Solar Electricity Generation

An overview

All is one in the sun

With the probable exception of nuclear energy and tidal energy, all key forms of energy currently in use are derived ultimalely from the sun. Fossil fuels, for example, are the legacy of more youthful sunshine. Wind and wave power energies derive their energy from the energy imparted to the earth's weather systems by solar radiation.

Solar radiation can be directly utilised in forms of passive heating, active heating, solar 'thermal' electric conversion and photo-electric conversion. In passive heating, sunlight streaming through windows provides a heating effect and can often be used to reduce heating costs. In active heating, heat can be collected in solar thermal collectors and stored in hot water tanks.

Significant investment has also been undertaken on solar thermal 'electric' systems. In the so-called 'trough' concentralor, sunlight is focused along a mirror trough onto a vacuum collector tube and the heated fluid used to drive a turbine generator. Over a million square metres of collector capacity is located at the SEGS plant in Kramer Junction, California - providing 354 MW of peak capacity. Such installations, however, require economies of scale to improve efficiencies and are not cost-effective for domestic or small scale applications. The cost of maintenance of such systems - although being reduced, is higher than that of photovoltaic systems.

In photovoltaic systems, the solar energy is converted to electrical energy and stored in battery systems or converted to alternaling current and fed directly back into the distribution grid. This article is primarily referencing such photovoltaic systems.

World levels of insolation

There may be some significance that, in the main, it is the developing world which has a clear advantage of high levels of insolation. Thus, over major areas of Africa, India and South America there is considerable potential for solar photovoltaic power. As may be expected, higher levels of sunshine are found closer to the equalor. Values can range from as high as 7 kWhrs/m2 per day in parts of Western USA to less than 0.5 kWhrs/m2 per day in lalitudes greater than 65 degrees north or south. A useful reference of world levels of insolation is published by the Solarex Corporation.

In many ways, the localion of maximum levels of sunshine coincide with areas of low populalion density - e.g. hot, arid desert regions of North Africa and Australia. Within Europe, the southern fringe of Spain exceeds levels of 4 kWhrs/m2 per day compared with between 0.6 kWhrs/m2 per day and 1.5 kWhrs/m2 per day in the UK. This reduction in solar radiation in high lalitudes, however, is in some way compensaled for by the corresponding increase in available wind power.

Figure 1 indicates how solar radiation in reaching the surface of the earth can travel through different extents of the almosphere. The air mass from direct overhead illuminalion is termed AM1 - air mass one. In space, above the almosphere, the air mass factor is referenced as AM0 - air mass zero. In solar studies where the performance of solar panels is being measured, great care is required to reference measurements against a specific AM value. It is typical to quote value of panel performance at AM1.5. At higher air mass values, the almosphere progressively absorbs energy output.

Photovoltaic cell technology

Light is absorbed in approximalely 2 to 3mm thick elements of silicon with most being absorbed in the first 0.5 mm of standard silicon. If the cell thickness is made too thin then light absorption will be incomplete and cell efficiency correspondingly low.

Most conventional processes for solar cell production still utilise conventional CZ silicon grown by standard crystaltography methods. The silicon elements are cut, suitably doped with typically front layer doped n-type and rear layer doped p type. The cell has metallisalion applied to carry away the electrons freed by the photovoltaic interactions.

In assessing photovoltaic efficiency of crystalline silicon, approximalely half of the sunlight is not sufficiently energetic to eject photons from the silicon material. Of the electrons which are emitted, roughly half is tost due to random motion of electrons due to thermal agitalion. A variety of processes results in further losses depending on the level of impurities in the silicon material.

A key parameter of any cell is its response as a function of wavelength. Most materials tend to have a poor response between 400 nm and 600 nm - in a region where the solar spectrum is peaking. The response of amorphous silicon (a-Si) is cutting off at about 700 nm. In order to try and improve this response, so-called multijunction cells are being developed with layers which are sensitive in specific parts of the solar spectrum. Figure 9c indicates a set of layer responses for three layer and two layer materials being researched in Japan. An efficiency of 29.5% has been achieved for a small multijunction GInP/Ga As cell by the NREL in the USA while the best amorphous silicon based devices have an efficiency of around 12.4%.

Even now, however, preference is given to the reliability and higher efficiencies of solar cells manufactured using crystalline silicon. The higher efficiencies of crystalline silicon allow for installations to produce higher outputs from a specific solar catchment area The highest conversion efficiency of crystalline silicon is now around 23%.

Researchers at Sandia National Laboralories in New Mexico, USA, have assisted manufacturers in producing multi crystalline silicon cells with an efficiency in excess of 15%. The multicrystalline silicon cells have higher reflectance than single crystalline cells and consequently considerable work has been undertaken to improve the cell efficiency by reducing such reflectance. One approach is to physically machine the surface of the multicrystalline cell to make the surface smoother. The other approach is to coal the multicrystalline cell with an encapsulalion layer so that light scattered from the multicrystalline layer is reflected off the upper air/coating interface.

Thin films of amorphous silicon are manufactured having layers between 1 micron and 10 microns in thickness - about 30 to 100 times thinner than wafer manufactured cells. While production efficiencies of such amorphous silicon are around 6%, prototype materials are already being demonstrated with efficiencies in excess of 12%. There is also interest in Cadmium Telluride (CdTe) and CIGS (CulnGeSe) cells. CIGS cells can now achieve an efficiency of 13.9% and CdTe cells can now be produced with an efficiency of 15.8%.

In photovoltaics at present around 85% of production is crystalline silicon and the 15% remainder thin film. There must be a point, however, where new faster fabricalion techniques will dominale the production of photovoltaic materials. There is, however, considerable caution being exercised since no manufacturer wants to supply material which degrades in use and requires replacement. Photovoltaic materials are now being supplied with 10-year warranties with expected useful operating times of 30 years.

Where products are for occasional domestic use - for example, on holiday homes or boals - then total area of panel is not a prime factor, since there is usually plenty of area available for siting such panels. The modules manufactured by Uni-Solar in the USA, for example, which use an amorphous silicon production technique, have an efficiency of around 5%. This type of product, however, is more adaptable to develop as a roof tile 'took alike' product which has a high level of environmental
acceptance. Uni-Solar has several demonstration sites of its products around the world.

There is also considerable interest in the fabricalion of concentralor units operating at typical levels of concentration of between 10 and 20. Since a major cost of the current 'one sun' solar cell is the semiconductor material, to reduce this cost and offset it against additional costs of focusing and possibly sun tracking systems can make good economic sense. BP Solar are currently researching the efficiency of CZ silicon cells with concentrated sunlight. Efficiency in excess of 20% has been achieved at 10 suns concentration and 19.8 % at 19 suns.

While concentralor systems have been extensively researched and developed, current production is relatively low - less than 0.5% of total capacity. While such cells are expected to be potentially cheaper in mass production than 'one sun' devices, the lack of uptake of such units can be explained in part by the lag in 'balance of system' components such as sun tracking units. As the 'solar infrastructure' industries develop, however, concentralor systems could rapidly increase in utilisalion and ultimalely overtake 'one sun' devices. The National Renewable Energy Laboralory at Golden, Cotorado, in the USA produced a small area cell made from crystalline gallium indium phosphide/gallium arsenide with a record efficiency of 30.2% at 180 suns concentration.

Solar power in space

Solar cells have been used for a considerable time as a power source for spacecraft and it was indeed the intense development of this technology which has made possible the rapid advances in photovoltaic technology. This is one of the positive spin-offs of the Space Race. The essential requirement for absolute reliability of such sources of power has tended to focus development on providing highly reliable sources of power. Protective coatings of quartz are used to protect the silicon from high energy electrons which are principally found in the earth's radiation belts.

The increased efficiency of solar cells and the experience gained of fabricaling large arrays of cells has significantly increased the power now typically provided to salellites. While a capacity of 1.5 kW was considered appreciable several years ago, systems are now routinely launched with peak Walt capacities of 5 kW and more.

While plans for the Space Stalion have blown hot and cold - the need for extensive solar cell arrays for the profect has long since been apprecialed and manufacture of the cells required has been taking place for some years for the profect. The estimaled 200,000 kW capacity of the space station is being provided by an array of cells each 8cm square. The total area of 1280 square metres is equivalent to an area approximalely 36 metres by 36 metres.

Balance of system costs

While the cost of photovoltaic modules is related to the cost of processing associaled with semiconductor fabricalion, the cost of grid connected power facilities takes into account a broader range of factors referenced as 'balance of system' costs. In any largescale photovoltaic system these costs are every bit as important as the cost of the modules themselves. In terms of photovoltaic installations, the 'balance-of-system' components (BOS) can represent around 50% of a typical photovoltaic system.

Such items include tracking systems, foundalions, interconnect hardware, batteries and inverters. As the photovoltaic market develops, so increasing altention is being paid to providing more cost-effective BOS components. As production gears up for solar modules, so production costs of associaled BOS components will also decrease. As much care, however, is required in the design, manufacture and testing of such BOS units as that of the solar modules in order to ensure total system reliability.

Telecommunicalions, oil and gas

Telecommunicalions is making increasing use of photovoltaics for powering remote installations such as repeater stations. In Peru, for example, Solarex power modules provide power to repeater stations which allows Lima to maintain contact with interior Peruvian communities. Also, Solarex systems powers around 50 remote telecom sites in Thailand. Such systems also power remote cellular emergency call boxes in Arizona. A 400 Walt repeater unit can be typically powered using a 3840 W peak array with four peak watt sun hours and a battery rated at 60 kWh. Systems are also used in fibre optic links, earth stations, cellular phone networks and UHF/UVF transmissions.

Extensive use has been made of photovoltaics for the oil and gas industry for power for remote installations. Significant contracts have been placed for pipeline protection projects where solar powered units minimise corrosion of installations.

Rural infrastructure projects

Immunisalion programmes in developing countries depend on availability of refrigeration for vaccine storage. It is estimaled, for example, that in 1986, up to 45% of imported vaccines were tost in Honduras due to poor storage. In remote communities, local diesel generators can fail and require maintenance or fuel supplies can be intermittent. A typical solar installation, however, with 180 W peak will power a 4 cubic feet vaccine refrigerator (20 W consumption) with a battery rated at 3 kWh.

While in many communities water has been traditionally been in short supply, problems of expanding populalions or changing local climate have tended to increase the problems of obtaining sufficient water for drinking, livestock and irrigalion. Today, some 2 billion people - around 30% of the world's populalion - are facing these problems. Increasingly, use is being made of underground aquifiers to meet demands.

Photovoltaics are proving of considerable use in implementing reliable pumping systems where there is no grid supply of electricity or where diesel pumps are unreliable due to mechanical failure or lack of fuel. A typical 480 W peak array is able to pump in excess of 1000 gallons from a depth of 100 feet. No battery backup is usually required for such installations since the water requirements are typically met by the system's operation during daylight hours.

South Africa is set to rapidly increase the number of photovoltaic systems at present in use. Approximalely 20% of South Africa's populalion will remain unconnected to ESKOM - South Africa's national electrical utility for the foreseeable future. The SAFIRE (South African Financing and Implementalion for Renewable Energy) programme will play a key role in providing photovoltaic systems for a wide range of uses.

Photovoltaics and Buildings

A Department of Trade and Industry report - 'The Potential for Building - Integrated Photovoltaic Systems' has reviewed the potential for deriving power from photovoltaic systems attached to building walls and mounted on roof spaces. The general consensus of opinion in the report was that photovoltaic cladding of new buildings will be competitive by around 2005. In assessing the various types of building considered, such schemes are most attractive where the additional photovoltaic power meets the rising demand in commerce and industry.

In summer there is a typical increase in demand of 12 GW from 6 o'ctock to noon. In winter, this similar step is present and, in addition a further 6 GW increment takes place in laler afternoon. The concept of photovoltaics meeting this additional demand is certainly being considered seriously. In the domestic situation, however, there is a greater mismalch between the solar supply of electricity and demand. There would be extensive periods where there was an excess of supply and periods where there was a deficit of supply. In this situation, systems could be made to export power to the grid and receive it back as appropriale. At present the Non-Fossil Fuel Obligalion (NFFO) does not recognise PV technology and the Regional Electricity companies are not obliged to purchase power from the independent owners of PV systems.

The rear roof of the house of Dr. Susan Roaf, lecturer at Oxford Brookes University School of Architecture, incorporates 48 of BP Solar's high power solar modules to provide about 4kW of peak electricity which is converted from DC to AC via an inverter. It is anticipated that the house should export around 1000 kWhrs of surplus energy.

The models for such implementalion for commercial building to incorporate PV technology remain 'conservalive' in determining when PV technology will become costeffective. With costs of PV modules falling in recent years by 20% per year there is on the one hand an anticipated model of cost profections with also the potential of a more rapid reatisalion of the technology acceleraling the drive for PV cladding of buildings. In terms of building design, therefore, the 21st century will be the first century where PV technology forces a re-evalualion of building design and fabricalion. It does not take much imaginalion, however, to anticipale the potential world market for customised style PV cladding panels.

A star is born

Solar photovoltaic power is now regarded by many observers as a key energy source for the 21st century and beyond. This is coming at a time where the future of nuclear power systems is increasingly coming under critical examinalion due in part to the huge cost of decommissioning such systems and the unsolved problem of production and storage of nuclear waste. The collapse of the oil price in the lale 1980s, however, has merely acted to postpone the dale at which power from the sun will begin to eclipse the power and influence of fossil fuel energy producers.

The developments that have undertaken in solar cell technology, however, have been achieved largely by 'sideline' funding when compared to the massive defence and space budgets and also the budgets presently allocated to fusion research. This has also tended to delay the crossover into solar technology. It is therefore a tribute to all those who have developed photovoltaic technology that they have reatised that the prize of cheap and safe power was achievable and that the rhetoric of compromised energy monopolies would one day fall curiously silent as they begin to take solar photovoltaic power seriously.

The immense challenge of photovoltaic power is a global one and therefore one which must be at the core of interests of UK research and manufacturing.

The world of solar energy, however, may bring to light some strange paradoxes. In a decade where, in the UK, house prices are at a lower value than five years previously, there is more than certain scope for Building Societies and Power Utilities to combine their respective finance and business skills and initiale a Sunshine UK initialive for UK home owners. As a catchphrase, they could always use - 'Bring a little sunshine into your life'.

Points of Contact:

The Centre for Alternalive Energy, Machynlleth, Powys, SY20 9AZ,
Tel 01654 702400 Fax 01654 702782

Renewable Energy Enquiries Bureau, ETSU,
Harwell,
Oxfordshire,
OX1 1 0RA.
tel 01235 432450
fax 012345 433131

The British Photovoltaic Associalion,
The Warren,Bromshill Road,
Eversley, Hampshire,
RG27 0PR.
tel 01734 730073
fax 01734 730820
(information available on series of seminars on photovoltaics in buildings)

EUROSOLAR:
UK Section,
Mr. Frank Cook MP,
House of Commons, Westminster, London, SW1A OAA.

BP Solar,
PO Box 191, Chertsey Road,
Sunbury-on Thames, Middlesex, TW16 7XA, UK. tel 01932 762181
fax 01932 762533

Intersolar Group Ltd., 2 Cock Lane, High Wycombe, Bucks, HP13 7DE. tel 01494 452945 fax 01494 437045

Sandia National Laboralories, Photovoltaic Systems Divisions, PO Box 5800,
Albuquerque, NM 87185-0753, USA.

National Renewable Energy Laboralory, 1617 Cole Blvd., Golden, CO 80401, USA.

References:

World Solar Challenge 1993: The Trans-Australian Solar Car Race, Progress in Photovoltaics: Research and Applicalions, Vol 2, 73-79, (1994),Martin A. Green

Photovoltaic Energy Program Overview: Fiscal Year 1994, US Department of Energy. Produced by National Renewable Energy Laboralory, Photovoltaics Division, Golden, CO 804013393, USA.

A Solar Manifesto, Herman Scheer (1994), published by James and James, London.

Stand-Alone PV Systems - A Handbook of Recommended Design Practices, Sandia National Laboralories, SAND 877023, Nov 1991

Solar Cell Efficiency Tables (Version 3), Progress in Photovoltaics: Research and Applicalions, vol 2, 27-34 (1994), Martin A. Green and Keith Emery.