design and
demonstration for those favorably selected applications. We would then
construct and test those demonstration systems. That completes the over-
all program.
Now in figure 2, looking at the wind energy conversion systems

which are primarily the towers, the rotors and the electrical genera-
tion. Where should we begin, and how can we put a wind generator
together that meets our requirements. First we started with systems
design studies of wind conversion systems. We would concentrate on
those without storage at this point, and these would be to identify
the size of machines that makes sense, hopefully focusing these into
applications.
We then select the ones with the most promise to actually have pro-
totypes running within the 5 years. We may pick out one or more combina-
tions for detail design of these prototypes. Once we have a detail design
245
of these prototypes, we would then select the most promising one and pro-
ceed with construction of that prototype. We would also at that time
begin bench testing and in-house testing of the components and subsystems.
Also in the systems designs we would come up with some advanced con-
cepts. These would be split out and paralleled to start studying these
advanced concepts, deciding which of these should be built and tested,
primarily at the component level.
What we have is a program going down three paths leading toward the
construction of prototjypes as quickly as possible, probably building very
heavily on the technology that was discussed by Dr. Hutter, and breaking
off in parallel component tests and modeling of rotors and the key elec-
trical conversion and Îso
i with advanced concepts.
i
We would test these prototypes within the 5 years, and all of this
would be input, then, into the design of the demonstrations for selected
applications.
In the area of the meteorological studies (fig. 3) we have mainly

started out to assess what the existing wind data are. Obviously, we
will need very close cooperation and coordination with an agency like
NCAA. The purpose is to identify sites for these first prototypes ajid
pick out the favorable sites. We must get the wind data we need to do the
the detailed designs of the prototypes.
Also we want to determine the wind energy potential in the U. S.
Using the existing data we would try to determine that potential. We
would determine the favorable wind sites in the U. S. Also we would
develop analytical techniques for wind prospecting so you don't have to
smother the landscape with anomometers.
We would make additional measurements, wind measurements that are

required in areas where there isn't sufficient data and feed this back
to up-date the wind energy potential estimates.
In the area of energy storage (fig. 4-) we would again look at the
various types of energy storage systems that are available. We would do
some design studies and select the systems that look the most promising
for fabrication and evaluation. In parallel, there are going to be some
advanced storage systems identified. We will do design studies on these
systems also and build and test those advanced components. All of this,
again, leads into the demonstration systems occurring after 5 years.
For the conversion systems in Fiscal 197M- we would immediately start

with several systems design studies, followed by a detailed design of the
prototypes that appear to be the most promising. Remember that we have
a key point here where we bring in the advisory groups to help in the
selection of this. We begin construction of that prototype at the begin-
ning of Fiscal 1976, and should be able to have a prototype and start
testing at the beginning of Fiscal 1977. We also build and test the
prototype components and sub-systems as soon as we've identified the sys-
tem that we have selected.
246
The meteorological studies are an on-going program. We envision
these studies would assess and make use of existing wind data, set up
instruments for making additional wind measurements as required, and
develop analytical techniques for wind prospecting.
The key points are that by the middle of Fiscal 1975 we would have
those prototype sites, and have made an up-dated determination of just
what the wind potential is in the United States in these selected areas.
By here we would identify these suitable sites for selected applications.
I haven't gone through all of the program here but very briefly have
tried to give you what we see as the approach at this point to getting on
with the job. The key thing is to do these things in parallel, carrying
them out, and to constantly be evaluating the direction we are going in.
DISCUSSION
Q: When do you expect to be able to make announcements on this or a

similar program being started?
Dr. Morse: I would say within the next several months.
Q: You announced the NSF funding program. How about the NASA?
Dr. Morse: The funding is for wind energy conversion. The projects or
part of the program that NASA would conduct for NSF and the grants or
contracts that it would issue and manage would be from those funds.
Mr. Thomas: And as I'm sure you are all aware, NSF is the lead agency.
They have the solar energy program. What NASA is willing to contrib-
ute at this point is manpower and facilities. We have laboratories,
experimental capability, and analysis and project management. That
would be NASA's contribution.
Q: Out of this million dollars from NSF? '.

Mr. Thomas: Yes.
Q: What is left over?

Dr. Morse: It's all left over. They are not going to pay their salaries
out of that. NASA Lewis is contributing the manpower, the facilities,
and the experience to the program. The money that I mentioned that
NSF has will be spent partially through NSF and partially through
NASA Lewis. Again, this is a program with strong university and in-
dustry involvement. We expect the bulk of that money to be spent
externally.
Q: Are there any sorting criteria that you can give us for which type
of proposals go to which, or do all of the proposals go to. NSF for
this time period?
Dr. Morse: All RFP's and program announcements will state clearly on
whether that should be directed to Lewis or to NSF. At the moment,
unsolicited proposals should go to NSF. They will be reviewed in
the usual way except that now Lewis will be included as a reviewer
on all proposals. I might also mention that proprietary proposals
will be handled accordingly.
247
Q: There seems to be an incompatibility — maybe it's only by inference —
between system design and your technology in the position phase. You
have a systems design that goes right on into the concept. You have
a technology phase which says there are some technology gaps which
have to be identified. Are you fellows getting together on this?
Dr. Morse: Yes, we are.
Q: That's one question. The second question is: Is your objective

merely to demonstrate a large rotor that can effectively turn out so
much power in some kind of wind, regardless of -what you use it far?
Or is it to integrate an objective as to how you will use this energy
in a system?
Mr. Thomas: The objective is to discover how you use the energy and
whether or not it can be practical and competitive.
I think the other point is, too — I mean, you clan sit back and show
all these little magic blocks up there on the Vu-graph. The way those
were arrived at is: What is a reasonable objective and goal to get
accomplished by the end of 5 years?
We set down our plan, what we thought we could do. Then those blocks
are really backed up by the actual tasks that have to be required in
each one of those areas, in terms of manpower and dollar expenditures
and what the hopeful output will be for each one of those blocks. It
would be a number.
Now, I won't go into detail on those, because, if I put up the defi-

nition of those tasks, give you the outline of a work statement and
told you what the manpower was and what the dollars were, there
wouldn't be much sense in putting out a competitive RFP because all
proposals would come with the same costs and the same amount of man-
power.
Now, obviously, what's come out of this meeting, we are going to get
criticisms, good and bad; depending on how much money NSF is going to
divert to this NASA effort depends on how fast we go down this path.
But we look on this as a joint venture with NASA and NSF, and to go
down those parallel paths together in the best and the optimum way
overall.
Q: Dr. Morse, does this mean that the proposals which have been submit-
ted this year already on the subject are — to use a word of Mr.
Zeigler's — inoperative?
Dr. Morse: No, that isn't so. We have ten unsolicited proposals, two of
which have been reviewed. The other eight are in various stages of
discussion and formalization. Some of these proposals may very well
be supported, based on their own merits. It may be that some of these
proposals fall very close to what we have in mind for an RFP, in which
case we would not go ahead with that, and indicate to the principal
investigator that, at the moment, we are going to hold off on that.
248
Q: Will the storage facilities or the mechanism for the program selected
for storage involve a concept of shipping?
Mr. Thomas: Can you elaborate on that? I'm not sure I know what you mean.
Q: Does the use of wind energy involve shipping in any way?

Dr. Morse: You mean like producing hydrogen and shipping it in some form?
Q: No. For sail, the idea being if the storage mechanisms were practical
enough — is that ruled out?
Mr. Thomas: We haven't really considered that, although the FCST panel on
transportation did recommend that the old sailing ships may hold a lot
of application today and that work be done in that area.
Q: I noticed in your 5-year plan you didn't have any kind of estimates
or projections on what kind of money would be allotted for wind re-
search. Will you comment on that?
Dr. Morse: I really don't think it is appropriate to comment on that,
since one never knows how those will turn out. We do have a 5-year
program. We have worked out a budget for all those areas, and the
only figure that is really a pretty real one is $12.2 million for
next year.
Q: It appears that you really intend mostly study programs the first
years, instead of hardware. Is that true?
Mr. Thomas: That's true, but those study programs are the first step in
really defining and identifying which way to go on the hardware.
Dr. Morse: Although I might indicate the program at Oklahoma State is

looking at two key components of the system, and that is a hardware
type of a program, the bulk of it is systems studies.
Q: You mentioned earlier you have provisions or are you making plans
for industry-university participation? Where in that program is the
flexibility to do that sort of thing?
Mr. Thomas: It seems very clear to me, since I drew all the diagrams,
but at the end of the first systems studies which will be in the
first fiscal year, we plan to do several parallel studies in the same
area. Each of these studies would probably have several different
concepts. The point then is to pick out the system or systems that
have the potential for practical completion within the 5 years.
Those systems would be reviewed by the advisory board. The advisory
board would be composed of representatives from the industries and
universities that are working in this area and representatives from
the utilities or other potential users.
Dr. Morse: I might also add that the problem of universities responding
to RFP's, if they have a 2-week response time or a very short response
time, is one that we have discussed and considered. We intend to main-
tain a strong university involvement as far along the program as it is
appropriate to. So we're concerned with maintaining the involvement
of the universities and industry. We would like to see a growing
industrial involvement. We would like very much to have them pick
249
up the ball and make something of it and the sooner the better.
Q: Is the planning information that you included here going to be in-

cluded in the proceedings of the workshop?
Dr. Morse: I would think a summary of what we said will be.
COMMENT: May I point out that we received recently from NSF two RTF's,
both two months after the deadline for response.
Dr. Morse: Right, that's the gentleman from Alaska.
COMMENT: Airmail takes one day.

Q: What do:you see, if anything, as the place in this program for small-
scale developments? In other words, on demonstrations and perhaps
the type of thing I was talking about the other day—a demonstration
of the home heating plant based on wind power.
Dr. Morse; I think it's in there, and I think that shows the power of
these workshops. We have been focused pretty heavily on a way to
make major impact using wind supplied energy in a major way. It's
become apparent from this meeting and other discussions that perhaps
one way to attack the problem is through the small size user located
type of a system. We are going to give small users serious consider-
ation in our planning, but youTre right, it was not a major part of
our thinking.
Mr. Thomas: In all fairness, it really wasn't left out completely. The
original systems design studies have been encompassed in the range of
50 kilowatts on up. And maybe we have to go down a little further
than 50 kilowatts, but that was the size that we were looking at.
There would be application studies in there. The whole point is to

identify those applications that can do the job. If it turns out to
be 50-kilowatt machines and you have to put up thousands and thousands
of them to do the job, that would very much influence the program.
But at the same time you wouldn't go just that way. You'd be looking
at the large ones, too.
Q: Do you expect there will be anything in your RFP's on this, or will

this be handled on the basis of unsolicited proposal type of thing?
Mr. Thomas: My own feeling is that that would be part of the RFP route.
Dr. Morse; I would agree with that.
Q: Would you clarify one point on your charts — the difference between
energy systems, energy resources, and solar energy?
Dr. Morse: Solar energy is a resource, and you need systems to use it.
NSF had a program while solar was in its infancy, in which we were
looking at energy systems of the United States — for example, a re-
port by Dr. Szego's company, several volumes on energy systems.
We were looking at geothermal, at coal gasification, at a variety of
2.50
other problems1. Solar energy grew to the point where it has .now been
singled out as a separate program area. This doesn't mean that it's
distinctly different from the resources systems.
Q: Is the $300,000 being granted to Montana State and Oklahoma State a

total figure for the both?
Dr. Morse: No, for the three — for NASA Lewis for this workshop, for
Montana, and for Oklahoma State. I'm sorry, that comes to $200,000
not $300,000.
Q: Are these FY 1973 funds?

Dr. Morse: Yes.
Q: Is it illegal for a privately financed corporation to earn a profit

on proposals that are submitted in response to RFP from NASA or NSF
for this wind generator?
Dr. Morse: I'm not a lawyer. It's my understanding that that is legal.
COMMENT: In response to solicited proposals, I believe all the things

that normally have been going on, such as the fee, are perfectly legal.
It's only in response to an unsolicited proposal submitted that we must
have some kind of a cost-sharing as there is no fee allowed.
Mr. Thomas: We are going down that route right now with some of our
RFP's, and these are either going to be cost plus fixed fee or just
a fixed price contract with a few in there. That's the way NASA
does business. Now, on the unsolicited proposals, I'm not quite
that sure, but even there there are some ramifications. I know that
some contracts let for unsolicited proposals have included a fee.
Q: Will NASA or NSF be handling the RFP's?

Dr. Morse: We haven't decided, but, since NASA is going to implement
this program, it seems reasonable that they will issue RFP's, evaluate
them, award the contract, monitor it, and follow it up. That doesn't
mean that NSF will not do the same. There is still a difference
between the total NSF program and the total NASA program.
251
Figure 1
FIVE-YEAR WIND ENERGY PROGRAM
Overall Program
Study, Design, Build and

Test Wind Energy Conversion
Systems without Storage
Gather and Assess Wind Data
Study, Design, Build and Test Design of Construct and

Energy Storage Systems Demonstrations Test Demonstration
for Selected Systems
Applications
Identify and Study Suitable
Applications for Wind Energy
Prepare Facilities and Deter,

mine Requirements for Major
Facilities
Phase 0 Studies
o Identify Applications for Large Impact
o Social and Environmental Impact
o Study Potential for Capital Cost Reductions
o Study Operation and Maintenance Requirements
o Economics of Wind Power
Figure 2
WIND ENERGY CONVERSION SYSTEMS
Action Diagram
Build and Test

Prototype Components
and Subsystems
Systems Design
Studies of Wind
Conversion Sys-
tems Without
Storage
Build and;Test
Advanced
Concepts
NASA-LeRC Design Build and Field Operate Tested Take Over of

of 100 kW Base- Test in Cleve- System on Operational
line System land Culebra System by
Puerto Rico
252
Figure 3
METEOROLOGICAL STUDIES
Select Sites for Prototypes J*—
Determine Nation's Wind-

Energy Potential
Design of
Assess Existing Determine Favorable Wind Demonstrations
Wind Data Sites for Selected
Applications
Develop Analytical Techniques
for Wind Prospecting
Make Additional Wind

Measurements, Cleveland
Test Site and Culebra
Figure 4-
FIVE-YEAR WIND ENERGY PROGRAM
Energy Storage Systems
Study and Design of Build and Test Design of Demonstra-

Comparative Cpl pni-ori Selected Systems tion Systems for
Evaluation of Systems Selected Applications
Energy Storage
Systems
\ S
/
253
ATTENDEES
Allison, H, Jack Budgen, Harry

School of Electrical Engineering Brace Research Institute
Oklahoma State University 72 Broadview Avenue
Stillwater, Oklahoma 74074 Pointe Claire, Quebec
Canada
Anderson, Bruce
Dubin-Mindell-Bloome, P.E. Carl, W.
42 West 39th Street Grumman Aerospace Corporation
New York, New York 10018 Bethpage, New York 11714
Barnes, Waldemar F. Chang, Howard

Environmental Quality Board California State University at
Box 11448 Santurce San Diego
Puerto Rico 00910 College Avenue
San Diego, California 92115
Barnhart, Earle
New Alchemy Institute Clews, Henry
Box 432 Solar Wind Company
Woods Hole, Massachusetts 02543 RFD #2
East Holden, Maine 04429
Beard, Norris
Intertech Corporation Cohen, Robert
Warrenton, Virginia 22186 NOAA Environmental Research
Laboratories
Beattie, Donald A. Boulder, Colorado 80302
1800 G Street N.W. Curran, Marcia
Washington, D. C. 20550 Office of Senator Metcalf
Old Senate Building
Seller, William Washington, D. C. 20510
Environmental Protection Agency
c/o Secretary of Natural Resources Dodge, Robert
Box 5887 Pennwalt Corporation
Puerta de Tiera, San Juan, 205 Hutcheson Street
Puerto Rico 00906 Houston, Texas 77003
Bergey, Karl H. Douglas, L. L.

865 ASP, Rm. 200 Boeing Vertol Company
University of Oklahoma P. 0. Box 16848
Norman, Oklahoma 73069 Philadelphia, Pennsylvania 19142
Blieden, H. R. Dubin, Fred S.
National Science Foundation Dubin-Mindell-Bloome, P.E.
1800 G Street, N.W. 42 West 39th Street
Washington, D. C. 20550 New York, New York 10018
254
Ekstrom, Rurik Hughes, William L.
Pneumatic Structures Institute School of Electrical Engineering
Antioch Pneumatic Campus Oklahoma State University
10251 Barcan Circle Stillwater, Oklahoma 74074
Columbia, Maryland 21044
Hutter, Ulrich
Eldridge, Frank R. University of Stuttgard
The Mitre Corporation Pfaffenwaldring 31
Westgate Research Park Hochschulbereich Vaihingen
McLean, Virginia 22101 7 Stuttgard, West Germany
Fields, Ray Jacobs, M. L.

National Science Foundation Jacobs Wind Electric Co., Inc.
1800 G Street, N.W. Route 11, Box 732
Washington, D. C. 20550 Fort Myers, Florida 33905
Friend, Gilbert Jessop, Brian R.

Community Technology Rural Electrification Administration
1717 18th Street, N.W. U.S. Department of Agriculture
Washington, D. C. 20009 Washington, D. C. 20250
Garr, Douglas Katsanis, Theodore

Popular Science Publishing Company NASA Lewis Research Center
355 Lexington Avenue 21000 Brookpark Road
New York, New York 11717 Cleveland, Ohio 44135
Green, Richard J. Lawand, T. A.

National Science Foundation Brace Research Institute
1800 G Street, N.W. MacDonald College
Washington, D. C. 20550 Box 900
Montreal, Quebec
Hausz, Walter Canada
816 State Street Lefever, H. R. r~
Santa Barbara, California 93110 Free Lance Experimenter
Route 1
Heronemus, William E. Spring Grove, Pennsylvania 17362
University of Massachusetts
Amherst, Massachusetts 01002 Livingston, Robert
Energy Resources Report
Herwig, Lloyd 0. P. 0. Box 1067
National Science Foundation Blair Station, S.S., Maryland 20910
1800 G Street, N.W.
Washington, D. C. 20550 Lines, Charles W.
Hewson, E. Wendell 825 Capitol Street
Oregon State University Washington, D. C. 20426
Corvallis, Oregon 97330
255
Lissaman, Peter B. S. Nixon, W. Barry
AeroVironment, Inc. Forrestal Flight Research Laboratory
660 South Arroyo Parkway Princeton University
Pasadena, California 99103 Princeton, New Jersey 08540
Loewentha1, S tua rt Noel, John M.

Army Air Mobility Research and Aerowatt Corporation
Development Laboratory 37 Rue Chanzy
21000 Brookpark Road Paris, France
Cleveland, Ohio 44135
Ormiston, Robert A.
Loftness, Robert L. NASA Ames Research Center
Electric Power Research Institute Moffett Field, California 94035
1140 Connecticut Avenue #1010
Washington, D. C. 20036 Oman, Richard A.
Grumman Aerospace
Lutzy, Everett Bethpage, New York 11714
Town of Hull Municipal Light Plant
15 Electric Avenue Powe, Ra Iph
Hull, Massachusetts 02045 Department of Mechanical Engineering
Montana State University
Madey, R. Bozeman, Montana 59715
Grumman Aerospace
Bethpage, New York 11714 Puthoff, Richard L.
NASA Lewis Research Center
Meyer, Hans 21000 Brookpark Road
Windworks Cleveland, Ohio 44135
Box 329, Route 3
Mukwonago, Wisconsin 53149 Ramakumar, R. G.
School of Electrical Engineering
Mockovciak, J., Jr. Oklahoma State University
Grumman Aerospace Stillwater, Oklahoma 74074
Bethpage, New York 11714
Rabenhorst, David W.
Morse, FredericK H. Applied Physics Laboratory
National Science Foundation Johns Hopkins University
1800 G Street, N.W. 8621 Georgia Avenue
Washington, D. C. 20550 Silver Springs, Maryland 20910
Morris, Charles E. K., Jr. Reitan, D. K.

NASA Langley Research Center Department of Electrical Engineering
Hampton, Virginia 23365 University of Wisconsin
Madison, Wisconsin 53706
Nelson, Vaughn
Department of Physics Rinard, George
West Texas State University University of Denver
Box 248 498 South High Street
Canyon, Texas 79016 Denver, Colorado 80209
256
Rowe, T. A. Smith, Beauchamp E.
Science Council of Canada S. Morgan Smith Company
3100 Carling Avenue P. 0. Box 2304
Ottawa, Ontario York, Pennsylvania 17405
Canada
Stodhart, A. H.
Robertson, Lawrence M. Electrical Research Associates
The Western Electric Industry Cleeve Road
320 Ash Street Surrey, England
Denver, Colorado 80220
Swann, Mark
Rom, Frank E. RD #1
171 Balmar Boulevard New Park, Pennsylvania 17352
Avon Lake, Ohio 44012
Sweeney, Thomas E.
Rotty, Ralph M. Princeton University
NOM Princeton, New Jersey 08540
13817 Arctic Avenue
Rockville, Indiana 47872 Szego, G. C.
Intertech Corporation
Savino, Joseph M. Box 340
NASA Lewis Research Center Warrenton, Virginia 22186
21000 Brookpark Road
Cleveland, Ohio 44135 Thomas, Ronald L.
NASA Lewis Research Center
Schwartz, Harvey J. 21000 Brookpark Road
NASA Lewis Research Center Cleveland, Ohio 44135
21000 Brookpark Road
Cleveland, Ohio 44135 Titterington, William C.
Sefic, Walter J. 930 Western Avenue
NASA Flight Research Center Lynn, Massachusetts 01905
P. 0. Box 273
Edwards, California 93523 Tompkin, Joseph
Consulting Engineer
Shepherd, George R. 1324 Jordan Drive, South
Mechanical Technology, Inc. Salem, Oregon 97302
8106 Thoreau Drive
Bethesda, Maryland 20034 Torrey, Volt a (KSP)
National Aeronautics and Space
Sherman, Mark Administration
New Alchemy Institute Washington, D. C. 20546
Box 432
Woods Hole, Massachusetts 02543 Wentink, Tunis, Jr.
University of Alaska
Simmons, Henry T. Fairbanks, Alaska 99701
Newsweek Magazine
1750 Penn Avenue, N.W. Wharton, James
Washington, D. C. 20006 Tillamook Public Utilities Department
906 Main Street
Tillamook, Oregon 97141
257
Wiesner, Wayne
Boeing Vertol Company
P. 0. Box 1681*8
Philadelphia, Pennsylvania 19i42
Vance, W. S.
Science Applications, Inc.
1200 Prospect Avenue
La Jolla, California 92037
VanSant, James H.
Institut de Reeherche Del'Hydro-
Quebec
1000 Montee St. Julie
Varennes, Quebec
Canada
Wilcox, Carl
Allis-ChaImers
Box 712
York, Pennsylvania 174-05
Wilson, Robert
Department of Mechanical Engineering
Corvallis, Oregon 97330
258 £. U. S. GOVERNMENT PRINTING OFFICE: 197l*-757-678/5lt08 Region Ho. 5-II

Wind Energy Conversion Systems PDF

Uploaded by

Copyright:

Available Formats

Wind Energy Conversion Systems PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Wind Energy Conversion Systems PDF

Uploaded by

Copyright:

Available Formats

https://ntrs.nasa.gov/search.jsp?

Edited by Joseph M. Savino

In recent years there has been a growing awareness in some quarters

Each participant making an oral presentation was requested to sub-

After each oral presentation, a question-and-answer period was

The workshop was structured to have technical sessions exclusively

The workshop was proclaimed to be a success almost unanimously by

INTRODUCTORY REMARKS, Joseph M. Savino . 1

OPENING ADDRESS, Alfred J. Eggers, Jr. . 2

WELCOMING ADDRESS, Seymour C. Himmel 4

SESSION I - IMPORTANT FAST DEVELOPMENTS

SMITH-PUTNAM WIND TURBINE EXPERIMENT, Beauchamp E. Smith 5

PERCY THOMAS WIND GENERATOR DESIGNS, Charles W. Lines 11

PAST DEVELOPMENTS OF LARGE WIND GENERATORS IN EUROPE,

INTRODUCTION TO VOIGT1 S WIND POWER PLANT, Joseph Tompkin . .<- 23

Guest Speaker: WHERE THERE IS A WIND, THERE IS A WAY,

SURFACE WIND CHARACTERISTICS OF SOME ALEUTIAN ISLANDS,

WIND POWER RESEARCH AT OREGON STATE UNIVERSITY, E. Wendell Hewson . . 53

WIND DATA FOR WIND DRIVEN PLANT, Arthur H. Stodhart 62

SESSION III - ROTOR CHARACTERISTICS

AN INTRODUCTION TO THE PRINCETON SAILWING WINDMILL,

THE USE OF PAPER HONEYCOMB FOR PROTOTYPE BLADE CONSTRUCTION FOR

ROTOR DYNAMIC CONSIDERATIONS FOR LARGE WIND POWER GENERATOR

THE EFFECT OF AERODYNAMIC PARAMETERS ON POWER OUTPUT OF WINDMILLS,,

VERTICAL AXIS WIND ROTORS - STATUS AND POTENTIAL, W. Vance . . . . . . 96

ADVANTAGES OF THE DIFFUSER-AUGMENTED WIND TURBINE,

BUCKET ROTOR WIND-DRIVEN GENERATOR,

SESSION IV - ENERGY CONVERSION SYSTEMS

WIND-POWERED ASYNCHRONOUS AC/DC/AC CONVERTER SYSTEM, D. K. Reitan . . 109

AN ELECTRICAL GENERATOR WITH A VARIABLE SPEED INPUT - CONSTANT

VOIGHT VARIABLE SPEED DRIVE, Joseph Tompkin 121

SESSION V - ENERGY STORAGE

ENERGY STORAGE USING HIGH-PRESSURE ELECTROLYSIS AND METHODS FOR

USE OF HYDROGEN AND HYDROGEN-RICH COMPONENTS AS A MEANS OF STORING

STATUS AND APPLICABILITY OF SOLID POLYMER ELECTROLYTE TECHNOLOGY TO

SUPERFLYWHEEL ENERGY STORAGE SYSTEM, David W. Rabenhorst 137

BATTERIES FOR STORAGE OF WIND-GENERATED ENERGY, Harvey J. Schwartz . . 146

ENERGY STORAGE BY COMPRESSED AIR, George C. Szego 152

EXPERIENCE WITH JACOBS WIND-DRIVEN ELECTRIC GENERATING

REVIEW OF THE WINDPOWER ACTIVITIES AT THE BRACE RESEARCH INSTITUTE,

ECONOMIC CONSIDERATIONS OF UTILIZING SMALL WIND GENERATORS,

WIND UTILIZATION IN REMOTE REGIONS - AN ECONOMIC STUDY,

.SESSION VII - WIND POWER SYSTEMS FOR LARGE-SCALE APPLICATIONS

TECHNICAL FEASIBILITY STUDY FOR THE DEVELOPMENT OF A LARGE CAPACITY

THE OREGON STATE UNIVERSITY WIND STUDIES, Robert E. Wilson 180

FRENCH WIND GENERATOR SYSTEMS, John M. Noel 186

SESSION VIII - TOWER STRUCTURES

A COMMENT ON TOWERS FOR WINDMILLS, H. P. Budgen 204

SOME EXTEMPORANEOUS COMMENTS ON OUR EXPERIENCES WITH TOWERS FOR

WIND MACHINES, P. B. S. Lissaman 208

WIND CHARACTERISTICS AND SUING 209

ROTOR CHARACTERISTICS 213

ENERGY STORAGE AND ENERGY CONVERSION SYSTEMS 218

PANEL DISCUSSION 225

NSF PRESENTATION, Frederick H. Morse 240

NASA PRESENTATION, Ronald L. Thomas 244

Lewis Research Center

It certainly is a pleasure to welcome all of you to this first

We are particularly happy that the National Aeronautics and Space