Search

Artificial Intelligence (AI) is increasingly playing a significant role in the design and optimization of renewable energy systems. Many AI approaches and technologies are already widely deployed in the energy sector in applications such as generation forecasting, energy efficiency monitoring, energy storage, and overall design of energy systems. This paper provides a review of the applications of key AI techniques on the analysis, design, optimization, control, operation, and maintenance of Solar Tower systems, one of the most important types of Concentrating Solar Thermal (CST) systems. First, key AI techniques are briefly described and relevant examples of their application to CST systems in general are provided. Subsequently, a detailed review of how these AI techniques are being used to advance the state of the art of solar tower systems is presented. The review is structured around the different subsystems of a solar tower system.

Central tower concentrating solar power (CSP) systems typically focus solar radiation upon a tubular solar receiver where radiation is absorbed and then transferred, by conduction and convection, into a heat transfer fluid. In this paper, a range of heat transfer fluids are compared, using energy and exergy analysis, and varying the tube diameter, tube wall thickness, and tube-bank flow configuration. The model optimises exergy efficiency including pumping work, assuming uniform flux, and neglecting the effects of thermal stresses, circumferential tube temperature variations and cost. Suitable temperature and pressure conditions are chosen for each fluid, based on a realistic configuration of an applicable thermal energy storage (TES) and power block (PB). The examined heat transfer fluids are molten salt (60% NaNO3, 40% KNO3), liquid sodium, supercritical carbon dioxide (sCO2), air, and water/steam. Results showed that liquid sodium at an elevated (540–740 °C) temperature range performed best, with a solar-to-fluid exergy efficiency of 61%. At a low temperature range (290–565 °C), sodium was still marginally superior to molten salt, even after allowing for some exergy destruction in a sodium-to-salt heat exchanger. Water/steam also performs relatively well in the receiver, although the difficulties of integrating it with large-scale storage make it a challenging heat transfer fluid for an integrated system. Using sCO2 as the heat transfer fluid appears infeasible due to excessively-high pressure stresses on the tubes. Air also appears unsuitable for simple tubular receivers, since poor heat internal transfer results in high losses due to much hotter external surfaces.

A cavity receiver geometry may represent a better option than an external receiver geometry for high temperature solar power systems because of reduced thermal losses. However, a limitation is that the small aperture necessary for an efficient (and affordable, practical) cavity receiver may lead to high spillage losses, negating the thermal efficiency gains. In this study, Monte Carlo ray tracing is coupled with heat transfer models to assess the benefit of secondary reflectors and spillage skirts on the performance of cavity receivers in the context of both dish and tower CSP systems. The heat transfer fluid is chosen as liquid sodium with inlet and outlet temperatures of 520 °C and 740 °C respectively. A spillage skirt and three different types of secondary reflectors, including conical, trumpet and compound parabolic concentrator (CPC) reflectors are investigated to decrease the energy loss. The results show that the spillage skirt helps in greatly decreasing the aperture size, while simultaneously reducing the spillage loss. The conical and trumpet reflectors increase the efficiency by reflecting rays outside the cavity inlet into the receiver, but the CPC does not achieve efficiency gains comparable to the other designs in both the dish and tower systems. For dish systems, highest efficiency is obtained with a combination of a spillage skirt and conical or trumpet reflectors, with a 1.0% increment in receiver efficiency compared to the best cylindrical receiver. For the tower system the efficiency gain is approximately 1.3% relative to the optimised cavity receiver alone.

Energetic and economic characteristics are studied for solar central receiver systems consisting of a polar heliostat field, a tower, a single-aperture cavity receiver, and an optional compound parabolic concentrator (CPC). System characterization and optimization are performed with a numerical model combining an in-house developed Monte-Carlo ray-tracing optical model, a simplified receiver heat transfer model, and a cost model based on the System Advisor Model (SAM). Based on the model, the effects of receiver temperature on the optical configuration of cost-optimal systems are elucidated, along with the benefits of using a CPC for improved energetic and economic performance. Under the assumptions made in this study, it is found that the overall minimum levelized cost of exergy is obtained by a non-CPC system with a receiver operated at approximately 900 K. A CPC benefits both the energetic and economic performance of systems only at elevated temperatures. The working temperature thresholds at which the energetic and economic performance benefit from the addition of a CPC are identified as 900 K and 1200 K, respectively. The general formulation of the model and broad range of values of the investigated parameters provide a universal predictive capability for studying techno-economic performance of concentrating solar thermal systems.

A methodology for in-situ measurements of radiative reflection and emission losses from a solar thermal receiver under high-flux irradiation is demonstrated. It combines radiosity analysis with photogrammetry and image recognition techniques to obtain directional and spatial radiosity distributions over receiver surfaces with a simple setup, mainly consisting of a camera. A CCD camera can acquire the radiosity in the visible range, which predominantly captures reflected solar irradiation. A thermal infrared camera can acquire the radiosity in the infrared range, which predominantly captures emission losses from the hot receiver surfaces. A hyperspectral camera can be used to obtain spectrally resolved results across a range of wavelengths. Images are taken from different directions in front of the receiver, and processed in software to obtain a point cloud via three-dimensional reconstruction, allowing the image data to be mapped onto a receiver mesh model. The receiver can be any shape, including those with complex-shaped cavity-like geometries exhibiting surface occlusion and light-trapping effects. These camera-based non-contact measurements allow for the performance of a receiver to be evaluated without interrupting its normal operation. The feasibility of the method is tested by quantifying the reflection losses from a multi-cavity tubular receiver under ~850 kW/m2 concentrated solar irradiation. The proof of concept is established by comparing the measured results with those from Monte-Carlo ray-tracing simulations.

To achieve further substantial reductions in the cost of concentrating solar-thermal power systems, recent efforts have been directed towards changing the working fluid in the receiver, adopting new approaches to thermal energy storage, and incorporating higher-efficiency power cycles. In this study, we present a novel CSP system configuration under development in ASTRI research program in Australia, incorporating a tubular sodium receiver, a high-temperature phase-change material storage unit, and a supercritical CO2 power block. The model is implemented using the SolarTherm framework in the Modelica language, and simulated using Dymola software. The novel system was compared to a ‘reference’ system with a molten salt receiver, two-tank storage, and subcritical steam Rankine cycle. For verification, the reference system was also modelled using the System Advisor Model software, and differences of 1% in annual receiver output and 2% in net electrical output were achieved when comparing Modelica and SAM simulations, despite difference operating strategies and time resolution. The results of the analysis of the novel system showed an improvement of annual solar-to-electricity efficiency to 19.6%, up from 15.6% for the reference system. This was achieved as a result of higher power cycle efficiency, despite the reduction in optical and receiver efficiencies resulting from the higher working fluid temperature. The ASTRI system was economically optimised for financial performance using 2012 as the basis year, and reached a minimum levelised cost of electricity of 11.42 ¢AUD/kWh for a solar multiple of 2.2 and storage capacity of 8–9 h at the location of Alice Springs, Australia.

This paper evaluates the annual dynamic performance and techno-economic feasibility of a solar fuels process driven by concentrated solar power (CSP). The process integrates solar supercritical water gasification (SCWG) of microalgae, syngas storage, and downstream Fischer–Tropsch (FT) synthesis, to produce renewable drop-in liquid transport fuels. Performance curves for key components, derived from design-point and off-design simulations in steady-state flowsheet models established in the previous work of the authors, are incorporated into a dynamic energy-based model of the overall plant in OpenModelica. Control units are implemented to model the high-level state transitions of the plant relating to ramp-up and ramp-down of solar-SCWG and FT units as well as syngas storage and dispatch. There is significant uncertainty about the achievable ramping time for FT plant, owing to the dynamics of the synthesis of long-chain hydrocarbons in the FT reactor, so this is modelled as a variable-duration ramp-up with no useful product output. A perfect one-day-ahead forecast scheduler is developed to regulate the supply of solar syngas to the FT unit to minimise these ramp-up events. The levelised cost of fuel (LCOF) for the plant is minimised using a genetic algorithm to adjust the solar multiple and syngas storage capacity of the plant, with weather data and location of Geraldton in Western Australia. The optimal configuration has a solar multiple of ∼4 and syngas storage of ∼20 h, and achieves an LCOF of ∼3.5 AUD/L (∼2.5 USD/L) and capacity factor of ∼72%. Sensitivity analysis shows that the price of the farmed microalgae feedstock has the strongest influence on the LCOF. The one-day-ahead syngas dispatch scheduler contributes an LCOF reduction of 19% compared to an immediate dispatch strategy. The importance of the FT ramping time in these simulations suggests that fast-ramping FT reactors (e.g. microchannel reactors) will be essential for the techno-economic success of solar-gasified FT fuels, and motivates further research and development on this topic.

Correlations for natural convection heat loss from solar cavity receivers are widely based on isothermal surface temperature assumptions, which do not occur in practice due to the local heat balance varying with position. An open question thus exists regarding the suitability of such correlations for non-isothermal conditions. This paper addresses this issue by presenting a new Nusselt correlation developed from an experimental investigation of natural convection heat loss from a non-isothermal scale-model cylindrical cavity receiver. Cavities that are considered in this work have length-to-diameter ratios of 1 and 2, are operated at peak temperatures ranging from 355 °C to 650 °C, and exhibit temperature differences along the cavity wall between 40 °C and 342 °C. Stagnation and convection zones, as well as view factor profiles, are observed to contribute to the wall temperature distribution as the cavity is inclined downwards. An energy balance undertaken for steady state provides insight into the effects of non-uniform surface temperature distribution and inclination-dependent surface areas on radiative and convective losses. Natural convection heat loss results from this work are compared with widely-used correlations from the literature that assume isothermal wall conditions, and systematic discrepancies are observed. The proposed Nusselt correlation which accounts for the temperature non-uniformity, cavity inclination and geometric aspect ratio is evaluated against experimental data from this and other studies. It is found to produce excellent predictions of Nusselt numbers for cylindrical cavity receivers in the Grashof number range of 2.6 × 105 to 1.4 × 107.

As concentrating solar power systems push towards higher temperatures and lower costs, it is critical that losses of overall system performance can be attributed correctly to the appropriate source. Up to now, this has been poorly done for the case of optical errors, since applicable methods do not exist to quantify how much different imperfections contribute to reducing the upper-bound efficiency of the overall system. Here, the exergy impact of varied optical design parameters—slope error, rim angle, mirror reflectance and sun-shape—is calculated for the first time. Slope error is shown to have the strongest impact. Also, dishes with rim angles significantly wider than the conventional 45° are shown to yield the best overall energy conversion. The resulting analysis method, broadly applicable in concentrating solar power, enables a new approach to quantitative optical system design.

Mixed convection around rectangular structures has great scientific value with various industrial applications. In this study, heat transfer in the mixed convection regime around a tilted rectangular cuboid is investigated experimentally and numerically. The cuboid has adiabatic wall boundaries, except for its frontal wall which is set isothermal and facing downwards when tilted. The effect of an upcoming horizontal flow on mixed convection around the cuboid was investigated focusing on flow characteristics for Reynolds numbers in the range of 10 4 Re 1.8 × 10 5 with low turbulence intensities of TI 5.5 % and Richardson numbers of 5 × 10 - 3 Ri 2 , well within the mixed convection regime. A numerical simulation based on the three-dimensional SST k - ω turbulence model, validated against data from wind tunnel experiments, was used to accurately estimate the total heat transfer rate and the average Nusselt number as a function of cuboid inclination and wind direction at different Ri and TI. The results show that convective heat transfer is enhanced at a characteristic inclination of ∼ 80 ° by ∼ 30 % in comparison to the case where the freestream flow is perpendicular to the isothermal wall. Flow behaviour in the vicinity of the cuboid demonstrated that laminar separation and turbulent reattachment were responsible for this enhancement, which is a behaviour not observed for isothermal flat plates but reported for flows around cylinders and spheres.

Temperature distribution in nonaxisymmetrically concentrating solar thermal receivers tubes is calculated for the steady-state solution with a Gauss-Seidel iteration in cylindrical coordinates. The classical plane-biharmonic thermoelastic approach to nonaxisymmetrically heated tube stress is applied. Calculation of the dominant axial thermal stress component is included. Validation is obtained with the linear-elastic thermal stress OpenFOAM® solver. Thermoelastic stress in stainless steel 316 Schedule 5S DN25 (1″) tubes containing liquid sodium is found to be 35% lower than in tubes containing molten salt. The difference is due to the higher conductivity of liquid sodium which maintains a smaller temperature difference between the front and back tube sides. A simplified thermal stress formula is shown to be erroneous if not implemented as originally documented. The sensitivity of tube thermoelastic stress to tube material and flow properties is illustrated with parameter variation, and the impact of solar concentrated heat flux is explored with some typical and ideal tube circumference flux profiles.

Sunshape and reflector surface slope error distributions are significant elements in modelling the optical behaviour of a concentrating solar power system. Different optical modelling tools implement these elements with various approaches. Discrepancies can easily accumulate in simulations of a large optical system as a result of incorrect implementations. This study reviews and verifies the implementations of these two factors in six tools that are widely used for optical modelling in solar energy research: Tonatiuh, SolTrace, Tracer, Solstice, Heliosim and SolarPILOT. The review incorporates three rounds of tests. Firstly, basic tests examine each factor carefully in simplified on-axis reflector–target configurations (round ‘A’). Secondly, off-axis effects are introduced (round ‘B’). Thirdly, full heliostat field simulations are verified (round ‘C’). All of the test cases are simulated with each modelling tool, and results are compared. Discrepancies were observed due to approximations inherent in the cone optics (convolution) methods, incorrect implementation the of pillbox slope errors, different approaches to setting the circumsolar ratio for the Buie sunshape, and different approaches to the calculation of blocking and shading losses in some tools. All issues are discussed fully, and solutions to most issues were implemented within the scope of the present study. Some remaining issues are noted. The study highlights the importance of careful implementation of these aspects of optical modelling and contributes to an improvement in the quality of several widely-used tools.

This paper presents FluxTracer, an advanced open source computer tool to assist in the analysis, design, and optimization of solar concentrators and receivers. FluxTracer is a postprocessor for Monte Carlo ray tracers used to simulate the optical behavior of solar concentrating systems. By postprocessing the rays generated by the ray tracer, FluxTracer can partition into volumetric pixels (voxels) a region of interest in three-dimensional (3D) space defined by the user and compute for each voxel the radiant power density of the concentrated solar radiation. Depending upon the set of rays provided by the ray tracer, it may be able to integrate the radiant power density in every voxel over time. The radiant energy density analysis described is just one of the analyses that FluxTracer can carry out on the set of rays generated by the ray tracer. This paper presents the main analyses that FluxTracer can provide. It also presents examples of how the information provided by FluxTracer can be used to assist in the analysis, design, and optimization of solar concentrators and receivers. FluxTracer is the first of a series of components of an open-source computational framework for the analysis, design, and optimization of solar concentrators and receiver, being developed by The Cyprus Institute (CyI) and the Australian National University (ANU).

FluxTracer is a new open source computer program being developed by the Cyprus Institute in close collaboration with the Australian National University. It is an experimental post-processor for Monte Carlo Ray Tracer (MCRT) simulations. By post-processing the rays generated by a MCRT, FluxTracer can provide information useful to analyze the distribution of the concentrated solar radiation in user-defined 3D-regions of interest within a solar concentrator. This paper presents the application of FluxTracer, to analyze the focal region of the heliostat field of the solar tower power plant considered as reference within the Australian Solar Thermal Research Initiative (ASTRI) and demonstrates the usefulness of the program in optimizing the dimensions of external cylindrical receivers or in providing guidance with regard to new receiver geometries of potential interest.

Liquid transport fuels produced from biomass are of growing importance, due to increasingly ambitious targets for CO2 emissions reduction. However, a mismatching hydrogen-to-carbon ratio in biomass feedstocks, versus that required for conventional fuels, requires that supplementary hydrogen be added or surplus carbon be removed in the production process, with many possible process configurations. Here, we consider these alternative configurations for a process incorporating a supercritical water gasification reactor, syngas reformer and downstream Fischer–Tropsch liquid fuel synthesis unit. The feedstock is microalgae, process heat is supplied using a concentrating solar-thermal collector, and additional hydrogen is supplied from photovoltaics-powered electrolysers. Using a dynamic techno-economic process model to capture solar resource dynamics, configurations are optimised for lowest produced fuel cost. Three syngas reformer types are considered: steam methane reforming (SMR), with solar heat driving the conversion of CH4 into syngas; partial oxidation/dry reforming (PO/DR), with added hydrogen instead serving that same purpose; and autothermal reforming (ATR), combining both H2 and heat. Furthermore, for SMR, both CO2 dumping and H2 addition cases are considered. At present-day 9.72 AUD/kg hydrogen costs, SMR with CO2 dumping is cheapest, yielding gasoline equivalent at 3.76 AUD/L. With cheaper hydrogen, the optimal configuration shifts to SMR with H2 addition, then ATR, then PO/DR, reaching a fuel cost of 2.99 AUD/L at H2 cost of 2.1 AUD/kg. The design of future biofuels processes will depend greatly on the cost of hydrogen.

High-temperature solar thermal energy systems make use of concentrated solar radiation to generate electricity, produce chemical fuels, and drive energy-intensive processing of materials. Heat transfer analyses are essential for system design and optimisation. This article reviews the progress, challenges and opportunities in heat transfer research as applied to high-temperature solar thermal and thermochemical energy systems. The topics discussed include fundamentals of concentrated solar energy collection, convective heat transfer in solar receivers, application of liquid metals as heat transfer media, and heat transfer in non-reacting and reacting two-phase solid–gas systems such as particle–gas flows and gas-saturated porous structures.

The present article addresses the design, mathematical modeling and analysis of a high temperature solar thermal energy storage unit based on molten soda-lime silica glass. The 126 kW hth storage unit is aimed to be used as one of the main components of a novel solar power-generation system intended for a continuous operation. The proposed design for the unit, as well as the restrictions imposed by its intended operation inside the power generation system, are thoroughly discussed. The development of the mathematical model used to calculate the efficiency and performance of the thermal storage unit during the different stages of the work cycle as well as the assumptions and simplifications made are comprehensively explained. The results obtained through the model are exhaustively analyzed. Special attention is paid to the assessment of the behavior of the storage unit to guarantee that the functional requirements are met; the performance of the unit is not evaluated exclusively from an energy standpoint, but an in-depth exergy analysis is also presented. The overall performance of the TES unit is satisfactory; the unit is capable of supplying the required 4 kWth output throughout the 16-h discharge while it reaches its fully charged state during the subsequent 8-h recharge. The proposed designed for the TES unit exhibits a round trip exergy-efficiency of 59%.

The use of concentrated solar thermal (CST) to drive endothermic processes, such as supercritical water gasification (SCWG) is an attractive option for green fuel production, and has been demonstrated at laboratory scale. However, there is a lack of understanding of the system-level challenges and economic feasibility of such a technological route. As such, this work is focused on the techno-economic analysis of algae-to-liquid fuel production via solar-driven SCWG-reforming and Fischer–Tropsch (FT) processes. A detailed steady-state physical model of the plant is developed in Aspen Plus software. Algae slurry, with 15.2 wt.% concentration of biomass, is considered as the feedstock. The total solar-thermal power delivered to the on-sun gasification and reforming reactors is assumed to be 50 MWth. There is an on-site syngas storage acting as a buffer between the intermittent SCWG-reforming and continuous FT units. The heat exchanger network is optimised for maximum syngas production, and thus liquid fuels including gasoline and diesel. The techno-economic evaluation of the system is carried out for an nth plant design for commercial units considering a plant lifetime of 30 years. The total cost of the plant includes capital investment, (fixed and variable) operating costs as well as a penalty cost for CO2 emissions. With 2016 as the basis year for costing, the proposed system has a levelised cost of fuel (LCOF) as low as 3.2 AUD/L (2.44 USD/L) of gasoline equivalent (at a solar multiple of 3.5 and 15 h syngas storage capacity) with the total capital investment of ∼160 million AUD, producing 7600 tonnes of fuel per year (∼145 bbl/day) and achieving a capacity factor (CF) of ∼70%. The sensitivity analysis of economic variables indicates that the feedstock cost, here algae, is the most influential factor, followed by the discount rate, and the capital costs of the FT, the syngas storage and the receiver unit. Although the estimated LCOF in this study seems to be relatively high compared to fossil fuel-based petroleum products, this technology has the prospect of becoming economically competitive in the near future based on the anticipated decrease in the cost of large-scale algae production and CST/FT plant scale-up.

For commercial solar tower systems of relevant size, state of the art of light collection and concentration (LCC) subsystems (heliostat field plus receiver envelope) do not achieve annual optical efficiencies larger than 65%. Achieving optical efficiencies substantially higher than that requires departing from the traditional mono-tower design. Although multi-tower systems were identified long time ago as candidates to achieve high annual optical efficiencies, so far no multi-tower concept has emerged as a clear alternative to the mono-tower system. In this article, we introduce a variant of the multi-tower concept that combines the modularity and high land coverage of linear focusing technologies with the low thermal losses and high operating temperatures of solar tower technology, and overall represents a disruptive departure from the traditional mono-tower design.

Supercritical carbon dioxide (sCO2) Brayton cycle is an emerging technology to be used as a power block with concentrated solar power (CSP) systems, tower type, sCO2 Brayton cycle has the potential to be competitive with traditional Rankine steam cycle. Most of the studies have been focused on the steady state analysis of this technology. This research has developed numerical models for five configurations of sCO2 Brayton cycles operating under quasi steady state conditions. The studied cycles are connected directly to the solar central receiver tower, which is used to provide heat input to the cycles; consequently, the heat addition is changing over time as a function of solar radiation. During the off load operation, the mass flow rate of the cycle is changing with the goal of keeping the turbine inlet temperature at 700°C. The compressor and turbine use a partial load model to adjust velocities according to the new mass flow rate. Also, the heat exchangers effectiveness are adjusted as they present off-design operation. In the recompression cycle, the model permits to explore the relationship between recompression fraction and the ambient temperature. It is demonstrated that the power generated by the cycle may be improved more than 6 % if the recompression fraction is continuously changed and controlled as a function of the ambient temperature.

This article reviews the state of the art in optical design, modeling and characterization of solar central receiver systems.

Bladed structures offer an approach to improve the efficiency of conventional concentrating solar power (CSP) cylindrical receivers, due to improved light-trapping via the cavity effect, and by allowing more tubes to be compressed into a smaller aperture, enabling the flux on the aperture to be increased without exceeding the peak flux limitation on individual tubes. In this paper, we present an optical model of a hypothetical bladed receiver mounted on the tower of the Sandia National Solar Thermal Test Facility (NSTTF). We examine the impact of receiver geometric parameters including receiver width, receiver height, number of blades, blade depth and blade angle, through analysis using ‘Tracer’, an open-source Python-based Monte Carlo ray tracing library. Validation of Tracer is provided, through comparison with results from other tools. At the optimal configuration, 15 blades with a depth of 4.5 m and angle of 63.9° from the vertical are spaced vertically over a 9.6×9.6 m back wall. In this configuration, the peak flux occurs on the back plane and is considerably lower than a corresponding flat receiver. The design-point receiver optical efficiency increases from 93.8% for a flat receiver to 98.5% for the bladed configuration, and is shown to be robust to sun position changes.