The chances of wind power joining the GW-Plus club were raised with a recent bid under the UK's third offshore round for a 1350 MW (1.35 GW) wind energy project off the Kent coast. And were it not for the present financial crisis, a certain US billionaire would be moving forward on a proposed 4 GW wind farm for Texas.

Can solar power follow? Until recently, most commentators would have had their doubts. Solar farms based on conventional semiconductor solar photovoltaic (PV) arrays can generate power at utility scale, albeit expensively and while occupying large areas of land. But concentrating solar rays by up to several hundred times offers a fresh slant on solar energy which is rapidly becoming viable and has great scaling potential. This is especially so when it is married with solar thermal power capture, thereby avoiding the use of silicon. Given the boosting of solar prospects by re-emergent concentrating solar power (CSP) technology, there is now optimism that solar might indeed get into the big league.

The word 're-emergent' is apt because researchers and engineers have been round this loop before, significant power stations having been built, notably in the USA in the 1980s. The Department of Energy (DoE) and the California Energy Commission installed a tower-based concentrated solar thermal power (CSTP) station in the Mojave Desert in which 1818 sun-tracking (heliostat) mirrors (total area 782,000 ft²) disposed circumferentially around the tower base concentrated the sun's energy to a receiver high up in the tower. Here it heated synthetic oil, a heat transfer fluid that conveyed heat down the tower to boilers on the ground where steam was produced, via a heat exchanger, and used to drive turbines connected to alternators. This facility was dubbed Solar One.

In 1995 Solar One became Solar Two when a second ring of 108 larger heliostats (total area 891,000 ft²) was added, bringing total power generation capacity to 10 MW. Although only a hundredth the capacity of a good-sized conventional power station (we can take 1 GW as a yardstick), this was useful scale and pointed the way ahead – even more so when the solar system was modified to use molten salt instead of oil and water.

The salt, a combination of sodium nitrate and potassium nitrate, constituted a high-temperature storage medium and enabled energy supply to continue through daytime interruptions to sunlight and into the night. Solar Two was decommissioned in 1999 after operating for five years. Bill Richardson, US Energy Secretary at the time, said the CSP system had demonstrated the ability of solar molten salt technology to provide long-term, cost-effective thermal energy storage for electricity generation.

More recently, Spain has adopted similar technology, while stepping up in scale. In the sun-drenched interior of southern Spain, a concrete tower rising 162 m from the plain near Seville is by day bathed in an ethereal white glow caused by sun rays reflected from surrounding arrays of sun-tracking mirrors. This solar power station, prosaically named PS20 and currently in start up phase, has a nominal capacity of 20 MW. When fully operational, it will be the world's largest commercial CSP plant feeding power into a national grid. Over 1200 heliostats, each with a 120 m² parabolic mirror, focus their rays onto a receiver/heat exchanger high up in the tower.

2^nd generation solar tower

Solucar Energia, responsible for the plant, says that the heliostat field/collector/steam cycle system used is a development of technology employed in an earlier CSTP tower, PS10, which is also operating commercially just outside Seville – while the new €80 million plant is a second-generation system incorporating improvements based on the earlier experience. PS10 has about half the number of similar heliostats which, focussed on a receiver located up a 115 m tower, enable about 11 MW to be generated via a steam turbine. PS10 has for two years been supplying enough power for several thousand premises in Seville. Solucar, part of the solar business unit of energy conglomerate Abengoa, proudly proclaims that the two plants between them can develop some 31 MW of peak solar power without emitting any greenhouse gas.

Daytime output is ideal for powering the air conditioners that keep the Seville townspeople cool in summer. At night air conditioning is not needed. Even so, power supply can continue for some time after sunset thanks to some of the steam generated during the day being stored in insulated tanks. Although water/steam heated to about 250ºC is less effective as a heat storage medium than a salt such as sodium or potassium nitrate that can be stored at 600ºC, Solucar opted for water for the sake of system simplicity and robustness.

PS10 and PS20 are part of a larger solar complex due for completion in 2013. Forthcoming Aznalcollar and Solnova 1 to 5 towers will move up the power curve still further at 20 and 50 MW respectively. However, the 'solarisation' of Seville is not wholly solar thermal based. Other elements of the scheme include Sevilla PV, claimed to be the world's largest low-concentration PV plant. Here, 1100 ft² of mirrors on 168 tracking devices direct sunlight onto some 373 ft² of PV surface, generating up to 1.2 MW. Company officials say this plant will produce around 2.4 GWh of solar power per year.

When finished, the Seville complex should be capable of generating more than 300 MW, providing power for some 180,000 homes – meeting most of the city's domestic needs. Almost a third of a gigawatt represents genuine utility scale and the complex should, moreover, prevent emissions of some 600,000 tonnes of CO₂ over its intended 25-year life.

Spain is roaring ahead with plans for additional CSP. The government has approved more than 50 projects and by 2015 total capacity should be over 2 GW. Torresol Energy alone intends to have 1 GW installed within a decade. Gemasolar, a 17 MW CSTP plant Torresol Energy is constructing in Fuentes de Andalucia, Seville, should start working in 2011. A molten salt thermal storage system will enable the plant to operate for at least 15 hours without sunlight. Torresol, a joint venture between Spanish engineering group SENER and Masdar of Abu Dhabi, announced at Abu Dhabi's World Future Energy Summit that it had secured the €171m needed to allow construction of the station to go ahead.

Revival

Meanwhile, the USA is revisiting the solar tower concepts it first came to grips with over a quarter of a century ago. This is considered worthwhile because of the growing conviction that the oil-based economy is unsustainable and the cost-reducing technologies that can now be applied to CSP.

An example of this evolution is the Luz Power Tower 550 system from BrightSource Energy Inc. LPT 550 is central to 7 projects totalling 1.3 GW recently agreed by Southern California Edison (SCE). A principal cost-saving feature is the use of thousands of small, flat mirrors that are simpler and cheaper to produce than the solar trough/parabolic mirrors used in previous systems. Efficiency is promoted by super-heating water, the heat transfer fluid used, to 550ºC so as to provide high pressure steam. To conserve precious water in the intended desert location, air is used to cool the exhaust steam back into water, which is then returned to the turbine in a closed-loop system.

BrightSource and SCE are targeting towers of 100 MW each, about 10 times the power of systems like Solar 1 or Spain‘s PS10. The BrightSource Energy technical team feels confident to do this on the basis of experience gained with units constructed at a solar energy development centre in the Negev Desert belonging to sister company BrightSource Industries (Israel) Ltd.

Assuming that the California Public Utilities Commission gives the green light, the first 100 MW units could be operating by 2013 at a site near Ivanpah, delivering some 286,000 MWh of electricity per year to Californians. The full 1300 MW of projects, when completed, should deliver enough clean energy to serve nearly 845,000 homes, avoiding more than 2m t of CO₂ emissions annually. Solar power towers will contribute about 900 MW of this. According to Stuart Hemphill, senior vice president renewable and alternative power at SCE, the scheme will make the utility America's largest purchaser of solar energy.

Alternative

Notwithstanding the high promise of 21^st century CSTP, these towers are not the only show in town. It is too early for any particular CSP technology to have become dominant and there is still plenty of room for innovation. Much will depend on the cost of the energy delivered – hence a constant drive to improve efficiency and reduce overall 'photon-to-electron' costs.

Several technologies are under consideration. For example, mirrors are not the only way to concentrate solar power. An alternative, familiar to small boys who enjoy setting fire to paper or other combustible material with magnifying glasses, is to use lenses. International Automated Systems Inc of Utah has trialled thin-film lenses in sun-tracking arrays. IAS says its lens panels are efficient and inexpensive compared with mirror-based solar dishes and troughs. It claims they need little maintenance, unlike other technologies that (it says) require frequent adjustment to maintain focus. The company has also developed a bladeless turbine that operates on rocket reaction principles rather than a flow of gas impacting on blades. International Automated Systems Inc claims enhanced reliability and reduced cost for its breakthrough.

An alternative to both conventional mirrors and lenses is an arrangement of segmented flat-plate reflectors configured to approximate to a single parabolic reflector. This Fresnel-like arrangement, though less efficient, is also less expensive to produce while the close arrangement of mirrors requires less land than separated heliostats.

Other technological possibilities include generating steam directly in the receiver tubes rather than via a synthetic oil heat transfer fluid or, looking further ahead, replacing the steam turbine cycle entirely by heating air and using the hot pressurised air to drive a gas turbine. The European SOLGATE project established the feasibility of this approach, sending air heated to 800ºC to the combustion chamber of a gas turbine.

Small gas turbines could feature in parabolic dish concentrators, whose operation can be likened to a vehicle headlight operating in reverse. Alternatively a Stirling engine can be used, as was done in the 10 kW EURODISH project. Although potentially quite efficient, such parabolic dish engines would be unlikely to achieve utility-scale power – a 15 m diameter dish, for example, would typically have a power capacity of about 25 kW. It would, therefore, be best suited to decentralised stand-alone systems. Another idea that has been tried is the solar chimney, in which air trapped in an inexpensive 'greenhouse' structure around the chimney base is heated by the sun and drawn up through the chimney, driving a turbine on the way.

Troughs

None of these ideas has yet become mainstream. One that has, however, relies on the fact that it is not necessary to focus to a single point. The solar tower and heliostat approach is challenged by an alternative that uses parabolic trough reflectors to focus energy onto linear receiver tubes containing the energy storage fluid. These trough/receiver units, which may be arranged to track the sun in two dimensions, can be installed on the ground. Power is needed to pump fluid round the tubes but, on the other hand, there is no need to pump uphill into a tower.

A notable installation of this sort is Nevada Solar One, which began operating in June 2007 from a 250-acre site in the Nevada Desert, just outside Boulder City, Arizona. At 64 MW nominal, this station is well up the power scale by contemporary standards and can generate around 134 GWh per year. Spanish and American interests came together to realise this project, the plant having been built by Acciona Solar Power Inc (formerly Solargenix Energy Inc), a part-subsidiary of Spain's Acciona Energy, in collaboration with USA's DoE and the National Renewable Energy Laboratory (NREL).

The plant is 98% CSP and 2% natural gas. Long arrays of parabolic trough reflectors focus the sun's energy onto more than 18,000 four metre-long receiver tubes. Sun-tracking motion is controlled by a system from Parker Hannifin. Schott Glass and Solel Solar Systems provided the efficient receiver tubes, which heat oil inside them to over 300ºC. The hot oil is used to boil water into steam, which drives a 75 MW Siemens reheat steam turbine. The 2% natural gas component is used to heat the oil transfer fluid when incident sunlight is interrupted.

A sure sign that concentrated solar can reach a utility scale approaching that of some conventional mainstream power stations is a 553 MW complex under construction in California for the Pacific Gas and Electric Company. Mojave Solar Park is a trough-based facility that will occupy 9 square miles of the Mojave Desert and is expected to become operational in 2011. Some 1.2 million mirrors and 317 miles of vacuum tubing will be used to capture the Mojave sun‘s heat. Israel's Solel Solar Systems, with its US arm Solel Inc, is responsible for the facility, which Solel says will supply 1388 GWh of energy annually at a price competitive with plants powered by fossil fuels.

Actually, a large solar trough-based facility already exists in the Mojave Desert. A 345 MW solar electric generating station (SEGS) was built in the 1980s by Israeli company Luz International Ltd, forerunner of BrightSource Industries (Israel) Ltd mentioned above.

Back in Europe, Spain's Torresol is working on TERMESOL and ARCOSOL, two 50 MW parabolic trough-based plants located in Seville and Cadiz. Other Spanish plants using this technology are Andasols 1,2 and 3 in Granada Province, with a combined capacity of some 150 MW. Andasol 1, developed by Solar Millenium AG, is Europe's first utility-scale parabolic trough power plant and went on line last November. Numbers 2 and 3 are under construction and a further four Andasols are planned.

Nor are conventional silicon-based solar PV arrays out of the picture. An enormous solar farm due to start generating in 2010 in Victoria state, Australia, will develop 154 MW, providing enough clean energy for 45,000 homes. Although most of the facility will be PV array-based, using high-performance space-grade solar cells, the A$420m development will also feature 2 MW solar PV power tower technology. Australia's Solar Systems and TRUenergy are the builders of this solar PV power station, expected to be the world‘s largest.

Sahara power

Deserts, with their high insolation and low habitation levels, are ideal places for solar power stations whether conventional PV array, solar thermal trough, CTSP or solar PV towers. In principle less than 1% of the globe's deserts could generate as much electricity as the world currently consumes. And there are deserts within 2500 km of 90% of the world's population.

These realities have prompted the serious suggestion that electricity generated by the North African sun could meet a substantial proportion of Europe's domestic electricity needs, if cabled across the Mediterranean by a grid network. In December 2007 Prince Hassan bin Talal of Jordan outlined to the European Parliament a bold vision for a string of giant solar power stations located along the Mediterranean desert coasts of northern Africa and the Middle East. More than 100 heliostat-based stations would generate billions of watts of solar power to meet power supply requirements in European countries and drive desalinisation plants locally.

The scheme, Desertec, is sponsored by the Trans-Mediterranean Renewable Energy Corporation and has attracted support from politicians and engineers throughout Europe, the Middle East and North Africa. Co-founder of the scheme, physicist Gerhard Knies, points out that most solar radiation beating down on deserts is wasted, serving only to heat the ground during the day so that it can be radiated back into the atmosphere by night. As he asserts, “we need to stop that waste and exploit the vast amounts of energy the sun beams down on us.”

Proponents like Knies believe stations built to date indicate that individual solar installations capable of generating 100 MW of power should be viable, and see no reason why up to 100 billion watts of power should not be generated by numbers of these in solar farms. Long-distance power transmission would rely on an evolving high-voltage DC grid, with AC transmission being retained for in-country national grids.

Tantalising prospects

Solar radiation has long tantalised humanity with its ubiquity and power – enough to satisfy mankind's power needs many times over. The solar PV sector first showed that this renewable energy can be turned into useful amounts of electricity, the most convenient medium of power delivery, though reliance on semiconductor materials like silicon and gallium make this method expensive at utility scale. The great promise offered by the alternative CSTP approach is evident from the current revival in the technology. Power generation is pollution-free, while CSTP is basically fixed-cost once the initial investment has been covered, offering budgetary predictability through long-term supply contracts. Given energy storage media such as molten salts or heated concrete, CSTP could meet base load requirements, as well as intermediate and peaking loads. From the point of view of a grid operator, CSTP behaves like any conventional steam cycle power station, an important factor for grid stability. Being by nature distributed rather than centralised, a solar power infrastructure is less vulnerable to hostile actions than present arrangements.

Currently, though, CSTP is two to three times more expensive than power conventionally generated – at least under normal accounting conventions. But, as the influential Stern Report pointed out, normal accounting conventions do not allow for external costs associated with fossil fuels. It is hard to allocate figures for ill health caused by pollution, environmental degradation, fuel security issues and other 'intangibles', but these 'externalities' are real and putting a value on their avoidance would level the playing field. Moreover, costs for CSTP, currently around 9-12 US$c/KWh, are predicted to fall to about 6US$c/KWh over the next decade, thanks to scale and improved technology. Given this prospect, CSTP stations would soon be competitive with fossil fuel generation, even under normal accounting practice.

Much therefore, will depend on scale and, as we have indicated, this is on a definite rising trend. With operational CSTP installations ranging from about 10 MW (eg Spain's PS10 at 11 MW) through 20 MW for PS20 to 64 MW for Solar Nevada One, and with BrightSource Energy targeting 100 MW towers for Southern California Edison, the trend is clear. Combine a number of tower or trough installations together in a solar farm and you move up another gear still – witness 300 MW for the Seville complex and 553 MW for Mojave Solar Park.

Half-gigawatt solar power stations are respectable utility scale and, given their imminence, there seems to be little reason in principle why a 1 GW station should not be produced. Solar would then be right up there among the big hitters of power generation. At the present rate of progress, the first such station might not be long delayed. If the Desertec string of 100 MW-plus solar pwoer stations proposed for the Mediterranean border were to be built, the result could be considered a well-spread linear solar farm, potentially delivering billions of kilowatt hours. That would be scale indeed, indisputably putting solar among the big hitters.