Optimising P-k-C* model for organic removal performance on H-CWs in Mediterranean climate

The P-k-C* is widely recognized as the most suitable in modeling CWs performance (Kadlec and Wallace, 2009; IWA, 2017), since it is a compromise between accuracy and computational simplicity to approximate the degradation processes for selected pollutants (Dotro et al., 2017). The kA indicates how fast is the treatment process; it depends on water temperature through the theta factor (θ) (Kadlec and Knight, 1996; Kadlec and Wallace, 2009) deriving from the Arrhenius equation. Due to the high sensitivity of this model to temperature, testing its application in different climate areas worldwide could be useful to gather more information, enabling a better comprehension. This study aims to demonstrate the applicability of P-k-C* model to describe the response of horizontal CWs (H-CWs) for domestic and agro-industrial wastewater treatment, and to evaluate key design parameters for the model optimization in Mediterranean semi-arid conditions. In particular, kA20 (m year-1) and θ values were assessed in two H-CWs located in Eastern Sicily and characterized by different design, hydraulic and organic load features. The model was assessed to simulate BOD5 and COD effluent concentrations at the outlet of the H-CW units. Calibration parameters, kA20 and θ, were found by summing and minimizing the squared differences between observed and modeled data, obtained by simultaneous adjustment of kA20 and θ for all samples (20<34). The coefficient of determination, R2, the Nash–Sutcliffe efficiency, NSE, and the root mean square error, RMSE, were used as statistical performance measures for the model. Results showed that the model provided an efficient approach to describe water quality response in terms of BOD5 and COD effluent concentrations. In particular, a worsening of the performance was observed when the influent and effluent concentrations did not differ a lot and/or when the flow was lower than the one used for the system design. Moreover, generally the model overestimated the smallest effluent concentrations and underestimated those highest. Most important finding was that a θ<1 should be used from practitioners to optimize H-CW design in Mediterranean conditions.