The study started with characterisation of greywater and then followed by the treatment methods designed by Zuma and others (Zuma et al. 2009). Greywater was obtained from one donour household in the coastal town of Kleinemonde, Eastern Cape Province of South Africa. The treatment performance of the FLFT was tested by obtaining samples of the untreated greywater (the influent) and the treated greywater (the effluent). For sample collection, 500 or 1000 ml Schott bottles (Sigma-Aldrich, Johannesburg, South Africa) were washed in 10% hydrochloric acid (Sigma-Aldrich, Johannesburg, South Africa), commercial detergent and MilliQ water (Millipore/Merck, Port Elizabeth, South Africa). These were then sterilised in the Model RAU-53Bd REX MED autoclave (Hirayama Manufacturing, Tokyo, Japan) at 121 °C/kPa for 15 minutes.
3.4.1. Materials and consumables
The pulp and paper mill flyash was obtained from SAPPI, Mandeni, RSA; lime from the Local farm, Grahamstown (Lime 1) and P & B Limeworks, Cape Town (Lime 2), RSA. The COD (test kit number: solution A (1.14679.0495) and solution B (1.14680.0495; range 500 - 10 000 mg/l), phosphate (test kit number: 1.14848.0001), nitrates (test kit number:
1.09713.0001) , ammonium (test kit number: 1.14752.0001) and chlorides (test kit number:
1.14897.0001) were purchased from Merck (Pty.) Ltd. (Johannesburg/Cape Town, RSA). The following agars were purchased from Merck (Pty.) Ltd. (Johannesburg/Cape Town, RSA).
3.4.2. Sample collection
The samples were collected as described by Tandlich et al. (2008). After collection, it was stored in sealed 5 litre plastic drums at 15 °C and used within 24 hours of collection. All microbiological, physical and chemical analyses were completed within 48 hours of collection, while the baseline treatment assessment was measured over a two-week period.
The greywater used for the treatment performance testing was collected from the bathroom greywater and the kitchen greywater. The composition was measured over a six-week period to establish the ranges of greywater parameter values.
3.4.3. Microbiological analysis
At the laboratory, the samples of the FLFT influent and FLFT effluent were analysed for the faecal coliforms (FC) on all occasions. Faecal coliforms were enumerated on m-FC agar (Merck (Pty.) Ltd., Cape Town/ Johannesburg, South Africa), at 44.5°C for 24 hours. The mFc agar was prepared according to the manufacturer’s instructions. The mFc agar (52 g) was dissolved in 1000 mL of deionised water by boiling for 1 minute, while being stirred.
The agar was prepared in 1 L schott bottles (Labotec, Midrand, South Africa), the media autoclaving was performed in a Model RAU-53Bd REX MED autoclave (Hirayama Manufacturing, Tokyo, Japan). Samples were inoculated onto plates in a LA1200 BII laminar-flow hood. All incubations took place in one of the following incubators: the Labcon incubator Model FSIM B (Labmark, Johannesburg, South Africa), the TS 606/3-I incubator (WTW, Weilheim, Germany), the Labcon low temperature incubator LTIE 10 (Labmark, Johannesburg, South Africa) and/or the Heraeus Model FT 420 (Heraeus Kulzer GmbH, Dormagen, Germany). Membrane filtration (MF) filters (nylon, size 0.45 pm) for microbial enumeration and sterile petri dishes (90 mm) were procured from Spellboud Labs (Port Elizabeth, South Africa). Bacteria samples were either spread plated or membrane filtrated.
Samples were analysed in duplicates or triplicates. The average of the replicates was recorded as the concentration. These enumerations concentrations were expressed in CFU/100 ml and all microbiological analyses were completed within 48 hours of sample collection.
3.4.4. Microbial and Physicochemical analysis
+ 3
Greywater samples were analysed for concentrations of ammonium (NH4 ), phosphate (PO4 "
), nitrate (NO3-), chloride (Cl-), sulphates (SO42-) (catalogue number 1.14848.0001), chemical oxygen demand (COD) (test kit number: solution A; 1.14679.0495 and solution B;
1.14680.0495) and faecal coliforms (FC). Turbidity (Tur), electrical conductivity (EC) and pH of the samples were also determined. Unless stated otherwise, all quantitative analyses were performed according to Standard Methods (APHA, 1998). Besides SO4 " were determined using the relevant kits as purchased from Merck Pty. Ltd. (Johannesburg, South Africa). The COD concentrations were measured using the respective reagents and the reflux closed vessel method (APHA, 1998). Spectrophotometric measurements were performed using the Shimadzu 1200 UV/VIS spectrophotometer (Shimadzu, Johannesburg, South Africa). The EC and pH of the greywater samples were measured using the portable testers from Hannah Instruments (Port Elizabeth, South Africa). FC was determined using the membrane filtration technique (APHA, 1998) on the m-FC agar (Merck Pty. Ltd., Johannesburg, South Africa). Cells were trapped on the 47 mm GF-6 membrane filters (pore size 0.45 pm, Spellbound Labs, Port Elizabeth, South Africa). Turbidity (Tur) was measured using the Lutron TU-2016 portable turbidimeter (Test and Measurement Instruments CC, Johannesburg, South Africa). The SO4 concentrations were determined using the United States Environmental Protection Agency (EPA) method 375.4 (EP A, 2008). Sterile plastic Petri dishes were purchased from EC Labs (Port Elizabeth, RSA). Duplicate measurements for all samples were conducted.
The treatment performance was tested by obtaining samples of the untreated greywater (the influent) and the treated greywater (the effluent).The reactor was fed twice a day with 3 litres of greywater in the morning and 2 litres in the afternoon. The influent and effluent samples were collected as composite fractions of individual feedings. Analyses on the influent and effluent were performed analogically as stated for the baseline greywater characterisation.
Efficiency of treatment was assessed by calculating the percentage of removal for each parameter as defined in equation (3.1 below). The percentage removal efficiency (PR) was calculated using the following equation:
Percentage Removal effincieny (%) = (Cinfluent Ceffiu en t) x ig g Equation...3.1 Cinfluent
In equation (3.1 above), PR is the percentage removal, while Cinfluent, effluent are the concentrations or values for a given parameter in the influent (subscript influent) and the effluent (subscript effluent). The PR values are dimensionless while the units for a given parameter are listed in the results and discussion section.