LETTER TO MR. EMILIO Q. DADDARIO FROM GEORGE L. CARY

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP77M00144R000600080014-5
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 19
• Document Creation Date: December 20, 2016
• Document Release Date: July 19, 2005
• Sequence Number: 14
• Publication Date: July 9, 1975
• Content Type: LETTER

Approved For Release 2006/10/18: CIA-RDP77M00144R0006000 q 47_0g01/a
CENTRAL INTELLIGENCE AGENCY _
W ASH I NGTON, D.C. 20505
c4L4 "- O (FT-
8 JUL AS15
Mr. Emilio Q. Daddario, Director
Office of Technology Assessment
Washington, D. C. 20510
Dear Mr Daddario:
As a result of your April 11 request to the Director that Dr. Ronal
Larson of OTA meet with CIA employees working in the solar energy field,
Dr. Larson did meet on two occasions afor nd Soviet solar employees.
He asked for information o
that our personnel informally review a draft OTA report entitled "Solar
On-Site Electricity" from a forei o viewpoint. I am forwarding
with this letter a memorandum o Soviet solar projects and
an article entitled "Solar Photovoltaic ower eneration." Also enclosed
is the copy of the draft report. Several analysts have read it and have
made marginal notations. T-n addition, we asked fossil fuel and solar energy
specialists on the staff of the Agency's
energy consultants, to comment on some o the com.parau e engineering
economic data presented in the report; their observations have been included
at appropriate points in the text. Inasmuch as the Agency's main focus
is on the foreign technology aspects of the issues involved, our comments
are limited to questions of technical fact and analytical considerations,
except for the aforementioned economic data.
We are happy to be of service to your office in these areas and it
is hoped that if you have any further similar requests you will be in touch
with my office.
Sincerely,
SIGNF~!. r
George L. Cary.
Legislative Counsel
Enclosures
Distribution:
Orig - Addressee
`
1
- DDS &T
$- OLC Subject
1
OSI
1 - OLC Chrono q
1
D/OER
OLC: DFM: cg (2Jul75)
,0 UTlo1
a M
Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
 STAT Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
 Approved For Rele0
Rk'
SUBJECT: Foreign Reference Material Relating to Solar
On-Site Electricity Generation
A resurgence of interest in solar energy utilization
has now taken place within the Soviet government. In
January 1974, as part of the May 1972 US-USSR Agreement on
Cooperation in the field of Science and Technology, the two
nations signed a protocol for Cooperation in the General
Problems of Utilizing Solar Energy. Dr. Richard J. Green
of ERDA is the US coordinator for joint solar projects. One
of the three problem areas deemed appropriate for joint
research is solar power stations, designated as Program I
under the joint agreement.
The Soviets have reinstituted their own effort with plans
for the construction of a solar power station with an output
of 1-2.5 MWe. The former design (reflectors mounted. on
moving flatcars circling the base of a tower supporting a
movable receiver-boiler) has been modified to reduce
construction costs. The new design involves reflectors
located in fixed positions around the base of a tower upon
which a stationary tubular boiler is mounted. The
heliostat system is made up of 1600 3-by-Sm reflectors;
the mirrors can be oriented on two axes and track via a
photocell-activated servo system. The maximum radius from the
tower to the outer circle of reflectors is 200m. The mirrors
reflect the sun's radiation onto a boiler mounted on a 40m
high tower. The boiler is a vertical cylinder, 7m high and
9m in diameter wrapped in a helical array of tubes. The
system is designed to produce 11 tons/hr of super-heated
steam at 400-500 deg. C and 35 atm.
The Soviets plan to generate 1.2 MWe with a non-condensing
turbine or 2.5 MWe with a condensing system. No storage
(either thermal or electric) is involved and the plant is
expected to operate for 2000 hrs in the year. They anticipate
that the overall conversion efficiency of sunlight to electri-
city will be about 17% (based on the following stated
assumptions: mirror reflectivity, 0.78; interception of
radiation by boiler, 0.9; optical absorptivity of boiler
tubes, 0.9; thermal efficiency of boiler, 0.85; the efficiency
of conversion of solar radiation to steam is about 55% from
which, assuming turbine efficiencies of about 30%, the overall
plant efficiency comes out to approximately 17%).
Approved For Release 2001 6 8 1?710600080014-5
 Approved For RelAaMc8 LC - L 0144R000600080014-5
SUBJECT: Foreign Reference Material Relating to Solar
On-Site Electricity Generation
Design studies are virtually complete and site evalu-
ations for future construction have been performed. In
addition, one heliostat assembly has been built and field-
tested. Apparently, the Soviets intend to begin construction
of a power plant with the above specifications in the near
future. Individuals associated with the project have
indicated that they believe cost to be the determining
factor in the success of the system. Estimates for capital
investment of 1000-2000 rubles/kW have been noted, compared,
with 150-180 rubles/kW for fossil fueled plants. (It should
be mentioned here that this solar plant is tiny in
comparison with normal Soviet power stations which tend to
be of the order of 300-600 MWe and larger in capacity. The
very large capital cost factor must therefore be viewed
as quite unrepresentative of a large commercial installation
to which the fossil fuel cost estimate probably does
correspond.)
Approved For Release 2000a1$ 'IP71J10?t}0600080014-5
 Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
"-DIT.gvt, 0 trje 1::ji1tftibb r7ftttG~
OFF ICE OF T[ CHNOLCG3Y ASSESSMENT
SOLAR Oil--SITE ELECTRICITY
Draft Final Report
April 11, 1975
Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
 Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
utilize
ompre-
uld be
=t. The
would
around
to the
sewage
_ rming
waste
-nssible
s"; by
.pasting
isively
profit
11 pur-
more
iple to
-t only
=ed.
during
ig but
-pecial
such
-oduc-
iority
'three
Iurces
)f the
_tstes"
icting
'here-
_ist be
-)erty.
parti-
- area
nj ury
~blem
bility
I the
--ople,
Solar Photovoltaic Power
Generation
Hajime Maeda
Toshiba Research and Development Center,
Tokyo Shibaura Electric Co., Ltd.
The report by The Club of Rome several years ago and the
oil-shock originated by the Middle East quarrel which again
flared up the year before last disclosed to the world the magni-
tude of the energy crisis. Japan is most seriously affected by
the crisis because its archipelago covers only a small area
(377,000km2 ), yet it is densely populated (109 million), while
natural resources, including energy resources, are scarce.
Aiming at clean (non-polluting) energy resources and energy
independence, Japan's "Sunshine Project" began on July 1st
last year under the guidance and sponsorship of the Ministry
of International Trade and Industry (MITI).
The project intends to develop four types of energy resources:
solar energy, geothermal energy, synthetic natural gas derived
from coal, and hydrogen energy. The solar energy plan consists
of four main items:
(1) utilization systems,
Conceptual Diagram of Solar Energy Utilization System
(2) thermal powr generation,
(3) photovoltaic power generation
(4) heating-and-cooling and hot water supply for dwellings.
As for photovoltaic energy conversion, silicon solar cells have
been used in special fields, for example, aboard spacecraft, for
communications power sources on remote islands, and atop
mountains, and for operating navigation lights on buoys and
automated lighthouses. As general power sources, however,
conventional solar cells are too expensive at their present stage
of development and there is thus the need to achieve a tech-
nological breakthrough, which will lead to their large scale
cost-down. Toshiba R & D Center is meeting the challenge of
developing commercially feasible solar cells, since the company,
as a comprehensive maker of electric products, is producing
electric power generation machinery as well as semiconductor
and electronic devices- Further, the company is one of the five
major makers of silicon single crystals.
The following is a brief review of the state-of-the-art of the.
R & D of solar photovoltaic conversion-
No.2 I TECHNOCRAT v%NWed For Release 2006/10/18: CIA-R DP77M00144R000600080014-5 19
 Human Being and?f FgyiplqJtof I iqse 2006/10/18: CIA-RDP77MOO144R000600080014-5
Non-reproducible resources
Reproducible resources
Locally important resources
Trivial as energy resources
Nuclear fusion
tHeavy hydrogen as fuel)
Nuclear fusion
(Lithium 6 as fuel)
Solar energy[
Tidal energy
Nuclear fission
(Breeder reactor)
Nuclear fission (Converter)
Geothermal energy
Hydraulic energyl
Animal energy
Wood ? Vegetable fuel
106 104 102 10 Year 102 104
2. General Features of Solar Photovoltaic Power Gener-
ation
2-1. Properties of Solar Energy
As prospective energy resources, there are atomic energy
nuclear fusion energy, solar energy, geothermal energy, synthetic
natural gas, hydrogen and global energies such as ocean energy,
wind energy and hydraulic energy. Some of the above are
included as a targets of the Sunshine Project. Among these,
solar energy has some characteristic features which distinguish
it from the others. These characteristic features, which neces-
sitate for its effective utilization a thorough understanding
thereof and a development of suitable uitlization techniques, are
as follows.
Solar energy incessantly falls on the earth in the form of solar
radiation, and is therefore, virtually endless. Its rate of supply
is immense (173 x IO6GW or 13 x 1020Kcal per year) or
equivalent to about 140 trillion K! of oil per year. However,
it also has some "adverse" aspects such as:
(1) It is dispersed in space, resulting in a very low power
density (-140naW/cm2 outside the atmosphere, and, at the
maximum, -100mW/cm2 on the earth's surface).
(2) On the earth's surface, the radiation power density varies
with time, between day and night, and through the year. The
variation is especially large in Japan on account of its location
around latitude 35?N. Thus, if the seasonal variation as well as
Present
the day-period variation of radiation intensity and the incident
angle at which it falls are taken into account, together with the
effects of bad weather, the average radiation power density
must be estimated at one tenth the above mentioned maximum
value.
(3) Solar radiation has a wide spectral distribution ranging in
wavelength from infrared to ultraviolet (Fig. 1), therefore, a
complete utilization of its total energy is impossible. Different
methods are effective in utilizing its different. regions of wave-
length. The heat collector, for instance, plays an important role
in the utilization of solar energy in the region of heat radiation
including infrared, while the silicon solar cell is effective in
utilizing the region of visible light.
Fossil fuel represents an accumulation of this solar radiation
over an extremely long period of time, while hydraulic and
wind energy represent its condensation in space. One liter of{
'
crude oil contains, on the average, 9,400Kcal of energy; this
corresponds to an accumulation of solar energy falling on 1ni
through a 4-5 day period with an efficiency of 100%.
2-2. Characteristics of Photovoltaic Power Generation
For a concentrated utilization of energy, as is general in
modern civilization, it is necessary to collect solar energy in
some effective way as mentioned above. In order to gaiin an
output power of ten thousand kW on the average on the earth's,
surface, we need an area of about one km2 if we assume ac
Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
 Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
Fig. 1. Energy Spectrum of the Solar
Radiation
(in Wave Lengthsli
Black Body Radiation:6000?K
Solar Radiation outside the Atmosphere
Solar Radiation at the Sea Level
Wave Length (Fm)
cident
th the
,ensity
.imam
ing in
ore, a
fcrent
wave-
It role
iation
ive in
rat in
gy in
in an
. rth's
,me a
conversion efficiency of 10`%.
An energy form which is suitable for utilization as a general
rat energy source is electricity. Processes for converting solar
tnergy into electricity are thermal power generation and photo-
ivltaic power generation. In the former, the solar energy,
crught by a collector system, convert water vapor or other gas
rigors, through a heat exchanger system, into a high tempera-
ture steam or gas, which is then supplied to a system resembling
an ordinary thermal power generating plant. Thus, it is subject
to scale merit, being rather inefficient as a small capacity gener-
sting unit while suitable as a large capacity centralized power
generating system. When it is utilized as a centralized power
generating system, the problem of the time variation of solar
udiation becomes crucial, and the development of energy
storing techniques, such as heat storage, storage batteries,
pumped storage power generation, etc., becomes important.
In Japan, with its small, narrow land area, construction of this
systertl within the country accompanies locational disadvantages.
In the "photovoltaic" system, on the other hand, powre is
actually being generated at every point on the surface of the
solar cell fabricated with a certain kind of semiconductor
material (as will be described later), and is therefore not subject
to the scale merit as is the thermal generation system. Since
semiconductor material, as the working material, is spread over
a large area, its cost and packaging technique are problematic
and require thorough probing. If these problems are solved, its
usefulness as a general use power source can be expected at all
ttvels, from small scale use for homes, to medium scale use for
i ldings, schools, offices and small factories, to large scale use
Asa Centralized power generating system.
A merit of the photovoltaic system is that it has, as its solar
cell material, silicon which has the following superior features.
First, its characteristic energy, or "band-gap", corresponds to
a wavelength of 1.2p, which happens to he in good coincidence
with the useable solar radiation spectrum range. Secondly, it
is the element most abundantly found on earth next to oxygen.
And thirdly, the technology related to silicon itself has attained
remarkable progress during the past quarter century in the
transistor industry field.
The silicon solar cell, in spite of a lapse of twenty years since
its invention in 1954, has been applied as a power source only
to a limited and specialized extent. The few installations per
year in Japan generate only several KW. The greatest reason
for its limited use as a. power source is its prohibitive cost. At
the present technological level,, the cost for IKWof maximum
output power is approximately Y 15 million (-S45,000), which
is several hundred times the cost for conventional power genera-
tion such as thermal power generation.
Thus, an essential developmental problem that bars the way
to the practical utilization of the photovoltaic system as a new
energy source is the development of a technology that leads
to a cost reduction of 100 times or more compared to its
present level and this cannot be attained through such "im-
provements" as the development of a mass-production tech-
nique, but rather, positive "technical innovations" are require
for its attainment.
There are proposed, as targets of energy development pro-
jects in Japan and the U.S., two, forms. of photovoltaic power
generation systems, the space station and the terrestrial station
(Figs.2, 3). In the former concept, which. is chiefly proposed by
ADL of the U.S. is a station with two 2km square solar cell
arrays which will be sent into orbit above the earth and the
power generated by it will be transmitted to earth in the forth
of microwaves. With this system, a constant and rnaximunm
Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
 radiation can b I W04 ~jQtCC~v ~l~t~ t WAV arpt~As`-RI t7a7iMpQi s4 ~77tt~Qc lSifoltS~y4 gether with-the prospect
so as to always to face the sun (the solar radiation source).
However, in this case, some new technical problems will havef
to be solved, (other than reducing the price of the solar cells),
such as the launching of such solar satellites, making the station
light in weight, effectively controlling the station, transmitting
and receiving high power microwaves, etc. The latter concept,
advocated by NSF (the National Science Foundation) of the
U.S., includes the development of a power station with solar
cell arrays spread over a land area of several km2 and a hydrogen
plant based on the electrolysis of water. The combination with
hydrogen gives a means, through the storage of energy, for
solving the problem of time variation of solar radiation as well
as a means for an efficient transference of energy to the users.
As was described above, photovoltaic generation does not
have the scale merit problem and is thus suited for both small
and large capacity generation systems. In consideration of the
special circumstances of Japan, such as its narrow land area and
Fig. 2. Electric Power Generation in Space
YO
o Microwave
Beam
c /f Transmitter Antenna
Fig. 3.
Terrestrial Solar
Photovoltaic Power
Station
(after F. R. Eldridge,
the Mitre Corp.)
Satellite Orbit
Solar Cell Array
of the development of an energy storing technique, medium and
small capacity power sources using the photovoltaic syste:rrr'
gains importance as target for development. Taking these into
account, the author and co-researchers propose the development
of a small capacity photovoltaic power source unit and magnifi-
cation of its capacity through a synthesis of the small capacity
power source unit, The target rated performances of the power
source unit are as follows:
Maximum output
1.2KW
Output voltage
IOOV (D.C.)
Solar cell surface area :
I Om2
Conversion efficiency :
12%
An example of the application of this unit in the smallest
capacity system, the case of utilization as a domestic power
source, will now be considered. As shown in Fig.4, the system
will be connected to the utility line together with an AC-DC
inverter and a controller. When the generation of energy by
the solar photovoltaic system is great enough, AC power is
supplied to each load from the solar cell array through the
AC-DC inverter. But when generation by the photovoltaic cell
is insufficient, such as during the night or during periods of bad
weather, the deficit power is supplied from the external utility
line. When generation by the solar cell array exceeds the load
demand, the excess power flows out into the external utility'
line and is supplied to other users of power in the neighborhood
or to a central pumped-storage power station. The inverter is
excited by the external power in order to maintain the quality
of power, and a controller is provided for optimum operation
of the solar cell array. Also, for an evaluation of the quantity
of power consumed at the home, the watt-hour meter is pro-
vided with reverse rotation when power flows out.
The advantages of the above power generation unit and the
generation system are:
(1) Constant supply of power to loads and the maintenance
of conditions for optimum operation of the solar cells can. be
achieved without requiring battery units.
Electrolysis
Facility
Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
 Fig. 4.
Approved For Release 2006/10/18: RA-RDP77M00144 000600080014-foof-tole Solar
(2) Specific facilities for power transmission and distribution
are unnecessary since most of the generated electric power is
consumed on the spot. In other words, the system represents a
suitable method of utilizating solar energy, which is distributed
in low concentrations spatially.
(3) Though essentially an auxiliary power source, its parallel
connection with the utility line equivalently increases the
generating capacity of the central power stations. It is especial-
ly effective therefore, as a measure for power consumption
peaks in the summer season in Japan.
(4) The average output per month per unit is about 90KWH,
which means that it can furnish a considerable part of the
average power consumption of an ordinary household (i.e.,
-150KWH). It will play a very important role during the swn-
mer peak period as mentioned above.
(5) When larger power is required in buildings, factories,
schools, offices, etc., the unit can be effectively utilized by
simply connecting many of it in parallel.
(6) It requires no special geographical conditions nor is it a
public nuisance, as some other power generation units.
The author and co-researchers plan to obtain, by 1980,
a prospect for producing the above unit at a price of around
Y150,000 ($450), that is, one hundred times cheaper than
the present cost. This figure was set, by considering the needs
of average homes, so as to set a price of Y240,000 ($720) per
unit together with the inverter and the controller, a price similar
to the price of home cooler's. If a maintenance cost of Y9,000
(S27) per year and a lifetime of 20 years are assumed, the power
generation cost will be #20 (Q' 7)per KWH, and if this is reduced
to about one third this level by means of allight collecting tech-
nique and others, it can favorably compete with the power
generation cost of newly constructed thermal power plants,
i.e. Y8.90 per KWH. Of course, the author does not assume
the cost competition as the final, or end result of solar energy
Photovoltaic System
for Power Generation
use rather the problem of environmental pollution and the
imminent exhaustion of present conventional energy resources
gave rise to the desire to develop new energy technologies,. and
the cost itself can be considered fluid if considered in. a social
context.
3. Problems in the Way W Industrialization as a Gen-
eral Use Power Source
3-1. Structure of Silicon Solar Cells and Its Fabrication
Process
For the discussion of technical problems, a brief explanation
of an example of a conventional solar cell will be given. As
shown in Fig.S, we start with N-type silicon wafer, which is.
doped with a group-Vchemical element and in which electrons
(negative electric charge) carry electricity- Then we dope boron,
Anti - reflection
Coating
/P-type Diffusion
Layer
f P- N Junction
.-N -type Silicon
Single Crystal
Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
 tricity is carried by JWPrQfj,(4IFgt'efq - -
"holes
At the boundary of the P- and N-regions, called the P-N junc-
tion, an electric field is built-in, and when a radiation of a
wavelength less than I.2?, which corresponds to the band gap
of silicon, reaches the P-N junction, electrons and holes are
created and separated by the built-in electric field toward the
N-type and P-type regions, respectively, and thus electric power
is generated. Of course, metallic electrodes are necessary to
take out. the built up power.
Created electrons and holes vanish with certain probabilities
before they reach the electrodes. Therefore, the conversion
efficiency is chiefly determined by the average time lapse
before the charge carriers "vanish".This is called the "life time"
and is determined by the degree of purity and perfection of the
crystal, and by the necessary distance traversed by the carriers
(therefore by the cell construction design). The output voltage
originating from the built-in electric field is nearly 0.6V, if the
voltage drop in the cell is neglected. In order to reduce this
voltage drop it is desirable that the series resistance within the
cell. is small. In this point, certain kinds of Ill-V group com-
pounds, for example gallium arsenide, are superior to silicon.
Also, the coating which is to prevent loss by reflection of
the solar radiation on the cell surface is very important and
obviously cannot be dispensed with. As seen above, the fabrica-
tion process of a solar cell consists of single crystal wafer fabri-
cation, P-N junction formation, electrode metallization and
anti-reflection coating.
3-2. Discussions of the Production Cost and the Energy Con-
sumption in the Production
The toughest problem for the industrialization of the solar
photovoltaic power generation unit is the reduction of cost.
In Fig.6, a cost estimation is carried out for aunit which is
capable of generating 1KW during optimum conditions.
The estimation, on the basis of the present day technology,
gives a prohibitively high value of $66,700. Let's look into the
reason for this.
In the first stage of cell production, silica, the raw material
is processed, through reduction and purification, into pure
polycrystal silicon. Then, a single crystal seed is immersed in
the melt of the above polycrystal silicon and is pulled upwards
at a few nim per minute, rotating slowly, generating a single
crystal ingot of cylindrical shape with a diameter of several
centimeters. The fabrication process for polished silicon
wafers, which constitutes the starting point for the fabrication
of all sorts of silicon devices today, begins with the slicing of
the above cylindrical ingot, and follows with cutting, lapping
and polishing. The slicing process causes a substantial silicon
loss of more than 70%. There is at present, no way to reduce
this loss. In addition, the succeeding processes result in sub-
stantial labor costs and reduced yields.
Fabrication of solar cells, mentioned in the preceeding
section, is carried out by the batch process, which requires
many processing steps and cacti wafer or each chip results again
in substantial labor costs and very low yields.
It is clear that the price reduction envisaged for making the
solar cells commercially practical cannot be achieved by merely
increasing the manufacturing scale. We present here, as a future
technology, a new ribbon crystal growth technology combined
with a fully automated continuous production system for pro-
cessing the ribbon crystals so developed into solar cells (Fig.8).
An estimation is carried out on the basis of the ribbon crystal
Present Technology Future Technology
Raw material
Assembly,,, ,, Polycrystal
Fig. 7. Energy Balance of Solar Cells
Present Technology
Evaporation, , Metallizotion
10.975kWH/kW 1,246kWH/kW
Power Consumed in I kW Solar Cell Fabrication
Fig. 8. Automated Continuous Fabrication
System of Solar Cells
Ribbon Crystal Pulling Junction Formation Metallization and
Anti-retlection Coating
growth rate of 80cm2 /min. which is the author's first milestone
to be reached in next 3 years. The cell production system is
designed to be matched to this ribbon growth rate. The solar
cell cost then becomes $800/unit, which is about an 80-fold
reduction of the present cost.
The amount of electric energy which is required to produce
the solar cells is another very important factor for the examina-
tion of their feasibility, since it would be rather ridiculous if
is
all
VI
rnl
&i,
flz
p9
tl
Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
 -Fig. 9. Silicon Ribbon Crystql d For Release 2006/1Q/18 - CIA-RfP77MQQi44 0080014-5---
F-, Pull:
Capillary Die for Ribbon
Growth
EFG Method
this energy exceeds that which the solar cells can generate
during their lifetime. Shown in Fig.7 is a comparison between
the energies needed in producing a solar photovoltaic generation
unit with an output power of 1 KW at the peak output by
present day technology and a unit developed by the future
technology, envisioned above which utilizes the ribbon crystal.
In the former case, an electric energy of about 11,000KWII is
required and it takes more than 10 years to generate this
amount of energy with this unit. With the application of future
technology it will take only one or two years to recover the
energy consumed during fabrication.
3.3. Silicon Ribbon Crystal Growth
The foremost object of the development of this technology
is to present the crystal in a form which is convenient for the
automation of the fabrication process of the cell elements,
which, as clarified in the previous section, accounts for the
majority of the high cost in the production of the present-day
trilar cell fabrication process. Another is the contraction of the
production cost of the polished wafer itself. Further, the
presentation of wafers with larger face areas may be added as
third object.
The ribbon crystal growth methods may be broadly classifi-
ed into two groups: the dendrite method(s) in which the crystal
itself is given a property to grow in a ribbon shape, and the
capillary method(s) in which melt is drawn out from a ribbon
shaped die and is solidified.
If silicon melt is rapidly cooled to below the melting point
under certain conditions, a crystal in the form of a leaf of a
coniferous tree is seen to grow. That crystal is called a dendrite
ind includes twin-boundaries within itself. Its direction of
growth is thecrystallographic direction and its surface is
the 4 111 } plane. Its width never exceeds several mm. By con-
trolling the conditions of the dendrite growth such as the
+tdocity of growth and the temperature, two dendrites and,
between them, a film shaped crystal, thinner than the dendrites,
xnd having parallel mirror planes may he made to grow. This is
cilkd a web crystal. So far, web crystals with a 3cm width
love been obtained, and a growth rate of the order of 20cm/
Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
Web Method
(after D.L. Barrett et. at.)
min is considered possible.
Research on dendrite growth was initiated in 1960 at the
Westinghouse Electric Co. and a web crystal was developed in
1967. However, on account of limited needs from the field of
semiconductor devices at that time, its further development was
subsequently slowed down.. In Japan, research on dendrites
was made by Toshiba also around 1960. Advantage of the web
method is that ribbon crystals with a comparatively large width
and a good degree of crystal perfection together with good
crystal surfaces can be obtained at a high growth velocity,.
while its disadvantage is that it requires a high level of produc-
tion technique for controlling the temperature distribution,
pulling rate and mechanical vibration. Consequently, multiple
pulling from a single crucible is very difficult.
The capillary method is theoretically based on the same
idea as that of the manufacturing of sheet glass or polyethylene
sheet. Since Stepanov and others in Russia applied the method
to germanium, many improvements have been made, and the
one attracting greatest attention at present is the method
developed for alumina by l lHelie Jr. and others in Tyco Labor-
atories and named EPG (Edge-defined Film-fed Growth) meth-
od. In this method, if a die material which is wetted with
silicon melt is chosen, the melt is supplied stably to the upper-
edge of the die through capillary action (as shown in the left in
Fig.9), and at the same time, the outer contour of the upper-
edge of the die (not the inner contour) determines the shape
of the ribbon crystal. These circumstances are exactly the
same as those in a case reported by Stepanov and others except
that the relative configuration of the die and the ribbon crystal
is different between the vertical and horizontal configurations..
As for the application of the FIG method to silicon, Tyco
Laboratories have carried out intensive development with funds
from the NSF and have succeeded in growing a ribbon crystal
with a width of 2.5cm and a thickness of 200p and in the trial
manufacture of a solar cell with an efficiency of 10%. Various
other methods of forming the crystal such as drawing down-
ward or laterally, drawing by supporting the melt with a film,
etc., have been proposed and. tested.
At present, the EFG method is more popular. but there still
remain many problems which must. be solved prior to its suc-
 cessful developmefatp yey tlrgfii ilel i, cQs~ ly~o l d lA } iR77pl 4Q 14440qW001D QCl1{}s a production capacity
material has not yet been found: since the silicon melt has such
a strong reactivity that it reacts with almost any material, the
die therefore, can be easily corroded and the silicon contaminat-
ed. Another problem is that sufficiently flat crystal surfaces
have not yet been obtained. In the future, a method based on
improvements in these points, a "modified" capillary method,
so to speak, might conic into the spotlight as an important
candidate for industrial solar energy technology.
3-4. Silicon Thin Film and Compound Semiconductors
Solar radiation absorption by silicon crystals is 80% and
90% when the thickness of the crystal is I Op and I OOp respec-
tively. Therefore, if the price of the silicon material goes up
in the future and thus occupies a substantial portion of the cell
cost, thin films with a thickness of the order of lOp will be
economically more advantageous than thicker materials, al-
though the conversion efficiency might be smaller by 10% or so.
From the above viewpoint, an attempt is being made to fabri-
cate solar cells by generating a polycrystalline thin film on a
low-cost basis with the crystal grain size controlled so as to be
equal to, or larger than, the order of the film thickness and
forming a P-N junction or Shottky barrier on its surface. As a
method for thin film generation, such processes as vacuum de-
position, sputtering, iron-plating and chemical vapor deposition
are now being considered.
For attainment of the fabrication of polycrystal devices at an
output efficiency that can be of practical interest and with a
good reproducibility, studies of physical and electrical proper-
ties of the polycrystal itself, and especially those of the crystal
grain boundaries, will be very important. The major practical
problem needed to be solved will be the discovery of a cheap
yet high-quality base material.
Besides silicon, there are, for the solar cell material, various
compound semiconductors with various different features. It is
possible, for instance, to select a material with a band. gap better
fitted to solar radiation than silicon. Further, it is fascinating
to think of superposing two or more materials. Cadmium
sulphide is low in both material and fabrication cost. Gallium
arsenic (GaAs) permits the decreasing of series resistance to a
small value so its conversion efficiency would not drop marked
by even under a high concentration of light. It also has a large
output per unit weight. For each of these uses and possibilities,
matched with the best features of each, may well be imagined,
but none of these can compare with silicon in an all-around
evaluation as a general use power source at the present time.
4. Present Status of R&D Schedule
The main problem for the industrialization of the solar
photovoltaic generator is the crystal growth technique, but
besides this there also exist various other important problems
which need to be solved. They are: such technical problems as
the preparation of raw material, solar cell fabrication and the
assembling of solar cell arrays; such system problems as energy
storage, collection and pursuit of solar radiation, and combina-
tions with solar heat air-conditioning; and such social problems
as the environment, architecture engineering, legislation con-
cerning electric power supply, and its popularization policy.
As for preparation of the raw material, for instance, if 2% of
the estimated total quantity of Japanese domestic power gene-
ration in the year 2000 (1.Sx 10 12 KWH), is to be furnished by
silicon solar cells with a thickness of 100p, 105 tons of poly-
crystal silicon is required, and this in turn requires, assuming a
of 5,000 tons per year.or 20 times tine present capacity. On a,,
global scale, it can give rise to a new resource problem. Tlus~
constitutes a further motive for the development of the thin
film technique. Further, an integration is being considered
between the process of reduction and purification of the raw
material and the process of ribbon crystal fabrication.
4-1. Japan's Sunshine Project
Concerning solar photovoltaic power generation, the penul-
timate goal is the development of a high-performance, low-cost
system by around 1990. The target and content of the enforce-
ment plan from 1974 to 1980 states that "Studies will be made
on the technical feasibility of producing photovoltaic generator
systems with a substantial price reduction equal to, or less than,,
1/100 the present price if the system is fabricated by the present
technology. For this purpose, research and development will Ibe
made on a new fabricating process for semiconductor crystals,
on solar cell devices as well as on peripheral technology." The
program was started with the following division of work:
(a) Basic research of solar cell;
- - - - - Electrotechnical Laboratory,
- Silicon solar cell -
(b) Silicon ribbon crystal by vertical pulling;
- - - Tokyo Shibaura Electric Co., Ltd.
(c) Silicon ribbon crystal by horizontal pulling;
- - - - - Toyo Silicon Co., Ltd.
(d) Silicon thin film crystal of particle acceleration growth
(e)
(f)
(g)
type;
----- Nippon Electric Co., Ltd.
Silicon thin film crystal of particle non-acceleration
growth type;
-----Hitachi Ltd.
Compound semiconductor solar cell;
- Matsushita Electric Industrial Co., Ltd.
New type solar cell;
- - Sharp Corporation.
Besides these, research on solar cells composed of various kinds,
of compound semiconductors is to be undertaken at each of the
above organizations. Further, under the management of Japani
Electric Machine Industry Association, a research committee
undertaking studies on utilization systems is to be created for
the purpose of conducting surveys and research on other inr-
portant technical problems, problems concerning utilization
technology as well as social problems.
The total appropriation allotted to the Sunshine Project in
the fiscal 1974 budget amounted to Y2.4 billion ($7.2 million),
from which expenses for the consignment of research and
development to private sectors in the field of solar photovoltaic
systems amounts to some #200 million ($0.6 million).
Also at universities and other research institutions, independ-
ently of the Sunshine Project, basic research such as on the
development of solar cells utilizing new materials, solar cells
with new structures, such as hetero-junction, etc., are beginning
to be undertaken actively.
4-2. Present Status of R & D in the U.S.
Research and development. of solar cells in the U.S. have been
carried out by NASA and under its strong sponsorship, with the
aim of power supply for space satellites, the accumulation of
technology has been immense. In 1974, $1.5 million was spent
for solar cell development, with about forty specialists engaged
in research activities in this field at the Jet Propulsion Labo-
Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
 rator Pasadena, Calif.; kti i sti o Research Laboratory is seeking funds to reactivate
Y, EAb ?Y6AffcF Green e . 01' ~ 'i ~ i Of~ ~ a i $4 ~thf~d which was
Ohio; Slid the Goddard Spacecraft ('enter, , Gree e t, a t r a r r i
direction of development in the above activities (for space dropped about seven years ago.
stations) is somewhat different from that for terrestrial applica-
tion, and lays more stress on such things as the weight factor, or 4-3. Recent Events
output power density per kg, and protection against radio-
activity, though, of course, some considerations are given to On October 26, last year, the "Solar Energy Research,
the lowering of cost. Development and Demonstration Act of 1974 was proposed in
As for terrestrial application, many research activities at the U.S. This includes an additional appropriation of $I billion
universities and other institutions have joind the research and over 5 years, in which $54 million a year is to be spent. for
development program under the supervision of the National photovoltaic systems. Moreover, the negotiations for a joint
Science Foundation these past two or three years. NSF's venture by Mobil Oil Corp. and Tyco Laboratories, IInc.,wliich
expenditure for 1974 in the field of photovoltaic generation is was made public last surmner, reached agreement and Mobil
S8 million. According to NSF's plan, by 1977, a level of solar- Tyco Solar Energy Corp. was established this Janaury. Owned
cell technology that could produce electric power at a cost of 80% by Mobil and 20% by Tyco, the new company will develop
$5 per watt at the peak power output could be attained on the and commercialize solar cells made of silicon. ribbon fabricated
basis of refining present techniques. The goal by 1979 is to by Tyco's EFG method- Mobil, it is said, intends to invest
demonstrate the feasibility of dropping the cost by a factor of some $30 million over the next 5 years.
10, to 50 cents per watt at peak power output. This would Also in Japan, a plan of Kyoto Ccrainic Co., Ltd. for intro-
involve such new techniques as fabricating cells in continuous ducing Tyco's technology and establishing, together with two
silicon ribbons. In 1981, a pilot line for manufacturing arrays or three other companies, a joint enterprize for manufacturing
with a cost of 50 cents per watt, and in 1986, a pilot line for silicon ribbons for solar cells was announced in past January..
arrays with a cost of 30 cents per watt would be started In Japan, the national budget for fiscal 1975 is now under
respectively, establishing a mass-production system in 1990. deliberation in the Diet. The unofficially announced total
Moreover, many private companies are conducting research appropriation to the Sunshine project is said to be only Y3.7
and development in this field: Centralab., 1-Ieliotek Division of billion. The amount originally demanded was x`16.8 billion.
Textron Inc., Sylmar, Calif.; Solarex, Solar Power Corp., In this field too, like other fields, the scale of developmental.
Braintree, Mass.; and Solar Energy Systems Inc., Newark, Del. investment seems to be two figures smaller in comparison with
Some oil companies have invested in the research activities at the U.S.
these organizations. In consideration of the fact that Japan has world's second
Achievements in price reduction have been announced by largest semiconductor industry? and that the solar photovoltaic
some organizations, but so far they are limited to refinements technology might provide a way out for Japan with its scarcity
of the present techniques. Concerning silicon ribbons and of natural resources and its ever mounting oil import bill, it is
silicon thin films, many organizations are now carrying out desired that development should be strongly advanced with a
research. The most prominent among them is the research on long range vision, undertaking sufficient examinations of its
silicon ribbon growth by the aforementioned EFG method feasibility at each stage of development, and also obtaining a
conducted at Tyco Laboratories, Watthani, Mass. Further, the nationwide concensus on this plan..
Fig. 10. Estimated Costs for Large-scale Solar-cell Arrays (after F. R. Eldridge, the Mitre Corp.)
$i0000.
1973 1975
i977 1979 198;
1.983 1985 1987
New Production Line to
Produce I MW System
Improved or New Production Lines to
R a D Produce 10 - 1000 MW System
$1000
R6
I I I I I _I
10MW
Large Scale Production
Feasibility to Produce
1000 MW System(s)
a Q
100MW 1000MW
Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
Cum. Power Capacity 0.1 MW 1 MW
 Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
TO MR. CARY
FROM: Don DATE 3 July 1975
Letter to Daddario, OTA, encouraging OLC-channeled communication and
transmitting requested materials.
NOTE:
Lyle raised the issue of our burgeoning contacts with OTA with Mr. Duckett
personally and at a 9: 00 meeting last week. Consensus was that we should
kee these contacts within OLC channels. Mr. Duckett agreed to so inform
The attached letter to Emilio Daddario transmits information to
UT.A w is was requested by one of their employees and also enunciates
the new policy of dealing strictly through OLC. You may want to send a copy
of the letter and revised //memorandum to Mr. Duckett and to
COORDINATED WITH (list names as well as offices):
STAT
ACTION REQUIRED BY GLC: Signature on letter
Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
 Approved Fo
Ol l fl r7M00144R000600080014-5
CENTRAL INTELLIGENCE AGENCY
WASHINGTON, D.C. 20505
7~ -/y/ 41
2 S JUN 1975
MEMORANDUM FOR: Dr. Ronal Larson
Office of Technology Assessment
Congress of the United States
119 D Street N.E.
Washington, D.C. 20515
SUBJECT: Foreign Reference Material Relating to
Solar On-Site Electricity Generation
1. Reference is made to the letter from the
Director of the Office of Technology Assessment to the
Director, Central Intelligence Agency, dated 11 April 75,
relative to Congressional interest in the problem of
utilization of solar energy for on-site generation of
electricity and auxiliary heat. We are returning to
OTA herewith a copy of the draft report on Solar On-Site
Electricity which you left with us. Several analysts
have read it and have made marginal notations. In
addition, we asked fossil fuel and solar energy
specialists on the staff of our energy consultants,
to comment on some of the
comparative engineering economic data presented in the
report; their observations have been included at
appropriate points in the text. Inasmuch as our main
focus is on the foreign technology aspects of the issues
involved, we have limited our comments mainly to questions
of technical fact and analytical considerations, except
for the aforementioned remarks on costs.
2. We are also forwarding per your request the
information on Japanese and Soviet solar projects which
were described by
at the time of your visit on 15 May.
FOR OFFICIAL SSE ONLY
Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
 Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
Q
Approved For Release 2006/10/18: CIA-RDP77M00144R000600080014-5
 Approved. ForRelffl R2 I 1 L C ECf 11 0144R000600080014-5
SUBJECT: Foreign Reference Material Relating to Solar
On-Site Electricity Generation
A resurgence of interest in solar energy utilization
has now taken place within the Soviet government. In
January 1974, as part of the May 1972 US-USSR Agreement on
Cooperation in the field of Science and Technology, the two
nations signed a protocol for Cooperation in the General
Problems of Utilizing Solar Energy. Dr. Richard J. Green
of ERDA is the US coordinator for joint solar projects. One
of the three problem areas deemed appropriate for joint
research is solar power stations, designated as Program I
under the joint agreement.
The Soviets have reinstituted their own effort with plans
for the construction of a solar power station with an output
of 1-2.5 MWe. The former design (reflectors mounted on
moving flatcars circling the base of a tower supporting a
movable receiver-boiler) has been modified to reduce
construction costs. The new design involves reflectors
located in fixed positions around the base of a tower upon
which a stationary tubular boiler is mounted. The
heliostat system is made up of 1600 3-by-5m reflectors;
the mirrors can be oriented on two axes and track via a
photocell-activated servo system. The maximum radius from the
tower to the outer circle of reflectors is 200m. The mirrors
reflect the sun's radiation onto a boiler mounted on a 40m
high tower. The boiler is a vertical cylinder, 7m high and
9m in diameter wrapped in a helical array of tubes. The
system is designed to produce 11 tons/hr of super-heated
steam at 400-500 deg. C and 35 atm.
The Soviets plan to generate 1.2 MWe with a non-condensing
turbine or 2.5 MWe with a condensing system. No storage
(either thermal or electric) is involved and the plant is
expected to operate for 2000 hrs in the year. They anticipate
that the overall conversion efficiency of sunlight to electri-
city will be about 17% (based on the following stated
assumptions: mirror reflectivity, 0.78; interception of
radiation by boiler, 0.9; optical absorptivity of boiler
tubes, 0.9; thermal efficiency of boiler, 0.85; the efficiency
of conversion of solar radiation to steam is about 55% from
which, assuming turbine efficiencies of about 30%, the overall
plant efficiency comes out to approximately 17%).
[[~RR - 3
Approved For Release Y086~1 QFLFIWAIR D M LYR000600080014-5
 Approved For Remm Ff A(: U ER L1~o0144R000600080014-5
SUBJECT: Foreign Reference Material Relating to Solar
On-Site Electricity Generation
Design studies are virtually complete and site evalu-
ations for future construction have been performed. In
addition, one heliostat assembly has been built and field-
tested. Apparently, the Soviets intend to begin construction
of a power plant with the above specifications in the near
future. Individuals associated with the project have
indicated that they believe cost to be the determining
factor in the success of the system. Estimates for capital
investment of 1000-2000 rubles/kW have been noted, compared
with 150-180 rubles/kW for fossil fueled plants. (It should
be mentioned here that this solar plant is tiny in
comparison with normal Soviet power stations which tend to
be of the order of 300-600 MWe and larger in capacity. The
very large capital cost factor must therefore be viewed
as quite unrepresentative of a large commercial installation
to which the fossil fuel cost estimate probably does
correspond.)
3. If there are questions concerning any of the
enclosed references or comments, please feel free to contact
us again.
Chief
(-'`-Physical Sciences and
Technology Division/SI
Attachment: a/s
- 4 -
Approved For Release / -R{ 'Rc7MI *R000600080014-5