The research of Roland Winston and Joseph O'Gallagher

Nonimaging Optics and Solar Energy

Nonimaging optics is a relatively new optical subdiscipline that has experienced dramatic growth since its inception more than thirty years ago. The term refers to the optics of extended sources in applications for which image forming is not important, but for which effective and efficient collection, concentration, transport, and distribution of light energy is. The nonimaging optics group at The University of Chicago specializes in the development of this new optical discipline and its applications in a wide variety of very different systems. These include fiber optics devices; detectors for faint light sources used in astronomy, astrophysics, and high energy physics; efficient lighting systems; and, most importantly, systems for the collection, concentration, and conversion of sunlight to other usable forms of energy. Over the past thirty years the work of our group has established the properties of this new class of optical devices, developed a formalism suitable for general analysis, and led to the discovery of many new concentrator designs.

The sun is the direct or indirect energy source for virtually all the dynamic processes occurring on our planet. It has the potential to be an effectively limitless source of energy for human society's needs, yet providing a practical and economical means of harnessing this supply has been an elusive goal. It is in pursuing this goal that the role of new concentration techniques is so important, since one means of reducing the cost and improving the performance of solar collector systems is through the use of optical concentration. Traditional image forming optical techniques have been used for centuries in solar energy, but by their nature such systems have narrow fields of view and must track the sun to achieve useful concentration. Nonimaging optics provides a general method for designing optical systems that can, in principle, reach the so-called "Thermodynamic Limit" of concentration. (This limit is the maximum possible concentration allowed by physical conservation laws and cannot be approached by traditional optical designs.) This in turn allows concentration ratios a factor of 2 to 4 higher than previously thought possible for a given angular acceptance range. In 1990, our group produced and measured solar flux levels exceeding those at the surface of the sun itself. This remarkable achievement brought solar concentrator research into a new domain and served as a dramatic demonstration of the power of the techniques of nonimaging optics.

Our work is both theoretical (described in some detail below) and experimental. The experimental component lies in the design, fabrication, test, deployment, and demonstration of solar concentrating systems based on these new principles. The most visible evidence of the usefulness of these techniques has been the emergence over the past three decades of a new type of solar collector, the Compound Parabolic Concentrator (CPC). This concept was originally developed at Argonne National Laboratory in collaboration with the group here at the University. CPCs are long reflecting troughs which maximize the concentration for a particular absorber so that no active tracking is required. The CPC has been under development for nearly twenty-five years, and the most recent designs are readily achieving their early promise of being able to provide efficient conversion of sunlight to thermal energy at temperatures up to 300°C without any tracking. Double effect absorption cycle chillers, which are now being commercially produced, can be effectively driven by utilizing these collectors; a demonstration of such a system is being carried out in Sacramento, California. This application represents a high-value application utilizing the CPC collector that can have a near-term impact on society's energy needs.

Useful applications of the methods of nonimaging optics occur wherever concentration is desired. Our group has been involved in projects ranging from planning for the use of CPCs in the rural village situation in India to the use of nonimaging terminal concentrators to increase the concentration of parabolic dish power generation systems being built by the Sandia National Laboratory. We operated a field test array of CPCs on a Navajo elementary school in New Mexico and have built optical models for one- and two-stage concentrator systems for use on deep space vehicles. Other designs have greatly increased the practical solar flux concentrations limits which can be achieved by any optical system. We designed and operated a multi-stage concentrator system which produced a solar flux equal to 84,000 "suns" (84,000 times the ambient level of sunlight). This exceeded by a factor of five the highest levels ever produced by conventional optical concentrators. Other experiments performed in collaboration with colleagues at the National Renewable Energy Laboratory's (NREL) High Flux Solar Furnace (HFSF) have demonstrated the effectiveness of using concentrated sunlight and advanced nonimaging secondaries to pump lasers and produce fullerenes (potentially useful new forms of molecular carbon). Other applications of this "high flux" capability that we have studied and/or demonstrated have included high temperature materials research, potential for the destruction of hazardous waste products, and most recently to provide a power source for solar thermal propulsion systems in space. Finally, employment of the techniques of nonimaging optics in the design of so-called two-stage concentrators has resulted in significant performance improvements in conventional parabolic dish concentrators for solar thermal electric power, dramatic changes in the approach to solar furnace design, and the commercial development of a two-stage photovoltaic concentrator with greatly reduced tracking accuracy requirements.

The subdiscipline that has come to be referred to as Nonimaging Optics was invented at The University of Chicago in the mid 1960s with the discovery that optical systems could be designed and built that approached the theoretical limit of light collection. The limit to which a beam of light can be concentrated (or disseminated) is fundamentally connected to its angular divergence (characterized by a half-angle q), by the well known sine law of concentration which can be simply stated as

An early impetus to the development of nonimaging optics was the realization that conventional imaging optics falls far short of the sine law limit. For example, a parabolic reflector achieves, at best, one quarter of the above limit.

Nonimaging optics has proven to be powerful because it departs from the methods of traditional optical design and aims to maximize the collecting or transmitting power of concentrating or disseminating optical elements and systems of elements. This is accomplished in part by applying the concepts of Hamiltonian optics, studying phase space representations of the light distributions being manipulated, and very often by applying thermodynamic arguments or radiative transfer methods wherever appropriate. Nonimaging optics is the optics of extended sources and has more in common with radiative transfer than with conventional optical design and relies on such notions as "Hottel strings." In considering extended sources, one is led to distributions in phase space and inevitably to the Theory of Radiance. These approaches have led to many new insights in understanding the operation of light transport and concentration systems. A measure of the importance that the field has attained can be obtained by examining the Table of Contents of the more than 600-page volume on the subject covering the first twenty years of activity in this field and which appeared recently in the SPIE Milestone Series. Furthermore, it is well known that the use of appropriately designed nonimaging devices allows one to achieve ultra-high solar fluxes. The optical designs for such systems have been described in a recent review.

So far there have been three distinct phases in the approach to nonimaging optical design. The earliest nonimaging designs were constructed by optimizing the collection of the extreme rays from the source to the target: employing the so-called "edge-ray" principle. Later, new concentrator types were discovered by placing reflectors along the flow lines of the "vector flux" emanating from lambertian emitters of various geometries. Finally, a few years ago the discovery that making the design edge-ray a functional of some other system parameter permitted the construction of whole new classes of devices with heretofore unimagined capabilities. These "tailored edge-ray" designs have dramatically broadened the range of geometries in which nonimaging optics can provide a significant performance improvement.

Finally, in recent years our group has become interested in trying to understand some of the concepts implicit in nonimaging optics in a physical optics context. In particular, the present position in the theory of Generalized Radiance (i.e. radiance in the scalar wave model) seems unsatisfactory in that there are many possible definitions, none of which satisfies all the properties which radiance intuitively ought to have except possibly in relation to particular narrowly defined types of source. Furthermore, radiance defined in these ways leads to details in the calculated values which are inherently impossible to observe. Our work shows how introducing the measurement process into the theory of radiance in a self-consistent way removes certain long-standing difficulties in the subject such as the fact that generalized radiance can take on negative values. A way out of the difficulties posed by the theory of Generalized Radiance is to introduce an "instrument function" which is reciprocal to the wave field being measured. Recently we have turned our attention to methods of calibrating the instrument function in an attempt to make the subject useful in practical radiometry.

Find more information and links to related information on this group's web page.

• Return to top.