Peter F. Varadi
P/V Enterprises, Inc. Chevy Chase, MD, USA
The successful utilization of Photovoltaic (PV) cells to supply the electric needs of satellites used for various purposes gave the idea and an assurance that PV could also be used successfully for large scale terrestrial applications. The obvious problem was that at that time the price of PV cells made for the satellites was about $300/Wp. By 1972, several of the US government organizations, which had needs for telecommunication in remote areas, decided to try to use PV systems to power those equipment. The two US manufacturers of solar cells and panels for space use, Heliotek (Spectrolab) and Centralab, both located in California, were contacted to supply PV cells connected in a module to provide 12 V dc power. Obviously the price of the space solar cells at that time was still over $100/Wp and therefore to use them for these terrestrial applications was not feasible, only PV cells which were rejects for space could be used. The amount of these rejects were not much and only a very few experimental telecommunication systems were made.
1972 was also the year that scientists and academics started to discuss the utilization of PV for terrestrial purposes. In May 1972, the 9th IEEE PV Specialist conference was the first to include a session on terrestrial PV. In the same year, the Institute of Energy Conversion was established at the University of Delaware (USA) for the research of thin film PV.
Dr. Joseph Lindmayer the head of the Physics laboratory of the Communications Satellite Corporation’s (COMSAT) Research Center in Clarksburg, Maryland was successfully working to improve the efficiency of PV solar cells for satellite use, which was an extremely important project for COMSAT as the transmission capacity of the communication satellites depended on the amount of electricity generated by their PV solar modules. Lindmayer developed the so- called "violet cells". This was considered a breakthrough because the "violet cells" were able to increase the efficiency, to convert light to electricity of PV solar cells by 50%, compared to the conventional solar cells produced at that time.
Lindmayer and his colleague Dr. Peter F. Varadi, who was the head of the Material Science Department of COMSAT Research Center, realized that PV could also be used for terrestrial applications, but for this application the PV cells had to be considerably less expensive than the PV used for space applications. They realized this would require a technology very different from the technology utilized to produce solar cells and solar modules for space applications.
After COMSAT’s management did not approve any expenditures to develop technology to produce inexpensive PV cells for terrestrial applications, Lindmayer and Varadi — two Hungarian refugees — decided on December 31, 1972 to leave their very good jobs and start a company, which they named SOLAREX, devoted entirely to develop low price PV products to be utilized for terrestrial applications, to provide electricity to power electrical equipments and sell these at a price, which would be attractive to users, for telecommunication and for other electrical equipment.
In 1973, two important conferences were focusing on the possibility utilizing PV for terrestrial purposes. One was "The Sun in the Service of Mankind" held in July in Paris. The other, the so-called "Cherry Hill Conference" was held in October 1973. This conference by coincidence was held at the beginning of the "oil embargo" and it brought the US government’s attention to the possibility of the utilization of PV to avert future energy crisis. While these conferences were important events directing the spotlight to the utilization of PV as an alternative to the nuclear and fossil fuel energies, but they did not tried to solve the issue of the commercialization of PV for terrestrial purposes.
When Lindmayer and Varadi after the first euphoria — that they were now entrepreneurs and pioneers in an extremely interesting field — faded away they reviewed the situation, and realized that while PV in the future will be surely the alternative to the fossil fuel and the nuclear energy sources, but that would be only in a very distant future. When they opened the doors of Solarex on August 1, 1973, they opened the door to a totally unexplored new field of science, technology and business.
Lindmayer concentrated on the technology. His idea was to reduce the cost of the machinery needed to produce PV cells and modules, to eliminate vacuum and batch processing, which was essential for space solar cells and that all the production should be under atmospheric pressure utilizing off the shelf equipments, production bands and automation. Also the technology to be developed should be able to produce good quality Silicon (Si) solar cells even from the cheapest low quality wafers. He also had ideas to develop new and less expensive technology to produce Si wafers.
Varadi realized, that if he would still be a scientist, which he was one month before Solarex was started, he would also give papers, predicting the future possibility of the $0.5-1/Wp PV cells and its utilization to replace nuclear and fossil fuel to power the utilities. But life looked entirely different, when he had to worry to pay the employees and obviously himself and Joseph and the suppliers and try to achieve that they should produce at least 10 000 to 50 000 Wp in the first year at a sales price of about $ 10-30/Wp.
This price was realistic, but it was obviously not for the utility market and therefore his idea was to develop a commercial and diversified market for PV, a market, where PV would be an inexpensive solution compared to what the customer was using at that time. He also believed to rely as little as possible on government-supported business. He believed, that if Solarex had hundred commercial customers and lost one, that would be only 1% of the business, but if the only customer was the government and due to changes in its priorities and the support was cut, 100% of the business would be lost and that would be the end of Solarex. Therefore Solarex was concentrating to develop the commercial applications of PV.
This idea was very interesting, but in reality the issue, to find hundred customers, was extremely complicated. PV modules could be used for a great variety of applications. The first question was to find out what these applications were. What nobody predicted happened in October 1973, shortly after Solarex started the production. This was the oil embargo. That was a wake-up call for the people in the USA and provided for the tiny PV industry — if one could call it an "industry" — an incredible free advertising and exposure. Solarex was located in the suburbs of Washington DC and was visited daily by reporters and TV crews.
The effect of this free advertising was so great, that when Solarex’s sales person paid to put a "bingo card" in a magazine, the post office had to send a small truck carrying at least four large bags filled with "bingo cards". Obviously this should have solved the issue to find applications for PV, but as it turned out some of the applications where for a housewife paying the family house’s electric bill got the idea of mounting solar modules on the roof of their house, which would provide them with free electricity from the sun. She probably would have gotten a heard attack if somebody would have called her and tell her what the price would be. But there were many applications, in which PV even with those prices would be cost effective. The problem was which of the many applications would lead to open up a market for PV? Varadi used to say: "Many people can make solar cells, but few people know where to sell them."
When Dr. Wolfgang Palz, at that time the chief for Energy Development at the French National Space Agency in Paris, visited Solarex in those days and got into conversation with Joseph about large PV systems for utilities, Varadi used to say: "Where is the purchase order?"
Without purchase orders for utility scale PV systems, the business started to develop. The United States had several organizations, which were taking care of vast areas where nobody was living, but communication was still needed. US Forest Service, Bureau of Land Management, Weather bureau, and in many states, e. g. Arizona, the police. PV was tried and was performing very well and these organizations used it. The first mass produced PV system utilized for these applications was the Solarex H and HP type series (6 to 10 Wp) (Fig. 1) of which many thousand units were fabricated and installed world wide from 1974 to 1979.
For navigation aids PV offered big advantages. Canada was the very first in the world to start a program to convert all of their navigational aids to solar power. Mr. Sunny Leung was the architect of the Canadian Coast Guard Program. The very first solar powered lighthouse in the world was the one at False Duck, Ontario (Fig. 2). This lighthouse was commissioned in 1981. It was a hybrid system with two diesels and had 22 pc. of 30 Wp Solarex modules (Fig. 3) The load was reduced in 1995 and the diesels were removed and 18 out of the 22 modules
Figure 1. The first mass produced PV system (1974-1978).
Figure 2. The first Solar Powered Lighthouse: Falls Duck, Ontario, Canada.
Figure 3. Solar modules — Falls Duck, Ontario, Canada.
were transferred to other locations. Presently, after 28 years of service, the original four modules are still operating the lighthouse.
The northern-most navigational aid installed by the Canadian Coast Guard, powered by PV is at Resolute Bay, 150 km from the magnetic North Pole of the
Figure 4. Navigational aid near the Magnetic North Pole, Canada.
Earth (Fig. 4). This navigational aid was also installed in 1981 and is still in operation.
The introduction of PV to power the US Coast Guard’s navigational aids was technically more complicated because of the environmental conditions: the relatively warm salt ocean waters, atmospheric conditions, high humidity and strong UV radiation, as failure was not an option, the PV modules required very protective encapsulation for the solar cells and their interconnections. The US Coast Guard under the direction of captain Lloyd Lomer, conducted for many years testing and based on that, established a testing procedure and came up with a specification to use only PV modules fulfilling that specification and test procedure (Fig. 5). The US Coast Guard converted in the early 1980’s all of the navigational aids to be powered by PV. President Ronald Reagan commended Lomer for "saving a substantial amount of the taxpayer’s money through your initiative and managerial effectiveness as project manager for the conversion of aids to navigation from battery to solar photovoltaic power." The savings utilizing PV to charge the batteries in Navigational Aids with PV for the Canadian and the US Coast Guards were substantial. It eliminated the need to send ships to periodically replace batteries.
Utilizing Arco Solar cells and modules two manufacturers in the Houston and New Orleans area, Automatic Power Corporation and Tideland Signal equipped large number of oil platforms with PV navigational aids.
Figure 5. US Coast Guard module.
The Suez Canal was reopened in June 1975. The navigational aids used at that time were replaced in 1982 by 628 modern buoys produced by the Resinex Corporation of Italy. All of them were equipped with PV powered lights, (Fig. 6). The utilized PV system, modules and charge controllers were supplied by Solarex Corporation, providing also engineering supervision of the installed PV systems.
This is also a good example that even at the beginning, when PV was still expensive there was a market for it, where PV was less expensive than the alternative. The success of utilizing PV for navigational aids is only one example.
Other important market areas developed where PV was very useful. Obviously telecommunication was one of the major area, but cathodic protection, water pumping and others including homes in developed countries e. g. in Europe or America, which had no access to the electric grid but needed electricity for lighting and for radio/TV reception.
This expanding PV production was many times affected by the availability of its basic material, the Si wafer. PV cells could use Si wafers the quality of which was not satisfactory to produce semiconductor devices. The availability of these relatively low cost Si wafers depended on the requirement for semiconductor devices. If the demand was high, the available not satisfactory wafers were also high, if the demand for semiconductor devices was low, there was a shortage of the so-called "solar grade" wafers and their price obviously increased. Lindmayer and Varadi decided, that their expanding PV business could not rely on the unpredictable availability and price of their most important material, the Si wafer. One alternative was to buy crystal pullers and fabricate their own wafers, but that would have required a very large amount of capital, which was not available to
Figure 6. Suez Canal — solar powered navigational aid — 1982.
Solarex and would not yield relative inexpensive wafers. Therefore Lindmayer started to develop one of his original ideas, to simply melt the Si material and cast it in a crucible. After the Si solidified in a block, it can be made into wafers. Lindmayer developed a very simple casting technique, which could also use relatively inexpensive Si and also developed the technology to produce PV solar cells from the wafers produced by casting.
The process of producing Si wafers utilizing the casting technique was extremely successful and today close to half of the world’s production of PV solar cells use these wafers which are called "multicrystalline" wafers to differentiate them from wafers produced in crystal pullers which are called "single crystal" wafers. Solarex made the World’s first large PV installations utilizing these multicrystalline wafers. The first was in 1982, a 200 kWp roof installed on Solarex’s factory in Frederick, MD, USA (Fig. 7). It was called the "Solar Breeder", because the electricity produced by solar cells was used to produce solar cells. The next one was a 300 kWp multicrystalline roof installed on the roof of Georgetown University’s Intercultural Center in Washington DC in 1983 (Fig. 8). Both of these systems are still in operation.
Arco Solar with ARCO financing built the first one megawatt PV power plant for utility grid support in 1982. The system was located adjacent to the Southern California Edison substation in the Mojave Desert community of Hesperia in Southern California.
The October 23-25, 1973 Cherry Hill conference, had two important effects on the development of PV:
It called on the attention of the US Government to the possibilities in PV for the Research Applied to National Needs (NSF/RANN) program. As a result, the
Figure 7. Solarex’s 200 kWp factory roof ("The Solar Breeder") — 1982.
Figure 8. Georgetown University, Intercultural Center, Washington, DC: 300 kWp building integrated PV — 1983.
US government recognized the need to expand research and development activities in alternative forms of energy, which included PV and established in 1975 the Energy Research and Development Administration (ERDA). The program received bigger recognition, when in 1977 the US Government consolidated the federal energy policy and ERDA was integrated with the Federal Energy Administration and with other federal energy functions and created a Cabinet level U. S. Department of Energy (DOE). The result was that during the presidency of Jimmy Carter (1977-1981) substantial money was earmarked for PV R&D and for demonstration project.
The other and for the future success of the utilization of PV for terrestrial applications an extremely important result was, that in January, 1975 the US government initiated a terrestrial PV research and development project, assigned to the Jet Propulsion Laboratory (JPL), which was patterned after the recommendations of the Cherry Hill Conference. At JPL, John V. Goldsmith who headed JPL’s space oriented PV activities became the Technical Manager also for the terrestrial PV program, the aim of which was to help the terrestrial PV industry to reduce prices and produce reliable PV modules. The JPL program was extended to Si material, crystal growth of Si, encapsulation of solar cells and also fabrication methods of PV modules. Goldsmith, who spent many years working in the field of space oriented PV systems, knew very well, that one of the most important issue was to achieve excellent quality PV modules.
To achieve excellent quality PV modules, utilizing the experiences from the space programs, JPL established a test specification for terrestrial PV modules and was buying blocks of modules from manufacturers who could qualify their products. The first block (Block I: 1975 to 1976) purchased 54 kW off the shelf modules, to establish what is available. The specification was simple, it required verifying the electrical test per manufacturers ratings and environmental tests limited to: temperature cycle and humidity soak. Five companies participated in Block I: M7 International, Sensor Technology, Solarex, Solar Power and Spectrolab. Even these simple tests required several design improvements during production.
In Block II modules totaling 127 kW were purchased. The design (e. g. the requirements of interconnect and terminal redundancy) and testing specifications (thermal and humidity cycle, structural loading) were extended. In this program Sensor Technology, Solarex, Solar Power Corporation and Spectrolab participated.
Included in the Block II program was also the introduction of a quality management program. At that time the ISO quality management system did not exist (it was introduced only in 1987). JPL introduced a Quality Management System, which was very similar to what ISO introduced ten years later. The manufacturer had to prepare a Quality Assurance Plan, including inspection criteria. The program had to be approved by JPL personnel. Manufacturers had also to agree, that a JPL inspector could be in residence and observe the execution of the QA program.
These two blocks were followed by three more block buys ending with Block V. The test specifications were tightened in each of them, based on the experiences in the previous Block buys. It was expected, that PV modules produced, which passed JPL’s Block V tests would be able to have a 20 year warranty.
In retrospect, the JPL Block I — V program bought a significant amount of PV modules (in the Block III program in 1978, 30 to 50 kWp was purchased from each of five manufacturers, which at that time was a large quantity). This attracted more companies e. g. General Electric, Motorola. Spire and Arco Solar to participate, and the manufacturers were eager to qualify even if they had to comply with rigorous specifications and their enforcement. This resulted that the quality and reliability of the PV modules were vastly improved. Without this program the expected failures would have destroyed the image and usefulness of PV. The JPL Block program provided manufacturers a means to evaluate quality and the resulting expected life of their PV modules and they were able to offer a warranty long enough to make PV competitive with other electric generation systems.
JPL’s final, the Block V test specification became the PV industry’s standard. Later it was somewhat revised in the USA and became IEEE1262 and also by the European Union’s Solar Test Center, JRC-Ispra (JRC-Ispra 503: Qualification Test Procedures for Crystalline Silicon PV modules). Finally these were merged and became the presently used International Electrotechnical Commission (IEC) standard (IEC 61215 Crystalline silicon terrestrial photovoltaic modules — Design qualification and type approval).
In spite of the participation of several companies in the "block buys", in the late 1970s only a few meaningful terrestrial PV manufacturers remained in the USA: Solarex, Inc., was privately owned, Solar Power Corporation started by Mr. Elliott Berman was supported, and was in 1975 taken over entirely by Exxon. When Bill Yerkes, President of Spectralab, left Spectralab in 1975, started Solar Technology International (STI). STI was acquired in 1977 by Atlantic Richfield (ARCO) and was renamed to ARCO Solar. Mobil in 1974 established Mobil-Tyco to produce thin sheet silicon wafers to produce solar cells.
After 1976, the situation changed. The interest to use PV became global. Besides of these US manufacturers of PV there was still not many in Europe. Philips had a small operation in Paris and sold PV modules in Australia. Philips however discontinued this venture, when they encountered technical problems with their product. Another entry in France was Photowatt, which participated in 1980/81 in the JPL Block IV program with a 38 W module. In the UK BP acquired a small company and was involved in successfully installing PV systems in many countries of the world.
The demand was worldwide and Solarex expanded outside of the USA. In Europe it established in Switzerland a manufacturing operation for multicrystalline silicon produced by the casting process and module assembly and started to manufacture flash simulators. It formed joint ventures in France and Italy to manufacture solar cells and also modules. It started a factory in Australia, also for the manufacturing of PV modules and later also solar cells. It acquired the company of its representative in the UK and manufactured electronics especially for cathodic protection systems. It also established a factory in Hong Kong and manufactured small PV modules and novelty items powered by PV cells.
ARCO Solar under the direction of Mr. Charles Gay developed its worldwide sales of its PV products. It established joint ventures in Europe with Siemens and in Japan with SHOWA Shell. ARCO Solar’s sales was also expanded to the southern hemisphere, which only had access to diesel generator sets.
The beginning of the 1980s brought a fundamental change for PV in the US. In 1980, President Carter, under whose presidency from 1977 until 1981, the US Government which had exhibited great support for PV, lost the election. The new administration under President Reagan who took office in January 1981 cut the budget and the support for PV was drastically reduced. This had a great effect on PV companies, whose primary customer was the US Government. Two companies were not much affected, because of they developed a global commercial market for PV. One was Solarex, the other was ARCO Solar.
In case of Solarex, Varadi believed in expanding the commercial market and minimize the government dependence, furthermore he did not believe, that the large utility market was imminent. He believed also, that Lindmayer was an extremely competent scientist with lots of practical ideas and his research and development staff was extremely good to work out his ideas. An additional pillar to support Solarex’s success was that John V. Goldsmith joined Solarex and brought his expertise of running large programs in the field of PV. Solarex therefore could be and became very diversified. It had a large international presence, having factories on four continents. Solarex started to design and install large PV systems, also developed a sizeable consumer business, manufacturing PV products for this purpose in Hong Kong and selling them in catalogues and establishing Renewable Energy oriented stores under the name of "Energy Sciences" in the best shopping malls in the Washington — Baltimore area.
Solarex formed a corrosion protection division under the direction of Mr. Ramon Dominguez, selling turnkey cathodic protection systems powered by PV. Under the direction of Goldsmith, it developed a business to make PV solar cells and panels for space application.
Solarex was also involved in producing miniature PV arrays to power electronic watches and calculators. To their big surprise, this business was lost suddenly because some Japanese companies started to manufacture the thin film a-Si (amorphous silicon) mini PV arrays. For calculators, this was not only cheaper, but its spectral response was also better. While the crystalline Si PV array worked better in a bar at candlelight, the a-Si worked better in an office illuminated by fluorescent light.
That was the first time thin film PV was used in quantity. The idea of utilizing thin film PV solar cells started with selenium cells, but the efficiency was very low. In the early 1970s, there were attempts to use the CdS/Cu2S thin film PV cell. A factory was built by Shell in Delaware (USA), but it turned out that the product was extremely moisture-sensitive and would have needed moisture impermeable hermetic sealed encapsulation. Many other thin film structures were experimented with, a-Si, CIGS (Cupper-Indium-Gallium di-Selenides), CdTe (Cadmium Tellurides). The first successful system was a-Si (amorphous silicon) developed by the group headed by Dr. David Carlson at the RCA Research Laboratories, in Princeton, NJ. When the government support started to decline, RCA did not want to support further the a-Si development work. Solarex realizing that the a-Si is a viable PV system acquired from RCA the a-Si patents, know-how and equipment and hired all of the scientists working on the subject and opened an R&D and manufacturing facility in Newtown PA, near Princeton.
ARCO Solar on the other hand introduced also a-Si modules and started to develop the CIGS system and set up a small facility for thin film PV modules. Interestingly the CdTe idea was dormant for 25 years, when it was developed and became an extremely important factor in the PV business.
In September 1983, AMOCO acquired Solarex. Solar Power Corporation, which was relying heavily on government programs, by being involved in several demonstration projects, became the casualty of the Reagan era and EXXON announced that it will discontinue its operation and Solarex acquired the Solar Power Corporation assets on June 15, 1984.
About ten years after the beginning of the terrestrial PV business with these consolidations only two major PV manufacturers, Solarex and ARCO Solar, both owned by a major oil company, remained.