Optimizing the design of solar power tower systems with multiple receiver

Abstract

We present a Solar Power Tower system with multiple receiver technology emphasizing the system optimization procedure. The optimization of the heliostat fieldand the receivers are addressed, involving the receivers heights, dimensions,inclinations and angular positions. A new method is developed in order to solve thisnew problem, based on an alternating greedy‐based heuristic method, already used todesign a Solar Power Tower system with one receiver.Maximizing the energy generated per unit cost, leads to a difficult high dimensionalglobal optimization problem with an objective function hard to compute andnonconvex constraints as well. The problem is reduced to two sub‐problems, mainlythe heliostats location problem and the tower‐receivers design problem. These sub‐problems are strongly dependant that is the reason to apply the alternating procedure.Regarding the heliostats location problem, when using multiple receiver technology,some difficulties appear. For each receiver, different feasible regions where theheliostats will be placed have to be considered, called sub‐fields. These regions aredetermined by the algorithm for each receivers configuration. Moreover, the shadow‐blockage calculation becomes more complicated, due to the different heliostat tiltsthat appear when near heliostats focus on different receivers. Therefore, the behaviorof the shadow‐blockage effects in the boundaries of the sub‐fields has to be carefullymodel in order to obtain a realistic solution. Note that, the common use of geometricpatterns to solve this location problem can not be applied directly due to the alreadymentioned difficulties.It is also worth noting that in the tower‐receivers design problem, the receivers areconsider independent between them, as each receiver aim to a different sub‐field, onlythe constraints to obtain a feasible solution have to be taken into account. Theproblem where the energy reached by one receiver is somehow related with theenergy reached by the remaining ones can be also studied.The main results are presented showing the viability of the alternating algorithm andthe heliostat location problem with multiple receiver systems. In Figure 1, an exampleof the sub‐field regions selection and the heliostats location for a given three receiversconfiguration, pointing to the North, East and West, can be seen.