This study aimed to determine the optimal conditions for organic matter removal from wastewater by Eichhornia crassipes (E.C). As a matter of fact, a complete factorial design was used to determine the effect of residence time (X1), plant density (X2) and initial chemical oxygen demand (COD) concentration (X3) on the phytoremediation process. The process's performance was measured on COD (Y1), NH4+ (Y2) and PO43- (Y3), with the results indicating a reduction of 8.59-81.71% of COD (Y1); 22.53-95.81% of NH4+ (Y2) and 0.54-99.35% of PO43- (Y3). Then, the first-order models obtained for COD, NH4+ and PO43- removal were validated using different statistical approaches such as statistical and experimental validation. Moreover, multi-response optimization was carried out through different scenarios. On the whole, the results obtained indicated that two serial ponds are required for an optimum organic matter removal by Eichhornia crassipes. Indeed, for the first pond, a residence time of 15 days is needed with a plant density of 60 ft/m2 and an initial concentration of about 944 mg/L. The second was the same residence time as the first with similar plant density of 60 ft/m2 and an initial load 192 mg/L (> 200 mg/L). Optimal organic matter removal from wastewater using Eichhornia crassipes requires two ponds arranged in chain.
Keywords: Eichhornia crassipes; Multi-response optimization; Phytoremediation.