Solar furnaces: concentrating 100,000 suns.
In a series of experiments conducted in January 1990, a University of Chicago researcher under the direction of physics professor Roland Winston achieved a record concentration of 84,000 suns, far in excess of the most intense artificial broadband continuous light sources that have yet been devised. Further refinements of the techniques designed by Winston's group are expected to yield a solar furnace design that can produce concentrations of more than 100,000 suns.
The many methods of concentrating sunlight can be grouped into two general categories: imaging and nonimaging, depending on whether the optics used are designed to form an image of the object (the sun). Concentrators using imaging optics are optimized to map all rays from each point on the object to the corresponding point on the image and to preserve the geometric relationship with other points. They operate best for rays that are close to the axis; in the process, a portion of the angular field of view is sacrificed. This loss must be compensated for by tracking the collector to follow the sun. Since an image of the sun is not needed for concentration, imaging methods that focus the sunlight set unnecessarily high standards. All that is necessary is for all the collected rays to fall within as small an area as possible on the absorber or target surface.
Nonimaging optical devices, on the other hand, have a much broader angular acceptance and are designed for rays from the extreme angles of the optical system. Nonimaging concentrators, with either an axis or plane of symmetry or sometimes no symmetry at all, operate by taking all rays entering at extreme input angles and guaranteeing that they emerge just grazing the edge of the exit aperture or absorber surface. Although the relative mapping of light rays from the object will in general not be preserved--and therefore the concentrator will not produce an image--all rays falling between the extreme angles will also reach the desired smaller target.
The most efficient nonimaging concentrator for a given desired angular field of view has a reflecting surface that makes all rays from the extreme input angle fall, after at most one reflection, on the edge of the absorber. This "edge ray" principle, together with the laws of refraction and reflection, is the basis for the design of all nonimaging concentrators.
Many of the light collectors designed according to nonimaging principles are referred to as Winston concentrators or compound parabolic concentrators (CPC), after the simplest reflector shape originally designed by the University of Chicago's Winston for both conventional solar thermal and photovoltaic applications. The term CPC refers not simply to one specific collector. but to a whole family of concentrator designs that employ nonimaging optics in a trough geometry to provide moderate concentration with a wide enough field of view to accommodate the sun's motion without tracking. Some versions do require seasonal adjustments; however, the most highly developed form corresponds to a design geometry that can remain completely stationary throughout the year. The most widely known version consists of a CPC-shaped sheet of reflective aluminum coupled to a thermally efficient evacuated tube receiver that looks like a long thermos bottle. In collaboration with Roland Winston, this design was developed for commercialization in the late 1970s by the Argonne National Laboratory solar energy group directed by William Schertz, now director of energy and environmental technology research programs at the laboratory. A newer experimental version, developed at the University of Chicago in collaboration with GTE Laboratories (Waltham, Mass.), has both the receiver and reflector combined and integrated together in a long evacuated glass tube with the outer glass envelope of the evacuated tube shaped into the CPC profile. The shaped portion is coated with highly reflective silver so that each module looks a little like a long fluorescent glass tube with one transparent surface so that sunlight can enter. The optical efficiency of such an integrated CPC is significantly improved relative to the earlier commercial version.
Nonimaging optics is extremely forgiving. It provides different order-of-tolerance requirements from what is usually required with imaging optics. In the nontracking concentrator designs, for example, slope and contour errors of several degrees are not even noticed, whereas in the usual tracking parabolic trough the surface must be optically true (and the drive mechanism tracked) to within a small fraction of a degree.
The use of CPCs for solar thermal collection developed rapidly in the late 1970s at Argonne National Laboratory and the University of Chicago. More recently, however, the Chicago group, with continuing support from the Solar Energy Research Institute (SERI) in Golden, Colo., has been developing a more-advanced two-stage concept. This configuration employs a primary imaging stage as well as a CPC device in the focal plane. In principle, a single nonimaging optical element could be designed that would attain very high concentrations and approach the theoretical limit of concentration. However, even for moderately high concentrations ([is greater than]100 or so), such a single element would become very large and unwieldy. A more practical design employs a nonimaging secondary concentrator at the focal plane of the primary imaging element, where the secondary can provide an additional concentration factor so that the concentration of the combined system approaches the allowed limit for the appropriate system design acceptance angle. "A two-stage concentrator works best if you have a moderately long focal ratio design, that is, where the focal length is substantially greater than the diameter of the aperture of the primary of the focusing mirror," said Joe O'Gallagher, a senior research associate and senior lecturer in the Physics Department at the University of Chicago and a colleague of Winston's since 1976.
Reflecting nonimaging secondary concentrators can be of either the CPC type or the so-called trumpet type, discovered by Winston and the late Walter Welford of the Imperial College in London. The trumpet type can be retrofitted to augment the concentration of more conventional focusing imaging elements used as primaries and have advantages for many solar applications. They are used mainly to increase thermal efficiency by increasing concentration or to relax optical tolerances and thus reduce the cost of the primary concentrator.
Solid dielectric (n[is greater than]l) CPCs were originally developed in collaboration with Argonne National Laboratory for seasonally adjusted photovoltaic concentrator arrays. More recently, at Chicago they have been developed as secondary concentrators for photovoltaic cells to be used in combination with lens primaries. In this application, they increase angular tracking, alignment, and slope tolerance requirements for the primaries and reduce system cost and complexity. In a recently developed two-stage photovoltaic concentrator, a good primary with an acceptance angle wide enough to accommodate the sun's movement for about one hour concentrates the light 13 times, while a totally internally reflecting secondary provides the remaining concentration factor needed to reach 100 suns and makes the savings associated with reduced cell area worthwhile. The key parameters that determine the shape of the secondary are its entrance diameter, the convergence angle of the light at the focus of the primary, and the refractive index of the material from which the secondary is fabricated.
Again, it turns out that nonimaging optics is very tolerant. For example, even in many demanding high-flux experiments, the concentrator can tolerate surfaces that have an angular deviation of about 1/4 degree; a lens that was 1/4 degree off would be terrible. But the most exciting application is in the new super solar concentrators, which pursue the very high concentrations that otherwise cannot be attained.
"A few years ago," said Winston, "I decided that it would be fun to try pushing the concentration limit to its maximum, to see if we could reproduce the surface brightness of the sun." Winston worked on the design for several years with O'Gallagher and Philip Gleckman, then a graduate student at the university. In campus rooftop experiments in February 1988, Winston's group, using a telescope parabolic mirror as a primary and an oil-filled nonimaging secondary, achieved an irradiance value of 4.4 kilowatts per square centimeter at an isolation of 800 watts per square meter, which corresponds to a concentration of 56,000 suns, or 68 percent of the radiance of the solar surface itself. This was almost three times greater than the previous record set at the one-megawatt paraboloidal furnace in Odeillo, France. Moreover, unlike the peak concentrated light produced at the French facility, which is a small portion of the energy collected lying at the very center of a highly nonuniform distribution, the University of Chicago group's record concentration is very efficient, using almost all of the light gathered by the primary, and is much more uniform. This concentrator, if deployed above the earth's atmosphere, would produce a light brighter than the sun's surface.
"In the design we used," explained O'Gallagher, "the focal ratio was 2.5 to make most effective use of the two-stage configuration. Basically, it was a small concentrator that was used to demonstrate the principles." The design is currently being scaled up in size in collaboration with researchers at SERI.
Certain refinements have already improved that record flux significantly. For example, the refractive index of the oil in the secondary decreases as it warms up, so the secondary's shape was no longer a perfect match to the edge ray condition for which it was designed. By replacing the oil-filled secondary with an all dielectric (sapphire) secondary, the drop in refractive index was avoided and the secondary performed for the refractive index for which it was designed. More importantly, the index of refraction of sapphire is 1.76 and the additional factor of n increased the allowed limit substantially. In January 1990, in a series of experiments carried out by Dave Cooke, a graduate student of Winston's, the Chicago group achieved a concentration of 84,000 suns. Even on the earth's surface, this corresponds to an irradiance substantially in excess of that on the surface of the sun.
Conventional solar thermal electric systems do not need these ultrahigh concentrations. Nonetheless, the ideas conceived by Roland Winston, Joe O'Gallagher, and their collaborators at the University of Chicago generated considerable interest at SERI, which is the nation's primary national laboratory studying and developing technologies for applying solar energy. When the ideas of Winston's group first became known, they coincided with SERI's need to develop a solar furnace for use in experiments on a number of new ideas for industrial applications requiring high concentrations, including materials processing, materials surface engineering, and photochemistry.
Al Lewandowski, the SERI engineer on the solar furnace project, began discussions with the University of Chicago group on the possibilities that their concepts opened up. As a result, Lewandowski developed a design that enables a flexible use of both the more conventional optics and the new nonimaging concepts in a high-flux solar furnace (HFSF). The SERI HFSF is located on top of South Table Mountain near Denver, Colo., and has a nominal power of about 10 kilowatts. In the HFSF, a flat heliostat tracks the sun and reflects the incoming solar flux onto the primary concentrator. The primary concentrator, made of 23 individual facets, concentrates the incoming beam onto the target plane located inside the test building. Once on the target plane, the solar beam, now concentrated to a peak of 2500 times its normal density, enters the nonimaging secondary concentrator. "Two unique features of the HFSF make this two-stage concentration possible," said Lewandowski. "One is the so-called off-axis design. The second closely linked feature is the long focal length of the primary concentrator. The latter, particularly, enables the beam reaching the target to be narrow and to couple with the secondary efficiently." In December 1989, SERI's HFSF became operational using conventional optics and attained a concentration of 2500 suns. In September 1990, using a nonimaging compound parabolic concentrator designed for SERI by Joe O'Gallagher and Dan Sagie at the University of Chicago, the furnace established a new world record for solar concentration in air, attaining an average concentration of 21,000 suns on a 1.5-centimeter-diameter target. A ray trace computer code calculated the peak flux to have exceeded 25,000 suns. The secondary concentrator has a compound parabolic shape, with an aperture of 5.9 centimeters in diameter and an exit of 1.5 centimeters in diameter.
SERI's current plans are to fabricate a secondary concentrator with a refractive index of approximately 1.5 and to use it to attain peak flux densities of over 50,000 suns in 1991.
"The significance of demonstrating that these very high solar flux densities can be attained at considerable power levels is that they make new process conditions, and therefore new industrial applications for solar energy, possible," said Meir Carasso, who manages the advanced industrial applications program at SERI. These new process conditions include very high temperatures, very high heating rates, and very high flux densities over the broadband solar spectrum of about 300 to 2500 nanometers.
In the area of materials processing, several early experiments made use of the ability of concentrated solar flux to produce very high temperatures. One process combines silica and carbon in a high-temperature reaction to form silicon carbide. A number of tests were performed in which the product was formed inside a graphite receiver/reactor, where internal cavity temperatures of 1750 [degrees] to 1800 [degrees] C were measured. SERI verified the success of the experiments via chemical analysis, which indicated that the samples consisted of 80 percent silicon carbide.
Another promising area of application is in materials surface engineering, a SERI project headed by Roland Pitts. This area exploits the ability of the concentrated flux to impose a high heating rate on the target. Recently, thin films of diamond-like carbon (DLC) have been grown on nickel and silicon substrates using methane and hydrogen as precursors. Experiments with silicon substrates have produced thin films of silicon carbide and DLC films, as well as graphitic carbon, depending upon the growth conditions. Current conventional technology using microwave plasmas or hot filaments requires large amounts of energy to grow similar films.
An important use for DLC thin films is in electronic materials. Thin coatings of DLC on heat sink materials provide electrical insulation properties approaching that of the commonly used silicon dioxide, combined with a thermal conductivity better than copper. SERI anticipates that large areas of substrate materials may be prepared before the electronic devices are built up on the material. Th's appears to be a particularly good application for two reasons: the solar furnace's ability to process relatively large surface areas and a batch-processing mode of operation lends itself to the daytime operation of the furnace. Another use for diamond films is in the area of low-friction (tribological) coatings. Because of their extreme hardness, low coefficient of friction, and high thermal conductivity, diamond films are viewed as the ultimate coating for many cutting tools and bearing surfaces. These applications account for a large part of the current commercial market in diamond films.
Depositing high-temperature superconducting films is another area SERI is researching. Metalorganic precursors for the ytrium-barium-copper-oxide group of superconductors were spin-coated onto single-crystal yttrium, stabilized zirconia, and magnesium oxide substrates, dried in air, then thermally processed in oxygen using the solar furnace at SERI. Zero resistivity was measured at a temperature of 74.7 K. SERI believes that the higher deposition rates and high heating rates made possible by the solar furnace may not only make it more economical to produce existing superconductivity films but will also enable development of completely new films.
Other experiments at SERI have demonstrated the technical feasibility of conducting self-propagating high-temperature synthesis reactions enabling the formation of advanced ceramic coatings of nickel aluminide, titanium diboride, and titanium carbonate on a variety of metallic substrates; cladding, such as stainless steel and nickel-chrome alloys on mild steel substrates; phase transformation hardening of steel alloys to obtain various degrees of surface hardness; and finally, radiative joining where very localized, high heating rates enable the local joining of adjacent structures without heating the whole structures.
To determine whether these applications are likely to be economically competitive, two studies, one conducted by Sandia National Laboratory and the other by SERI staff, compared the cost of photons for materials processing using a solar furnace to that using arc lamps and a laser. These studies concluded that better economics existed under most circumstances for the solar furnace for a large area of the United States where insolation is high.
One area of application for the super solar concentrator that Winston's group is working on involves the direct solar pumping of lasers. Researchers have worked on solar-pumped lasers for over 25 years. Because of the relatively low solar flux available using conventional concentrators, solar pumping of solid-state crystal lasers has achieved low conversion efficiencies of about 1 percent. With high-refractivc-index CPCs it may be possible to use laser materials with overall conversion efficiencies in the range of 5 percent. Current SERI-sponsored research at the University of Chicago aims to
demonstrate this conversion efficiency using a solar-pumped dye laser.
Both the technical feasibility and the resultant conversion efficiency of a solar-pumped dye laser are closely tied to the availability of a high solar concentration, because the concentration has to be high enough to overcome the threshold requirement for lasing and because the laser's overall efficiency increases as the concentration increases. For this reason, attaining concentrations higher than 50,000 suns at SERI's HFSF will be a particularly significant milestone. A solar-pumped dye laser is attractive because it promises to have a higher overall conversion efficiency, tunability, and higher lasing frequencies. In addition, the dye laser promises to be more reliable since it will eliminate the cooling and thermal gradient limitations of solid-state lasers. Thus, SERI believes the technical feasibility of a solar-pumped laser should increase the economic viability of this powerful
Demonstration of a solar-pumped dye laser at SERI's HFSF is scheduled for early 1992.
Conceptually, applications of a solar-pumped laser include all areas in which current lasers are used. Other categories of applications include the photochemical dissociation of toxic compounds such as polychlorinated biphenyls and the photosynthesis of inorganic high-value substances such as ceramic carbides and borides.
The full range of possible uses for the solar furnace has yet to be realized. Three independent studies, now in progress at the National Research Council/Energy Engineering Board, SRI International, and MIT, aim to identify the most promising among a plethora of new and emerging applications. Meir Carasso predicts that researchers from industry and academia will want to use the furnace for a variety of experiments. "This is a significant new resource for the country, and we actively encourage its use by all," said Carasso. "The application of the furnace is limited only by our imagination and creativity. "
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||new solar power developments|
|Date:||Feb 1, 1991|
|Previous Article:||New life from old oil wells.|
|Next Article:||Junkers Ju 52: Europe's ubiquitous transport.|