Achromatic doublet on glass fresnel lenses for concentrator photovoltaic systems

Resumen

In the last years, the constantly growing energy demand of the population on earth, together with the increasingly topical environmental ethics, resulted in relevant advances in the field of the renewable energy sources. Among them, photovoltaic (PV) energy is one of the most widespread around the world and has often achieved a levelized cost of electricity (LCOE) comparable with conventional fossil fuels. Within this field, a special interested is dedicated to the concentrator photovoltaic (CPV) technology, whose measured efficiencies are the highest among all the existing PV technologies, achieving the record efficiency of 43 % in 2016. This thesis is dedicated to the development, manufacturing, and characterization of a novel optics for CPV applications capable of either increase the maximum attainable concentration or to widen the tolerance to assembly errors, tracking, temperature, and spectral variation, while simultaneously maintaining competitive costs. Nowadays, most of the CPV modules in the market are based on hybrid silicone on glass (SoG) Fresnel lenses. Such technology, whose success has been ensured by its excellent stability under outdoor exposition and by the easiness of the processing required for the manufacturing, has an important limitation: the chromatic aberration. Such phenomenon causes light with different wavelength to focus at different distances from the lens, limiting the maximum attainable concentration to a factor of slightly more than 200 suns, if no light spillage is assumed. In modern CPV module, a secondary optical element (SOE) is commonly coupled with the solar cell in order to increase the concentration and, at the same time, smooth the spectral distribution of the light over the cell. The implementation of SOE in a CPV module, apart from introducing a new source of losses or fails, results in additional costs due to manufacturing, assembly, etc. of the SOE itself The work carried out in this thesis aims to the development of an novel cost-effective achromatic lens for CPV applications, named achromatic doublet on glass (ADG) Fresnel lens. The proposed lens technology, thanks to the achromatic design resulting in a reduced chromatic aberration, allows to increase the maximum attainable concentration and to smooth the spectral distribution of the light, without the implementation of a SOE, with obvious benefits from the financial and the reliability points of view. The lens design consists in a rigid glass substrate on which a plastic piece featuring Fresnel grooves on both its sides is glued using an elastomeric materials. As a consequence, by choosing the pair of materials with the adequate optical properties, the well-known concept of a thin achromatic doublet can be applied, while maintaining low costs thanks to the simple manufacturing process. In fact, the plastic piece is manufactured using high reliable industrial processes such as injection molding, compression molding, or hot embossing. Finally, the ADG sandwich is obtained by lamination, which is envisaged to be similar to the method commonly used to laminate conventional flat PV modules. In the first part of the thesis, a comprehensive investigation over the materials that may be potentially used to manufacture ADG lenses is presented, together with the reasons which led to the current ADG design. Ray-tracing simulations based on Montecarlo method have been used as a tool to assess which pair of materials provided the strongest reduction of the chromatic aberration. Among all the materials available, a thermoplastic elastomer (TPE) and the polycarbonate (PC) with high transmission in the ultraviolet (UV) region have been selected. With this configuration, ray-tracing simulations predicted a relative optical efficiency (i.e. the ADG optical efficiency normalized with the optical efficiency of a reference SoG lens used as benchmark) of 92.6 % and, in addition, the maximum attainable concentration of the simulated ADG lens is the double than the concentration simulated for the SoG lens. Furthermore, the simulations results have been used to understand the tolerance of the lens structure to manufacturing errors with the purpose of understanding what is the quality required in order to achieve a high performance optics. Afterwards, the method for manufacturing ADG lenses was defined and the required equipment (a laminator machine) was manufactured ad-hoc at the laboratory of the Solar Energy Institute, Technical University of Madrid (IES-UPM). A huge number of prototypes were manufactured and characterized at the IES-UPM laboratory, experimentally demonstrating from the beginning the achromatic behavior of the prototypes and, consequently, the feasibility of the concept. Later, the work focused on the optimization of the ADG prototypes. Three ADG generations have been developed. From the first generation, whose measured relative optical efficiency is equal to 85.5 %, many things have been optimized in order to achieve, for the third generation, the relative efficiency value of 89.4 %, which is close to the theoretical limit value predicted by ray-tracing. The third generation with respect to the first includes an improved design, new materials with enhanced transmission, a surface treatment aimed to enhance the adhesion between the materials of the sandwich, and a completely automated machinery used in the manufacturing process. Also, the challenges deriving from the manufacturing of arrays composed of many ADG lenses have been studied in the thesis. The main obstacle was the absence of a mold designed to inject the parquet of plastic all in once. Nonetheless, manually gluing the individual lenses, several parquets composed of many lenses have been successfully laminated and characterized resulting an overall efficiency only 1 − 2 % lower than the elementary units composed of one individual lens. The main reason of the efficiency reduction is the high dispersion of the measured efficiency values, which in turn is caused by the handmade procedure employed to glue together all the lenses composing the array and by a possible border effect during the lamination. However, both these limitations would be easily avoided in a utility-scale production line. Finally, complete CPV modules including modern high-efficiency triple-junction (3J) solar cell have been assembled and measured outdoor over a time lapse of more than four months. Two modules were assembled. The first is a mono-module, i.e. a system composed of one ADG Fresnel lens and one 3J solar cell, while the second is a full area module composed of 10x5 lenses and as many 3J solar cells. The mono-module was used in order to highlight the potential of the ADG technology and to obtain the best possible results attainable with the current development level. Conversely, the full module was needed to show that the ADG technology is actually possible and that, in only three years and a half of development, was already conceptually demonstrated and experimentally tested in relevand industrial environment. The characterization of the module demonstrated that the achromatic design, apart from providing a higher concentration, contributes to reduce the sensitivity of the module to ambient conditions such as temperature and spectrum. In the last months, a cost analysis of the developed technology is carried out in order to prove that the ADG technology, apart from the technological advantages already remarked, provides also a significant economic saving. The pair of materials selected are cheaper than the silicone used for conventional Fresnel lenses. This, together with the enhanced annual energy yield provided by the improved thermal and spectral stability, demonstrated that the contribution of the lens cost to the LCOE is significantly lower for the ADG lens than for te SoG lens. Such difference results to be even higher if a high concentration scenario is considered.