This research was directed at reducing arsenic levels in the effluents generated at the Canelands facility that manufactures monosodium methyl arsenate. Two effluent streams containing arsenic have to be considered, a raw water stream that is treated on site and a brine stream that is disposed of by sea outfall. Removal of arsenate from aqueous media by coagulation was investigated and models were developed describing selected variables that influence the removal of the arsenate. Three coagulant systems were investigated, namely aluminium(III) coagulation, iron(III) coagulation and binary mixtures of aluminium(III) and iron(III). Researchers have studied individual aluminium (III) sulphate and iron(III) chloride coagulation. No detailed research and modelling had, however, been carried out on the use of binary mixtures of aluminium (III) sulphate and iron (III) chloride coagulation of aqueous arsenate, nor had individual aluminium(III) sulphate and iron(III) chloride coagulation of arsenate been modelled at relatively high arsenate concentrations. The models that were generated were validated statistically and experimentally. The variables investigated in the aluminium(III) model included initial arsenate concentration, pH, polymeric flocculent concentration, aluminium(III) concentration and settling time. The variables modelled in the iron(III) coagulation were initial arsenate concentration, pH, polymeric flocculent concentration, and iron(III) to arsenic mole ratio. The modelling of the binary coagulant system included initial arsenate concentration, pH, iron (III) concentration, aluminium(III) concentration, and flocculent concentration as variables. The most efficient arsenic removal by coagulation was iron(III), followed by the binary mixture of aluminium(III) and iron(III) and the weakest coagulant was aluminium(III) sulphate. Scale-up coagulations performed on real raw water samples at a 50 litre volume showed that iron(III) was the most efficient coagulant (on a molar basis) followed closely by the binary mixture, while aluminium(III) coagulation was considerably weaker. The residual arsenic levels of the iron(III) and the binary coagulation systems met the effluent discharge criteria, but the aluminium coagulation system did not. Leaching tests showed that the iron(III) sludge was the most stable followed by the sludge of the binary mixture and the aluminium(III)-based sludge leached arsenic most readily. Settling rate studies showed that the flocs of the iron(III) coagulations settled the fastest, followed by binary mixture flocs and the aluminium flocs settled the slowest. The flocs of the binary mixture had the lowest volume, followed by the iron(III) flocs, while the aluminium(III) flocs were the most voluminous. Based on current operations of the raw water treatment plant the aluminium(III)-based coagulation is the most cost efficient. Given a relative costing of 1.00 for the aluminium(III) coagulation, the iron(III) chloride-based coagulation would be 2.67 times more expensive and the equimolar binary mixed aluminium(III)/iron(III) system would be 1.84 times the cost of aluminium(III) coagulation.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:nmmu/vital:10981 |
| Date | January 2004 |
| Creators | Rubidge, Gletwyn Robert |
| Publisher | Port Elizabeth Technikon, Faculty of Science |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Thesis, Doctoral, DTech (Chemistry) |
| Format | xi, 216 leaves, pdf |
| Rights | Nelson Mandela Metropolitan University |
Page generated in 0.0022 seconds