Solar, Utility Scale, Wind Power

Management of growth: Concentrating Photovoltaics

Issue 1 and Volume 11.

Along with wind power, argues Ralf Leutz, concentrating photovoltaics (CPV) offers one of the most promising utility-scale, and sustainable, renewable energy technology options. CPV has the potential of an efficiency reaching and exceeding 50%. And in the right locations, will have electricity costs low enough to be competitive with fossil and nuclear power.

Once upon a time, concentrating photovoltaic (CPV) technology was a purely academic discipline. Sandal- clad people wrestled with mirror monstrosities through two oil crises, firmly in the shadow of solar thermal and always belittled by Big Oil.

That was until the space industry became interested.They had developed multi-junction devices capable of matching band-gaps and lattice constants of advanced semiconductor materials to the spectrum of the sun. Deep Space 1, a NASA probe launched in 1998 on a path extending past Mars to examine our star, was the first spacecraft to have CPV modules on board.These were linear Fresnel lens concentrators made of silicone under a superstrate of glass. The reason for using concentration was to supply sufficient flux at a distance from the sun where natural sunlight reaches only a quarter of the 1365 W/m2 power density found near Earth.

In those days, pioneering company Entech was more successful with its space designs than with the terrestrial variety. Mark J. O’Neill, Entech’s founder, must be credited with a number of the inventions still advancing the field including solar Fresnel lenses, inflatable lenses, stretch lenses, and prismatic covers for redirecting light away from the cell grid.

Then, the budding CPV technology was crushed by the return of cheap oil and political negligence. Solar was sidelined. Those were the days when photovoltaic panels brought the voice of the Maximo Leader into the Sierra Maestra, when solar panels appeared at police posts in the Caprivi Strip, or in Himalayan monasteries.The power of solar photovoltaic installations was measured in Watts. Meanwhile, concentrating solar thermal power had already installed 354 MW in the Mojave Desert in California. For a while, the future for large-scale solar looked like thermal – then it, too, ground to a halt.

All the same, the concept of providing premium power at a premium price, and at utility-scale, had proved feasible.Those linear troughs in the Mojave are still running, and provide peak power to Los Angeles, making money on that city’s air- conditioning habits.Thermodynamically speaking, however, it isn’t very efficient to concentrate sunlight in order to heat oil in order to generate electricity in a Rankine cycle, just to run a compressor in a reverse Rankine cycle in order to reduce the heat of large quantities of air.

AN ENERGY CONTEXT – IN NUMBERS

* The installed power generation capacity throughout the world is currently about 2 TW (www.ipcc.ch). * It is estimated that growth of energy consumption and replacement of power plants older than 20 – 30 years will require 6 TW to be built in the years to 2030.The cost associated is US$5.2 trillion.Welcome to the world of Tera Dollars! * In China alone, every week sees the commissioning of 1 GW of generation plant, mostly coal-fired. * Globally, photovoltaics as a whole approaches 10 GW installed, of which concentrating technologies account for one tenth of a percent. CPV installations must increase by several orders of magnitude in order to off set additional fossil-fuel based capacity. * Four to five orders of magnitude is daring for a maturing technology, which is only now seeing demonstration plants on the megawatt scale being built in Spain (primarily at the ISFOC – Instituto de Sistemas Fotovoltaicos de Concentracion, in Castilla-La Mancha) and in the USA in Arizona. * A trillion is a figure with twelve zeros. Numbers in excess of a trillion are either called zillions by the uninhibited, or quoted in scientific nomenclature, where giga (G, 109), tera (T, 1012), peta (P, 1015), exa (E, 1018), zeta (Z, 1021), and yotta (Y, 1024) are used as prefixes. * Sunlight reaches the earth with a power of 0.2 x 1018 W. Over the course of the seconds of the year, this adds up to 5 yotta Joule (YJ). Global electric energy production is currently 0.2 ZJ, while total global demand for energy is around 0.5 ZJ.

* Thus, assuming an overall efficiency of 10%, solar supply is a thousand times higher than global energy demand.That includes issues of demand curves, day and night, and differing technology.


COMPETITIVE ELECTRICITY COSTS

Energy as a commodity is as vital to the world as food, water, health or sanitation. Energy is also big business, and competitive. In the right place, the cost of CPV will be competitive with fossil and nuclear power.

CPV has the promise of an efficiency reaching and exceeding 50%. Its inherent complexity of concentrating optics and tracking contributes to its high-tech appeal.At good solar locations, CPV will have levelized electricity (LEC) costs low enough to be competitive with conventional fossil and nuclear power. LEC is a price tag for electric energy, covering investment and depreciation, operation, and performance, over the lifetime of the generator. Note that LEC refers to energy rather than to power. The difference between energy and power is time.The integration of power yields energy. For CPV, the main parameters LEC is sensitive to are direct solar irradiance, efficiency, and interest rate.The initial investment, the often cited $/W installed is in fact marginal for LEC calculated over 20 years of energy generation. Hence the difference in installation costs of $3/W for CPV and installation costs of $1/W for thin-film will look much less pronounced in terms of LEC, especially when land prices are considered.

Solar technology and efficiency will match latitude and application. While building integrated PV (BIPV) in higher latitudes will most likely be thin-film, power plants in the sun belt of 20*-35* will be more likely to use CPV or solar thermal power plants.

Large companies have understood the complementary nature of solar power conversion systems and are moving away from their former regime of exclusively marketing a single technology. This is the case for Sharp, which had been slow adopting concentrating technology, even though the company was a leading participant in a Japanese government-sponsored CPV research programme which left Daido Steel and Kenji Araki with dome-shaped, nonimaging Fresnel lenses. Another example is Emcore Corp., one of the three leading manufacturers of multijunction silicon devices. As of summer 2007, Emcore not only markets PV cells, but has also designed a CPV module. Current orders and options for cells have exceeded 100 MW, while those for modules are reaching 20 MW.The joint venture partner for the modules is the South Korean packaging company DI Semicon.

TECHNOLOGIES – INNOVATION AND DEMONSTRATION

The list of participants at the Institute of Concentration Photovoltaics Systems (ISFOC) in Castilla La Mancha, Spain, reads like the Who’s Who in CPV companies, with most of the major players apparently setting up demonstration plants there. ISFOC is currently constructing a 3 MW CPV installation with participation from a number of these companies, such as SolFocus.The Californian company, working with more at least $77 million venture capital, is one company working with reflective mirror-based Cassegrain systems. The technology uses designs by the world’s best known scientists in the field of nonimaging optics. SolFocus has now designed, manufactured and installed the first 200 kW of the 500 kW set to be installed over the next few months. Recently,Arima Eco, a Taiwanese computer manufacturer applied for, and was awarded, a slot at ISFOC.

THE EFFICIENCY CHALLENGE

Many producers selected for ISFOC use a variety of refractive optics. There are Fresnel lenses, or combinations of Fresnel lenses and glass, or mirror-based secondary optics. In the case of Isofoton, the Fresnel lens is a totally internal reflecting (TIR) lens.The aspect ratio, defined as ratio of focal length over lens diameter, of this module is smallest in comparison with the others. Fresnel lenses can be optically fast, with aspect ratios smaller than unity. A fast system has generally positive implications on acceptance half-angle and tracking error sensitivity, but dispersion and optical efficiency are usually affected negatively.

The acceptance half-angle theta (*) of concentrating systems is connected to the geometrical concentration ratio (C) of collectors by the relation C = 1/sin *.The half-angle of the solar disk at 0.28* sets the limit of concentration. The geometrical concentration ratio is a theoretical value. In practice, the optical efficiency of a concentrator must be considered.The product of optical efficiency and geometrical concentration ratio will result in the optical concentration ratio. The challenge in solar concentration is to reach the maximum optical efficiency. In reflective systems, the local surface slope error doubles the specular error of the reflected ray. This is particularly critical in Cassegrain optics, where primary, secondary, and sometimes tertiary optics must be aligned. In refractive systems, the limiting factor of concentration is dispersion.

Modern receivers are multi-junction devices where three or more cells are stacked.The conversion efficiency of the top cell is limited by the intensity of the incident ultra-violet radiation, and the Ohmic losses caused by the local hetrogenity of the irradiance and the top cell in turn limits the performance of the cell stack. Total device conversion efficiency also depends on the differences in current generation between cells, a function of the spectral irradiance distributions. The more complex the multi- junction device, the higher are the demands on spectral and local uniformities. This is the reason why most dispersive systems use secondary optics, for example kaleidoscope homogenizers.

Future generations of multi-junction devices will have to be adjusted not only for the dispersion of novel highly UV- transmissive optical plastics, but also for the spectral reflectance of advanced mirror materials. Both advanced plastics and highly reflective broadband aluminium mirrors are under development. While new materials will help advancing the frontiers in efficiency and optical concentration, the current challenges for CPV module producers are manufacturing issues. Concentrator elements measure in decimetres, lenses and mirrors alike are 10 cm-40 cm in diameter, in order to supply light at a concentration ratio of 400 to 750 suns onto cells measuring millimetres across. The problem is to mass manufacture large optical elements in optical quality and at reasonable cost.

MANUFACTURING BOTTLENECKS

The emerging CPV industry is currently experiencing supply chain issues where not everything it needs is readily available. The traditional optical industry has been focusing on imaging optics, using manufacturing processes not suitable for supplying large areas at low cost.The photonics industry has pushed the use of plastics, and highly reflective materials, of coatings and filters – among other elements – but issues remain concerning large area production, and delivering high spectral transmissivity, not to mention a 25-year lifespan in the great outdoors. Thus, we are witnessing the development of a supply industry dealing with the bottlenecks in CPV.

However, the bottleneck is not the cells. The three manufacturers are increasing their output annually fivefold to tenfold. Big names such as Spectrolab (Boeing), Emcore, and Azur Solar will be joined by several other companies in the near future. After all, every major university has a clean room with reactors where semiconductors can be grown. Cell processing and commercialization are the next steps.

The bottlenecks will be in the supply of lenses and mirrors, where vast areas, square kilometres initially, many thousands of square kilometres possibly, must be covered in optics of excellent quality. Once you accept that the plastics industry will have to multiply its production capacity, the point is understood.

STANDARDS AND SIMULATION

Flat plate PV has embraced the principle of selling lifetime kilowatt hours, as opposed to selling modules rated at a certain power.Two parameters need to be known in order to be able to guarantee a 20-year lifetime of energy generation: solar irradiation, and module performance.The industry is confident about module performance and longevity. While there might be some uncertainty due to the effects of climate change on solar irradiance at a given location, the effects of module price and installation cost are small when compared to the effects of module efficiency on LEC over twenty years.An exception to this rule is the cost of investment, since the interest rate can easily double investment costs.

CPV companies have adopted a lifetime production guarantee for their modules. Sol3g guarantees 80% of initial performance after 20 years, equivalent to an annual efficiency reduction of 1%, if a proper maintenance programme is followed.Amonix, Guascor Foton, and Green and Gold Energy are among other companies offering similar warranties.

The CPV industry, convened by Bob McConnell of NREL, has also just passed a performance standard, the IEC 62108, entitled: ‘Concentrator photovoltaic (CPV) modules and assemblies – Design qualification and type approval’ (which can be viewed at www.iec.ch). Performance rating, tracking, and safety standards are next.The performance rating aims to introduce a power rating and later add an energy rating for CPV. The challenges are inherent in the concentrating technology. The stationary concentrators, in particular, resist any simple tests. Furthermore, in contrast to flat plate collectors, concentrators may have complex optics, making testing more difficult. Their optical efficiency is usually significantly dependent on the angle of incidence.

Any complete characterization of stationary concentrators must be based on outdoor testing of at least six months, between solstices. Measurements may only be reduced to this period, if solar radiation values are known for the rest of the year, and can be correlated with detailed measurements from similar sites for the estimation of direct and diffuse irradiances for longer periods.

Moreover, many concentrator optics and receivers are sensitive to spectral changes of irradiance.These are difficult to model in data sets and complicated to incorporate in simulations. In addition to the vital outdoor testing, two avenues have opened to further the understanding of lifetime concentrator performance. These are computer simulations and indoor physical testing.

Computer simulations have reached a level, at which multi- junction cell performance based on incidence parameters can be accurately predicted. Nonetheless, these models are not applicable to general problems, and have yet to be fully accepted by the potential examiners of CPV systems. Optical ray-tracing has come a long way, but the simulation of some surface properties, for example those caused by common manufacturing processes, and the spectral ray-tracing of some highly complex optics, remain a fairly elusive goal. Even so, in the lab, we find that conservative ray-tracing, guided by experience, coincides very well with test results.

CONCENTRATORS – THE NEED FOR A SOLAR SIMULATOR

Standards cannot, however, be based on what is possible in a scientific laboratory. Mass manufacturing of CPV modules requires in-line performance testing of each and every module. Only if tested in this manner can modules be rated, and rated modules be matched, sold, and erected within a power plant of expected lifetime energy production.

This necessitates the use of a solar simulator for concentrators. This simulator needs to produce uniform light at the right spectral distribution, at the right power density, and within a collimation angle similar to the half-angle of the sun. The latter is a serious physical challenge. Only a simulator and rigorous quality assurance can make CPV the giant in energy production it deserves to be.

The introduction of standards will have a positive effect on the technologies used in the manufacturing of CPV modules. While any standard is the smallest common denominator of the views of the participating parties, standards offer the tools to rate the performance of technologies. The current emerging market for CPV is going through a phase in which designs are numerous and often novel. This is laudable, and absolutely necessary. Performance standards will in the long-run cause inefficient designs to drop out of the market. One of the purposes of standards is to make the market more transparent both for users and manufacturers.

A FUTURE TECHNOLOGY EMERGES

Just a handful of CPV designs will come to dominate the market. Large customers will force the industry to produce modules with the technology suited to their location, and their understanding. Cultures based on technological approaches are evolving.There won’t be a clash of cultures, but there will be a detailed comparison of efficiency values, measured over lifetime, with a testing and performance standard.The rating of CPV systems will be in $/kWh LEC.

Oil is at $100/barrel, climate change is here, the scientific and political stage for the success of solar energy is set. Markets are huge. Management of growth is the key. Once demonstration plants run according to plans and expectations, just two years should show that CPV does indeed deliver.Twenty years on, with tera Watts installed, CPV will be part of a tera euro market.This is the promise, and this is the challenge.

Dr Ralf Leutz is a former lecturer at Marburg University. He is cofounder of Concentrator Optics (www.concentratoroptics.com), a company focusing on the design, prototyping and manufacturing of Fresnel lenses for CPV. Concentrator Optics plans to produce a solar simulator for concentrator modules. Ralf recently organized and hosted the International Workshop on Concentrating Photovoltaic Optics and Power (www.concentrating-pv.org). e-mail:[email protected]


CONCENTRATING PV – THE BASIC PRINCIPLE

The principle of concentrating PV (CPV) is quite point. When the teeth run in straight rows, the lenses act as straightforward. In the familiar ‘flat-plate’ PV modules, a large line-focusing concentrators.) The concentration ratio can vary: area of photovoltaic material (usually crystalline silicon) is if the light that falls on 100 cm2 is focused onto 1 cm2 of PV exposed to the maximum naturally occurring sunlight. Normally, material, the ratio is considered as 100 suns. If the light from that maximum is achieved by installing the modules at an 10 cm2 is focused onto 1 cm2, the ratio is 10 suns. If the incline optimized for the latitude, but sometimes they are concentrated sunlight light falls onto a well designed CPV cell, installed on moving frames that can follow, or track, the sun as the cell will produce at least 100 times, or 10 times, the it passes across the sky. The PV cells perform under direct electricity. In fact, the conversion efficiency of cells increases (sunny) or diffuse (cloudy) radiation conditions, but output is at under concentrated light, so the correlation is likely to be its highest when the maximum amount of light falls on the cells greater than one-to-one, depending on the design of the solar (assuming there are no detrimental effects from overheating). cell and the material used to make it. While commercial The amount of light that falls on a cloudless day (this varies concentration ratios are around 200 to 300 suns, as much as according to location and season) is regarded as one ‘sun’, 1000 suns is expected for future concentrating PV systems. which is defined as 1000 W/m2. As most CPV systems use only direct solar radiation, these Concentrating PV systems use lenses or mirrors to focus installations almost always involve trackers – rotating about sunlight onto a small amount of photovoltaic material. (Usually either one or two axes – to keep the sun focused on the solar a Fresnel lens is used, a flat lens that uses a miniature cell. This means that CPV is best suited to regions with high sawtooth design to focus incoming light. When the teeth are levels of direct, rather than diffuse, solar radiation, particularly arranged in concentric circles, light is focused at a central the sunbelt regions. -JJ


ROLL OUT FOR NEW LOW-COST CPV DESIGN ANNOUNCED

SolarOr, a new start up company with facilities in Israel, says it has developed a novel concentrating PV solar panel based on a patented passive tracking optical design. This enables cost reductions by using around 40% of the silicon in comparable PV cells, instead using lower cost optical elements, SolarOr says. The company adds that its designs are inherently lower cost, retailing at about 50% of standard, commercially available solar panels with similar output power. This low cost is expected to continue to scale downwards with further reductions in Si PV cell cost, as additional supply is brought on line over the next 24 months. The translucent panel design enables a new and unique ‘Architectural PV Glass’ to be used for deployment as skylights and glass walls in residential and commercial structures. Such panels generate electrical power, while enabling daylight inside buildings, and certain models also have the option of incorporating pipes for heating water. In addition, this feature will also remove heat from the PV cells, further increasing their power generation efficiency. In future the product portfolio is expected to include solar panels using standard silicon PV cells, with an initial price projected at $1.5/W, and using multi-layer, high-efficiency silicon PV Cells for higher output versions. -DA