Assessing Removal Kinetics of Organic Matter in Rockplant Filters

Rockplant filter wetlands are a potential alternative to supplement onsite wastewater treatment systems. Aplugflow with dispersion (PF/D) model and an ideal plugflow (IPF) model combined with firstorder kinetics were testedfor the prediction of treatment performance in rockplant filters. Six mesoscale rockplant filters growing narrowleafcattails (Typha augustifolia) were studied to verify the models with actual internal performance data. Tracer studies wereconducted during summer, fall, and spring to estimate effective values of model parameters. Firstorder kinetics wasapplicable for organic matter removal, but the effective rate constant (KT,eff) values were consistently less than empiricaldesign rate constant (KT,emp) values when plants were not dormant. The KT,eff in planted cells showed less temperaturedependence than expected, in one case varying as little as from 0.0310 to 0.0324 h1 throughout the year. This might haveresulted from the offsetting effect of prolonged hydraulic retention time caused by evapotranspiration when temperatureincreased. Both IPF and PF/D models were reasonably accurate for performance prediction, but the increased complexityin the PF/D model resulted in little improvement in performance prediction. The accuracy of the IPF model was enhancedby using effective data derived from tracer studies, which reduced least square errors up to 90% compared to using the samemodel with empirical design values. Seasonal effects were not significant when using media of smaller particle size. Plantsalso contributed to the reduction in temperature dependence in treatment performance. The results of this study verified thatthe firstorder IPF model could reasonably predict treatment performance.