Concentrating Solar Power (CSP) plants unleash the power of the sun for low-cost, fossil-fuel free energy generation. Individual mirrors, called heliostats, track the movement of the sun to reflect its rays onto a central receiver where the heat is turned into electricity.
In order to minimise the risk of investments made in CSP, detailed planning, simulation and prototyping precede the execution of projects. Numerical simulation tools provide assurance that the plant will perform as expected. They can also play an important role in curtailing the need for expensive physical testing.
sbp sonne is now proud to announce sbpray, a comprehensive software suite for the creation and detailed simulation of large scale heliostat fields.
sbpray traces the path of sun rays as they are reflected onto the central receiver by individual heliostats. To provide a high degree of accuracy, each simulation involves the exact tracking of millions of rays. Yet, to keep run-times to a minimum, calculations are massively parallelised on the graphics card. Thanks to sbpray’s multi-core technology, the performance of entire heliostat fields can be accurately evaluated in a matter of seconds.
Rays are affected by various loss factors on their way to the receiver. Therefore, atmospheric attenuation, cosine, shading, blocking and spillage factors are all considered by when simulations are performed. Losses are faithfully estimated by taking into account exact heliostat and receiver geometry as well as mirror errors. A choice of varying sun models offers additional accuracy.
Since every single ray is traced, reliable predictions can be made about the overall performance as well as the performance of individual heliostats in the field. High resolution receiver flux maps constitute a welcome by-product of the detailed ray tracing. This means that heliostat aim point strategies can be simulated to ensure optimal heat distribution on the receiver surface.
The placement of heliostats around the receiver tower is crucial to the plant’s energy output. If packed too closely, they may obstruct one another in the form of casting shadows on their neighbours or blocking their reflected rays. On the other hand, heliostats placed far away from the receiver are susceptible to losses due to atmospheric attenuation and spillage. In other words, field generation seeks to find the equilibrium point between losses associated with close packing and those associated with more sprawling distributions.
sbpray has various field generation algorithms already built-in. These range from simple parametric methods such as radial stagger or Fibonacci series to more elaborate algorithms. The latter discard parametric forms but take field performance into account on-the-fly, leading to the most efficient fields to date.
Of course, sbpray also permits the import and simulation of field layouts generated elsewhere.
Previously, simplified models and very crude assumptions had to be made when optimisation and simulation where coupled. These simplifications have led optimisation algorithms onto false tracks and lead to sub-optimal overall results. Nonetheless, run-times were large such that optimisations were only possible for fields with very limited heliostat count.
This is where sbpray really shines. Its massively parallel core can perform fully-fledged analyses at high speed. This allows for an automated workflow whereby field parameters are iteratively adjusted, evaluated and improved until an optimal set-up is found.
The optimizer is not limited to stubbornly optimise the energy output of a field but all simulation data such as individual heliostat performance or the energy cast on individual receiver quadrants are available and can feed into the objective function to be optimised.
The sbpray has an easy-to-use graphical user interface. To really unleash every facet of sbpray's power, a well documented python interface is provided. This API permits the specification of limitless scenarios including arbitrary site topology, receiver or heliostat geometry. It further gives access to all simulation data for evaluation and further use in user-generated programs.
cosine: decrease of the effective reflective surface due to oblique incidence on the mirror
shading: when a heliostat casts a shadow on another one
blocking. when a heliostat blocks radiation reflected towards the receiver by another one
atmospheric attenuation: radiation that is scattered on aerosols and does not reach the receiver
spillage: radiation missing the receiver e.g. due to beam widening or tracking errors