Water treatment residues (WTRs) are the by-product from the production of potable water. They consist mainly of the precipitated hydrous oxides of the treatment chemicals, and materials removed from the raw water. This study investigated the range of treatment processes and residues produced in South Africa, and two WTRs were selected for testing on selected soils and mine materials. A questionnaire was developed and sent to water treatment authorities across South Africa. Information on the treatment chemicals, dosages, volumes and current disposal practices, and a sample of WTR from each treatment plant were requested. Eleven, of 21 authorities, returned completed questionnaires, representing 37 water treatment facilities. Organic polymers were the most commonly used treatment chemical, with most plants also using lime. Other less frequently used chemicals and additives were Alz(S04)3.14I-hO, Fe2(S04)3, FeC!), sodium aluminate, activated silica, activated charcoal, CO2 and bentonite. Information given regarding residue thickening and disposal was poor. Samples from Rand Water, Umgeni Water (Midmar), Midvaal Water Company, Amatola Water and Cape Metropolitan Council (Faure) were received or collected. An additional sample from Faure was also received, representing a change in the treatment process. These samples were analysed for a range of chemical and physical characteristics. These analyses showed that the WTRs had the potential to supply some plant nutrients (Ca, Mg, Fe, S) but that metal toxicity may be a problem, in particular Mn in the Faure WTR, and that P adsorption may be severe. The samples selected to test the potential for land disposal were from Rand Water and Faure. A pot experiment tested the growth of Eragrostis tefJ, Cenchrus ciliaris and Digitaria eriantha in mixtures of Rand WTR and material from a coal mine i.e., a sandy soil material, spoil material and coal combustion ash, at rates of 0, 50, 100, 200 and 400 g kg" with a uniform fertiliser treatment applied to all mixtures. The grass was harvested on three occasions and the mean total yield (dry mass) determined, as well as nutrient uptake. The pots were leached after each harvest and the pH and electrical conductivity determined. The soil, spoil and ash were characterised and pH, EC and water retention characteristics of the mixtures determined. Growth of the grasses in the ash treatments was poor and these were terminated. Eragrostis tefJ grown in the soil showed a decrease in mean total yield with increasing WTR application rate, but yield was good up to the 200 g kg" treatment at the first harvest, declining substantially by the second harvest. In general C. ciliaris and D. eriantha grown in the soil showed a decrease in mean total yield for all harvests with increasing WTR application. The yield of E. /ejJ, grown in the spoil, increased up to 100 g kg,l WTR addition, but decreased thereafter. Digitaria eriantha showed a decrease in yield, and C. ciliaris an increase, with increasing WTR application rate , but for all treatments the differences were non-significant. The pH and EC of the leachates generally increased with increasing WTR addition. The concentration of nutrients in the grasses did not indicate any deficiencies or toxicities. As the growth of grass was poor in the ash treatments, another pot experiment was established to test the growth of two creeping grass species grown in the Rand WTR as a cover over the ash material. Cynodon dactylon and Stenotaphrum secundatum were grown in 20, 40 and 60 mm layers of Rand WTR, with and without a fertiliser treatment. Both species performed best in the 60 mm layer with fertiliser, and C. dactylon performed better than S. secundatum. The former species was more tolerant of the high pH, but both have potential as cover vegetation on the ash dumps when these are covered with Rand WTR. A further glasshouse study investigated the effect of Faure WTR mixed with a nutrient poor sandy soil on the nutrient uptake and seed yield of common dry beans (Phaseolus vulgaris). The WTR was added to the soil at 0, 50, 100, 200 and 400 g kg" each with five levels of fertiliser (0, 25, 50, 100 (recommended optimum) and 150 %). Bean pods were harvested once the plants had senesced. The number of pods and mass and number of seeds per treatment were determined. The seeds were analysed for nutrient uptake. Interveinal chlorosis and necrotic lesions were evident on cotylendonous and new leaves in the WTR treated soils, the severity of the symptoms increasing with increasing rate of WTR. Additional pots were established at the 400 g kg" rate (without fertiliser) and leaf material collected for chemical analysis. This showed that Mn toxicity was the cause, with leaf concentrations about 12 times the recommended 100 mg kg" upper limit. However, mass of bean seed was highest in the 400 g kg" Faure WTR treatment with 150 % fertiliser. Nutrient translocation to the seed seemed to be relatively consistent regardless of treatment, with little accumulation ofMn. The data collected illustrated the range of conditions and types of WTRs produced in South Africa, and that in some instances these residues have favourable characteristics for land application. The use of the Rand WTR showed that it could be applied to the spoil medium at relatively high concentrations without severely negatively impacting on grass growth, but more caution should be used when applying this material to the soil medium. While the grass did not grow in the ash treatments, it would seem that with suitable species the Rand WTR could be beneficially applied to ash material as a cover layer. The use of the Faure WTR on a sandy soil seemed to potentially improve the yield of the indicator crop, but caution should be exercised due to the possibility of Mn toxicity. The use of additional fertiliser would seem to be essential. Further research would require that field scale investigation of both WTRs be conducted, as well as further studies of applicat ion rates and techniques in laboratory and glasshouse investigations.

Thesis (M.Sc.)-University of Natal, Pietermaritzburg, 2003.