For concentrating Solar Tower (ST) power plants, heliostats must be cleaned to maintain high productivity, but this comes at the cost of cleaning expenditures. Striking the correct balance remains challenging, due in part to the fact that soiling losses are location-dependent, stochastic, seasonal, and spatially inhomogeneous across the field. In this paper, novel reflectance-based cleaning policies are developed that trigger and prioritize cleaning of different solar field sectors based on reflectance measurements. In contrast to existing approaches, these policies have the potential to mitigate the effect of stochastic soiling losses and allocate finite cleaning resources by considering the spatial inhomogeneity of soiling. The optimization of the policy is conducted using the approximate Markov Decision Process (MDP) paradigm that utilizes a simulation model based on a recently developed physical soiling model. The proposed approach is applied to a case study on a hypothetical ST plant located in South Australia. The proposed policies are benchmarked with other traditional time-based cleaning policies and a previously developed reflectance-based policy. The results indicate a considerable benefit of sectorial reflectance-based cleaning strategies to other benchmarked policies (i.e. ∼2% savings on total cleaning costs). Moreover, in case where the per-cleaning costs (e.g. water, fuel) are significant compared to the fixed costs (e.g. truck depreciation), the savings of proposed sectorial cleaning policies are greater (∼10% savings).

Sectorial reflectance-based cleaning policy of heliostats for Solar Tower power plants

Picotti G.;Manzolini G.
2020

Abstract

For concentrating Solar Tower (ST) power plants, heliostats must be cleaned to maintain high productivity, but this comes at the cost of cleaning expenditures. Striking the correct balance remains challenging, due in part to the fact that soiling losses are location-dependent, stochastic, seasonal, and spatially inhomogeneous across the field. In this paper, novel reflectance-based cleaning policies are developed that trigger and prioritize cleaning of different solar field sectors based on reflectance measurements. In contrast to existing approaches, these policies have the potential to mitigate the effect of stochastic soiling losses and allocate finite cleaning resources by considering the spatial inhomogeneity of soiling. The optimization of the policy is conducted using the approximate Markov Decision Process (MDP) paradigm that utilizes a simulation model based on a recently developed physical soiling model. The proposed approach is applied to a case study on a hypothetical ST plant located in South Australia. The proposed policies are benchmarked with other traditional time-based cleaning policies and a previously developed reflectance-based policy. The results indicate a considerable benefit of sectorial reflectance-based cleaning strategies to other benchmarked policies (i.e. ∼2% savings on total cleaning costs). Moreover, in case where the per-cleaning costs (e.g. water, fuel) are significant compared to the fixed costs (e.g. truck depreciation), the savings of proposed sectorial cleaning policies are greater (∼10% savings).
Approximate dynamic programming
Cleaning
Concentrating solar power
Heliostat cleaning optimization
Heliostats
Reflectance
Solar tower
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11311/1161244
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 3
  • ???jsp.display-item.citation.isi??? 3
social impact