MODELLING OF FAECAL COLIFORM REMOVAL USING THE STELLA SOFTWARE. A CASE OF HORIZONTAL SUB-SURFACE FLOW CONSTRUCTED WETLAND IN TANZANIA

Abstract

A number of studies have been conducted concerning wetlands systems in East Africa; most of which focused on the performance of constructed wetlands in the removal of TSS, Org-N, NH3-N, NO3-, NO2- and BOD5. However, little attention has been given to the quantification of the contribution made by each process on the total removal of Faecal Coliform (FC). A research study was conducted to investigate the possibilities of optimising the FC Removal in Horizontal Sub-Surface Flow Constructed Wetland (HSSFCW) at University of Dar Es Salaam in Tanzania. The HSSFCW polishes the effluent from maturation ponds which form part of Waste Stabilization Pond (WSP). Modelling of FC by using STELLA program was the methodology used to achieve the main objective of this study. The modelling procedures that were considered in this study were the problem definition, conceptualisation, mathematical equations formulation, calibration, verification and sensitivity analysis. The validation was not considered because of data deficit in this study. The processes that were considered in the model were inflow, outflow, sedimentation, filtration, die-off of bacteria and growth. Also the study determined the best removal mechanism and parameter which was very sensitive to total removal of FC. Grab sampling method was used for collecting samples from the inlet, middle section and outlet and were analysed for FC (Counts per 100ml), BOD5 (mg/l), NH3-N (mg/l), DO (mg/l) and pH. The temperature (0C) and flow rate (m3/d) were also determined for the samples that were collected between February and April 2007. Other input data included solar radiation (cal per m2 per d) which was obtained from literature. All parameters mentioned above were used in the modelling processes. The observed results were; FC: 1,163,636±268,675 (inlet), 884,636±344,564 (middle) and 45,909±19,373 (outlet). BOD5: 141±35 (inlet), 119±25 (middle), and 68±11 (outlet). NH3-N: 10.92±0.47 (inlet), 9.93±0.55 (middle), and 8.01±0.64 (outlet). DO: 5.91±0.23 (inlet), 5.76±0.20 (middle) and 5.59±0.26 (outlet). pH: 6.28±0.14 (inlet), 6.67±0.37 (middle) and 7.62±0.72 (outlet). However, the temperatures: 29.77±1.63 (inlet), 30.20±1.63 (middle), and 30.64±1.57 (outlet) and Flow rates: 1.7±0.003 (inlet), 1.42±0.100 (middle) and 1.05±0.106 (outlet). For the modelling results, the removal efficiency of FC on those processes was; die-off processed: 44.164%: 3,422,203±1,129,699: Sedimentation; 5.826%: 451,433±149,914: Outflow; 0.053%: 4,075±1,796 and filtration; 0.001%: 94±31. However, the growth and inflow processes were 47.836%: 3,706,710±1,285,500 and 5.593%: 473,347±403,928 respectively. The study established that calibrated results were lower than compared to observed results by 91%. From this result, it can be concluded that the model had optimised the FC reduction in HSSFCW. Also the study established that, the growth and inflow processes increased the FC within the system while the rest reduced. On sensitivity analysis, the study noted that solar radiation was very sensitive in decreasing the FC within the system. The study further established that pH was very sensitive in increasing the FC within the system while temperature, DO, BOD5 and NH3-N were sensitive within the relative ranges of ±10%. The optimal FC was observed at 0% of the relative change of temperature, DO, BOD5 and NH3-N in this model. Hence it can be concluded that solar radiation was the best parameter that reduced the FC concentration within the system. Therefore, it can be recommended that the model be used as tool for designing and managing the HSSFCW after being validated in order to optimise the effluents quality that is being discharged to the receiving water bodies.