Resumen de Captador anidólico de canal paramétrico para central termosolar
It is a fact that plants producing electricity from solar energy can provide safe energy with predictable costs unaffected by geopolitics and global energy markets constraints (International Energy Agency, 2007). However, the use of solar radiation to generate electricity in a cost effective way remains still a challenge (International Energy Agency, 2014). Current Parabolic Trough Collector (PTC) technology is the most mature technology when compared to Central Receiver (CRS) and Fresnel systems. Since the first Solar Electric Generating Systems (SEGS) plants built in the Mojave Desert in California in the 80’s using the LS-2 and LS-3 generations of PTCs that marked the beginning of modern development of Concentrated Solar Power (CSP) plants worldwide (Fernández-García A. et al., 2010), today more than 2GW of this technology are currently on operation worldwide (CSP World Map, 2015).
Even if solar thermal energy power plants are not yet competitive in terms of cost of electricity when compared to fossil fuel plants, there is enough room for improvements, in both efficiency and costs, to reduce or even eliminate this gap (International Energy Agency, 2014). The use of molten salt and water as heat transfer fluids instead of oil allows achieving more efficient plants (Zarza E., 2004, Estela, 2012). From an optical point of view, parabolic trough collectors optics is far from the thermodynamic limit and there is room to bring the concentration closer to the physical limit (International Energy Agency, 2014) – as it is currently being done with the Fresnel collectors (Collares M., 2009) – and to increase the operating temperature and the cycle efficiency while comparatively reducing heat losses to the atmosphere and reducing wind loads and receiver distance to the mirrors to minimize structural costs.
In that regards non-imaging optics (also called anidolic optics), developed in the 80’s by Welford and Winston (Welford W. T. et al., 1989), is a powerful tool used in the development of solar concentrators. During the last three decades this branch of the Geometrical Optics has undergone a major evolution that has resulted in the progressive development and optimization of, mainly, photovoltaic concentrators.
Regarding PTCs there have been some attempts to optimize its optics while maintaining the parabolic shape (Rabl A. et al., 1982). Some developments focus in increasing the concentration through a reduction in the acceptance angle, which places a higher demand in tracking accuracy and reduces tolerances with respect to wind loads, quality of mirrors, control and mounting imprecisions. With this strategy, even if the concentration is improved, the ratio concentration relative to the thermodynamic maximum limit remains constant if the rim angle is not changed. In fact, this value can only be modified using secondary concentrators or new optical geometries. There exist proposals to improve the PTC performance by means of secondary concentrators sometimes with complex shapes and, in other cases, with secondary mirrors in contact with or very close to the absorber tube (Ciemat, 1990, Collares M. et al., 1991). These solutions force designers to locate the secondary inside the evacuated tube, which is unpractical from a manufacturing point of view. There exist other proposals that use commercial evacuated tubes and accommodate the gap between the glass envelope and the receiver tube without losses being compatible with the placement of second stage mirrors outside the glass envelope but generating a large separation of the receiver to the primary reflector (Rabl A., 1985). This approach penalises the concentrator in terms of structural cost so that it is more practical to reorient the design towards a Fresnel type concentrator where both the absorber and the secondary remain mechanically indendent of the primary.
Other non-parabolic primary mirrors solutions have been proposed with a sizeable gap between the optics and the absorber (Benitez P. et al., 1997, Cannavaro D. et al., 2013). In this case the proposed primary curve is no longer a parabola. In order to make the distinction with the traditional parabolic shape and taking into account than the new curve is a parametric one I have decided to use this name in this work.
It is from this point that the motivation for the present work arises. In this thesis the use of anidolic optics to obtain parametric trough concentrators with improved optical, thermal and structural efficiency is investigated and compared to current parabolic trough solutions with the aim of increasing the overall efficiency of current solar thermal power plants.
The present document is structured in 7 chapters:
In chapter one a thermodynamic description of the solar radiation, and a discussion on the maximum work production rate obtainable from it, is done as a first step to a further analysis of the physical limit to its concentration and as a necessary introduction to understand the concept of etendue.
In chapter two etendue and its conservation from the point of view of thermodynamics are introduced. Both issues can be seen as central in this field and are the pillars on which the rest of the thesis stands. The chapter concludes defining the thermodynamic concentration limit of solar radiation which will be used, a posteriori, as a reference to understand state of the art concentrators’ potential for improvement always from the point of view of the concentration.
Chapter three starts with a review of the parabolic trough collector’s optics. A symmetric non imaging Parametric Trough Collector (PmTC) is proposed. The new concentrator is designed guaranteeing etendue matching for a flat absorber. The flat absorber can be replaced with a multi-tube receiver supported by a thermal insulating base and with transparent covers similar – in section - to the receiver used in the Compact Linear Fresnel Concentrator (CLFC) developed in Australia by Mills (Mills D.R. et al., 2000) for Direct Steam Generation (DSG). The optical design method and the design assumptions are explained. A numerical shape optimization is run and an analysis of the characteristics of the final optics and its merits is performed through comparisons with state of the art optics. The proposed concentrator adds geometrical flexibility that may help to reduce the wind loads and may also allow the definition of a less expensive structure with a better controlled separation between the primary reflector and the ensemble absorber - secondary.
In chapter four a new symmetric non imaging Parametric Trough Collector (PmTC) for a circular evacuated receiver in this case is proposed, applicable for oil or molten salts as heat transfer fluid but also for DSG. In this case the secondary concentrator can be manufactured by partially mirroring a diameter adapted glass tube - either internally or externally - or alternatively by means of a commercial evacuated receiver and an independent arc of circumference external secondary reflector. The optical design method and the design assumptions are explained. A sensitivity study is run and an analysis of the characteristics of the final optics and its merits is performed through comparisons with state of the art optics. The proposed concentrator adds also geometrical flexibility regarding wind loads and maintains high rim angles and a controlled separation between the primary reflector and the ensemble absorber - secondary.
As Fresnel concentrators and Parabolic or Parametric Trough concentrators share the same concentration limit, a comparison in terms of concentration ratio relative to the maximum thermodynamic limit among all of current state of the art Fresnel solutions is done in chapter five. It is out of the scope of this thesis to compare Fresnel concentrators with Parabolic or Parametric Trough concentrators at a plant level. In depth studies analyzing in detail total optical losses (including cosine losses), receiver geometry with thermal losses, plant configuration and costs should be necessary to address the mentioned comparison at plant level.
In chapter six, Computational Fluid Dynamics (CFD) has been used as a “virtual” wind tunnel to compare the flow around a single module of the two PmTCs. Velocity vector field and mean values of aerodynamic coefficients were computed in a range of pitch angles and compared with commercial PTCs (LS2 y LS3 / Eurotrough). Two case studies - 2D and 3D simulations - have been analyzed. Results confirm that the PmTC for circular receiver behaves very similarly to the LS2 and LS3 geometries for the drag, lift and moment coefficients. The PmTC for flat absorber shows the worst performance in comparison with the other three collectors. In addition all rays suffer two reflections before reaching the absorber in this collector while 15% of the rays undergo secondary reflection before arriving to the absorber in the case of the PmTC for circular receiver. These facts and the non-availability of a commercial solution for such kind of receiver have induced the author to select the solution of circular receiver with commercial evacuated receiver an external secondary as the option to be evaluated at plant level in comparison with current PTC technology.
In chapter seven a commercial 50 MWe oil PTC power plant model without thermal energy storage located in Seville (Spain) is build up. A second simulation is run with the proposed PmTC. The collector area was chosen similar to the area of the PTC to assure similar single loop mass flow rates. Since the anidolic designs ignore transmission, absorption and reflection optical losses; calculations of the optical efficiency by means of Montecarlo raytracing simulations using real slope errors distributions and considering Fresnel losses were performed. Annual simulations allowing thermal losses and electricity annual electricity yield calculations allow the final comparison to be made.
The main conclusions of the thesis are:
1 Two Parametric Trough Collectors for flat and for circular absorber geometries obtained using non-imaging optics are presented and discussed. The flat absorber solution can be replaced by a multi-tube receiver which has the advantage of a better heat transfer from the absorbing coating to the heat carrier fluid so that large number of tubes improves the thermal efficiency. As the pumping power increases almost as a cubic power of this number, Direct Steam Generation remains the best application for this kind of receiver. The circular absorber solution can be solved with a partially mirrored big glass tube hosting an eccentric absorber tube inside or by using a commercial evacuated receiver with an external secondary. In this case either oil, molten salt or water can be used as heat transfer fluids.
2 A comparison in terms of concentration ratio relative to the maximum thermodynamic limit with state of the art trough collector based on a parabolic shape and with state of the art Fresnel collectors was done. The first comparison shows that the ratio concentration relative to the thermodynamic maximum increases up to 90% maintaining rim angles greater than 80° with the same effective acceptance angle than current LS3 / Eurotrough optics. Thus both proposals have the potential to deliver considerably more energy onto the same absorber perimeter for the same effective acceptance angle. In addition the shape of the proposed collectors having important discontinuities in the middle of the primary reflectors were identified as opportunities to reduce wind loads while adding flexibility to control the separation between the ensemble receiver-secondary mirror support substructure and the concentrator torque structure.
On the other hand, the etendue-matched two-stage Fresnel concentrator with multiple receivers has a product efficiency times the concentration relative to maximum in the range 55% up to 85% which is higher than the 60% one obtained with the best PmTC solution developed in this thesis. This second comparison is performed only from an optical point of view since both optics share the same concentration limit but it is out of the scope of this thesis to compare Fresnel concentrators with Parabolic or Parametric Trough concentrators at a plant level.
3 Detailed Computational Fluid Dynamics (CFD) wind tunnel simulations used to compare the flow around a single model-scale module of the two proposed concentrators and the LS3 / Eurotrough parabolic trough in a range of pitch angles, confirm that the solution for circular receiver behaves very similarly to the LS3 / Eurotrough geometry for the drag, lift and moment coefficients and that the solution for flat absorber shows more than 25% penalization in terms of maximum drag and moment values and more than 50% penalization in terms of maximum lift. This results together with the fact that in the flat receiver collector all rays suffer two reflections before reaching the absorber while only 15% of the rays undergo secondary reflection before arriving to the absorber in the case of the circular receiver solution, and the fact that it does not exist a supplier with a commercial solution for the multitubular receiver have motivated the selection of the solution for circular receiver with commercial evacuated receiver an external secondary as the option to be evaluated at plant level.
4 To understand the potential of the selected solution compared to the parabolic baseline two commercial 50 MWe oil parabolic trough power plant model without thermal energy storage located in Seville (Spain) models were built up. In the case of the parametric concentrator the collector area was chosen similar to the area of the parabolic collector to assure similar single loop mass flow rates. Since the hypothesis of zero transmission, absorption and reflection optical losses was made in the design process calculations, optical efficiency calculations using real slope errors distributions by means of Montecarlo raytracing simulations considering Fresnel losses were performed. Contrary to initial expectations, comparison with commercial Parabolic Trough Collector shows an optical penalization at zero solar incidence of 5.1% due to the reflectivity and additional dirtiness of the secondary mirror, to an increase in the end losses and to the Fresnel losses. Annual simulations show a 2.5% reduction in electricity yield when compared to the reference PTC plant. The improvement relative to the optical performance is due in part to lower thermal losses. Thus, the proposed solution does not show improvement compared to the reference PTC plant for a design loop oil outlet temperature of 391ºC.
