Elsevier

Energy

Volume 135, 15 September 2017, Pages 610-624
Energy

Design, optimization and optical performance study of tripod heliostat for solar power tower plant

https://doi.org/10.1016/j.energy.2017.06.116Get rights and content

Highlights

  • Concept of tripod heliostat explored to address challenges in conventional system.

  • Systematic cost optimization protocol for real heliostat field load conditions.

  • Optimization protocol also considers heliostat deformation as one of the constraint.

  • Ray tracing is used in conjunction with optimized heliostat to ensure performance.

  • Techno-economics of two heliostat designs compared based on cost and performance.

Abstract

Heliostats account for about 50% of the capital cost of power towers. In conventional heliostats with vertical pedestals and azimuth-elevation drives, the support structure contributes 40–50% of this cost due to heavy cantilever arms required by the large spanning structures. Additional costs are imposed by costly, difficult to maintain drive mechanisms. Here we show that a tripod heliostat can substantially address these shortcomings. We have presented the protocol and results of systematic structural analysis of heliostats with aperture areas of 62 and 100 m2. We have included effects of shape on load reaction and structure cost. An in-house ray-tracing software is incorporated to estimate the performance penalties due to deformation under gravity and wind loads. The analysis shows that the additional energy collection by a less-stiff, larger heliostat more than offsets the spillage due to the greater deformation of the same.

We have demonstrated that the economics of power towers are strongly governed by the structural cost of the heliostats rather than by their optical performance. We have brought down the cost of a tripod heliostat to $ 72/m2 which is less than half that of the conventional systems and meets the target set by the US National Academy of Engineering.

Introduction

It is uncontroversial that mass deployment of solar power is the backbone of a successful transition to a renewables-heavy economy. While other renewable energy sources (e.g. wind, biomass, hydroelectricity etc.) are derivatives of solar-energy, tapping directly into solar radiation presents the opportunity to increase solar-to-electricity efficiency by an order of magnitude. There are two main ways of doing so: solar-photovoltaic and solar thermal. While the former is more popular currently, it is (pending a revolution in electricity storage) an intermittent power source. On the other hand, with inexpensive thermal storage, the latter has the potential to directly replace coal as the source of base-load renewable power [1].

Currently, solar thermal (or concentrating solar power) plants fall into four main categories based on the manner of radiation harvesting: (a) parabolic trough collector, (b) linear Fresnel, (c) power tower and (d) parabolic dish. Solar to electricity conversion efficiencies of these technologies are in the range of 11–16%, 8–13%, 18–25% (concepts) and 12–30% respectively [2], [3]. Typical operating temperatures of the heat transfer fluids in these collectors are, 350–550 °C, 390–450 °C, 500–1000 °C, and 550–750 °C, respectively [2]. Clearly, the power tower (also called Central Receiver Solar Power) technology can achieve the highest range of temperatures: thus, the highest Carnot efficiency. Further, the higher operating temperature and the possible temperature differential in the storage system can significantly reduce the cost of thermal energy storage (TES) [2]. Also, the high temperatures attainable in power tower receivers allow their use for solar-hydrogen generation: an important intermediate step in manufacturing what are called “solar-fuels” – the “green” alternative to transportation fuels.

In power tower plants, heliostats (in the solar field) constitute the largest single cost element: about 40–50% of the total plant cost [4]. According to a Sandia National Laboratory report, the target heliostat cost for the molten-salt based power towers should be under $100/m2; for a commercially attractive Levelized Electricity Cost (LEC) of 0.059 $/kWh [4]. While comparing different heliostat designs, the cost must, of course, be weighed against performance. For example, a less expensive heliostat, made using less rigid supports, will deform more under operating loads (gravity, wind etc.) and result in a larger “spot” on the receiver. Hence, it is possible for a heliostat with lower specific cost but having lower optical and tracking accuracy to lead to higher levelized electricity cost [5].

This is a complex problem to address. Heliostat performance is affected by structural deformations due to wind load and superimposed sag due to gravity: which are themselves dependent on elevation angle and constrained by structural stiffness [6], [7]. However, commercially available software that works really well for structural analysis do not readily allow the subsequent optical analysis.

It is, nevertheless, a very important and useful analysis. The cost and efforts in initial on-field structural alignments can be greatly reduced with the proper design [7]. According to a review by Coventry and Pye (2013), the heliostat performance and heliostat total cost play a critically important role in determining Levelized Cost of Electricity (LCOE): from a sensitivity analysis of system with a 250 $/MWh baseline they showed that a design that gives 1% performance improvement is equivalent to reducing solar field cost by about 2.3% [8].

The total annual energy collection of the heliostat field varies with the diurnal and seasonal movement of the Sun [9]. Energy collection from single heliostats and the receiver flux distribution can be calculated using ray-tracing software such as, STRAL [10], HFLD [11], SOLTRACE [12], and MIRVAL [13]. Sansoni et al. [14] have developed software tools in integration with the Zemax ray tracing package to optically design a heliostat field and calculate power collected by therefrom. Mechanical performance of the structure under wind load was analysed using finite element method by Zang et al. [15] and commercial structural analysis software are readily available.

Software for optical analysis of deformed heliostats are much more specialized and very few papers have been published with this analysis. Yuan et al. [16] have published results of deformation of standard heliostats under gravity and wind load while Meiser et al. [17] have simulated parabolic trough collectors under deformation.

For a given receiver size, the optical performance of the heliostats is affected by their structural deformation and tracking accuracy. This is because the beam error induced by deformation results in substantial radiation spillage from a given receiver geometry [18]. These deformations happen in the heliostat support structure mainly due to wind load and gravity load (self-weight). Deformations affect optical performance and hence the annual energy collection (hence LCOE). There is plenty of low-hanging fruit in this area. For example, groups that reduced wind drag simply by creating gaps between facets found improved annual energy collection with lower heliostat costs [19], [20]. Hence, it is essential to consider deformation while designing the heliostats and develop tools to calculate annual energy collection by deformed heliostats.

In this paper, we have used a commercial structural analysis software (STAAD Pro) whose output was fed to a ray-tracer we developed in-house to execute exactly such a calculation.

In this work, we have applied this protocol to a “tripod” heliostat [21]: which offers considerable advantages over a conventional pedestal heliostat. A typical conventional (or pedestal) heliostat has been schematically shown in Fig. 1(A). It has a vertical pedestal (1) with a horizontal cantilever arm (2) supporting the reflectors (3). The azimuth (4) and elevation drives (5) are located at the junction of the horizontal arm and the vertical pedestal (6). These drives as much as 35% of the heliostat cost [4], [22], and also present maintenance challenges not least of which is the frequent failure of the gear box: which is more common than structure or facet failure [6]. Also, in the conventional design, the heliostat is supported on a central pedestal with horizontal cantilevered arms. It may be noted that material of construction (MOC) requirement of the cantilevered arms increases super-linearly with the unsupported span of the structure. Hence, making the cantilevered span longer (while maintaining the same degree of structural deformation) in vertical pedestal type heliostats requires super-linearly greater material: hence is costlier. For large span heliostats, therefore, there exists an opportunity to reduce costs by reducing the unsupported length. This objective can be readily achieved with the tripod arrangement as shown in Fig. 1(B). In this case, the triangular formation of the members (8) supports the large span of the structure through the yoke connections (9). This also eliminates the gearbox failure problem of conventional heliostats, since the load is distributed onto three legs (7) and there is no concentration of mechanical stress of magnitudes equivalent to those at conventional heliostat supports. Hence, the tripod heliostat geometry is eminently suited for mass deployment in the solar field of central receiver power towers.

The objective of this paper is to critically examine the factors governing the cost of tripod heliostats and attempt to use these insights to minimize these costs. The analysis has been performed for 100 m2 and 62 m2 heliostats.

Section snippets

Qualitative description of tripod heliostat

It is useful to describe the tripod heliostat concept in some detail. Tripod heliostats have been described previously [21] and we have built on some of the existing features. The schematic of our tripod heliostat structure is shown in Fig. 1(B). This structure consists of 4 sub-assemblies namely, Tripod legs (7), Equilateral triangle of the tripod (8), Yoke assembly (9) and, Mirror fixing assembly (10). At least two legs of the tripod are equipped with linear actuators (e.g. hydraulic systems)

Optimization and structural design of tripod heliostat support structure (100 m2 and 62 m2)

The desired basic functionalities of any field heliostat are as follows. (a) It should support the mirrors with sufficient rigidity: restricting maximum structural deformation to within tolerable deformation limits under all operating load conditions. (b) It should survive an extreme wind load of 140 km/h and seismic load. (c) The structure should be able to orient itself to any desired orientation in the field. (d) It be able to efficiently track the Sun and direct rays onto the receiver

Conclusions

We have carried out structural optimization of tripod heliostat with 100 m2 and 62 m2 reflector area, with maximum deflection restricted to 3 mrad. The results show that the optimum tripod triangle size is always in the vicinity of 3000 mm which is more or less independent of the yoke height. Load reaction also passes through a minimum for different yoke heights but tends to shift towards lower triangle sizes and higher load reactions. Hence, the heliostat shape with triangle size of 3000 mm

Acknowledgement

One of us (VCT) is grateful to UPL Ltd. for supporting my fellowship.

References (29)

  • G.J. Kolb et al.

    Heliostat Cost Reduct Study

    (2007)
  • A.B. Zavoico

    Design basis document

    (2001)
  • D. Todd Griffith et al.

    Structural dynamics testing and analysis for design evaluation and monitoring of heliostats

    J Sol Energy Eng

    (2014)
  • J. Coventry et al.

    Heliostat cost reduction - where to now?

    Energy Procedia

    (2013)
  • Cited by (13)

    • An optical-mechanical integrated modeling method of solar dish concentrator system for optical performance analysis under service load

      2022, Energy
      Citation Excerpt :

      For example, Zang et al. [29] calculated the structural deformation of the tower heliostat under self-weight and wind load by the finite element analysis method, and estimated its focusing accuracy according to the maximum deformation value of the mirror surface. Thalange et al. [30] optimized the structural parameters of a new tripod-supported heliostat with a constraint on the optical error value estimated by the maximum deformation value, and they also pointed out that the analysis of the optical-mechanical integration of the heliostat considering the effect of structural deformation is complex and still lacking in research. The surface reconstruction method uses the deformation information of the discrete points of the concentrator mirror to re-fitting the reflector surface, and then uses ray-tracing method for optical analysis.

    • A coupled structural-optical analysis of a novel umbrella heliostat

      2022, Solar Energy
      Citation Excerpt :

      Yuan et al. calculated the deflection under gravitational loading via a finite element analysis (FEA) in ANSYS-Mechanical, and the resulting deformed heliostat geometry was analyzed using the Breault APEX optical engineering software tool to evaluate changes in beam size and shape. Thalange et al. (2017) also simulated heliostat structural analysis, but with a different commercial software (STAAD Pro), and the output was extracted to exactly calculate optical performance with ray-tracing code. Some similar work has also been done for parabolic trough collectors by Meiser et al. (2017) and Yang et al. (2018).

    • Vertical-rocking-horizontal vibrations of a rigid disk resting on multi-layered soils with groundwater level

      2021, Applied Mathematical Modelling
      Citation Excerpt :

      It raises a concern about the vertical dynamic response of a rigid disk on a multi-layered soils with groundwater level which is still yet unclear. Horizontal and rocking dynamic responses also deserve further attention especially when superstructures (machine, heliostat, offshore wind turbine and bridge, ect.) are subjected to lateral excitation force, wind and wave load [32-38]. To solve the dynamic response of a layered system, the thin layer method divides each layer into numerous thin sub-layers.

    View all citing articles on Scopus
    View full text