Characterisation and Chemical Removal of Organic Matter in South African Coloured Surface Waters
By far the highest priority for further research on the treatment of coloured water in South Africa was found to be the need for characterisation and removal of unwanted organic compounds in these waters. Little information is available on the true character and properties of the local coloured waters, and more specifically of its high variability in locality and time (spatial and temporal), as well as the many complexes that it forms with other substances, notably metals. There is also a lack of knowledge on the effect of treatment processes, and in particular coagulation, on the removal of the different constituents of the coloured water. There was, therefore, a need for a more fundamental characterisation of natural organic matter (NOM) in South African coloured waters and classifying the coloured surface water sources, and to use this for establishing the treatability of the different classes of coloured water.
Considerable work on the characterisation of organic matter in coloured waters has been done overseas, notably in the UK (Water Research Centre (Wrc); Severn Trent Water), USA (AWWA Research Foundation), Australia (Australian Water Quality Centre, CSIRO and Monash University) and Norway (Norwegian Institute of Water Research). These included land use (catchment) studies, colour and organic matter characterisation, bench-scale treatability studies and continuous flow studies. A study at the University of Cape Town has, however, shown that South African coloured waters have considerably higher colour levels than in these countries, and that especially in standing waters (such as dams and lakelets), the colour intensity is very high by international standards. The results of the NOM characterisation performed overseas can hence not be applied directly to local waters to assess its treatibility by existing processes or new processes that are being developed.
A project was therefore undertaken to characterise the natural organic matter in South African coloured surface waters and to develop operational coagulation diagrams for the removal of the organic matter, in order to improve the effectiveness and cost-efficiency of treatment of these coloured waters.
The aims of the project were as follows:
- Establishment of a coloured water characterisation and classification approach which is relevant and practical for use by water suppliers and design engineers in South Africa (i.e. development of characterisation methods and establishment of appropriate control parameters)
- Characterisation of the natural organic matter according to the adopted approach in each of the major sources (rivers, dams or lakelets) of natural coloured water in the country
- Classification of coloured waters into main types based on the characterisation results
- Performance of extensive bench-scale coagulation tests using ferric and aluminium salts, to determine the extent of removal of natural organic material and to develop operational coagulation diagrams for the main types of raw coloured waters in South Africa
- Application of characterisation data and coagulation diagrams to assess treatability of each of the main coloured water types
- Drawing up a manual on the treatment of South African coloured surface waters
The overall conclusions of this study are that, for the waters of the study area:
- Differences between the waters, apart from turbidity, lie in the amount rather than the nature of the organic content, which appears to be very similar in all the supplies. This finding has resulted in considerable simplification of the whole subject.
- Most of the organic matter has a high UV absorbance, indicating a high aromatic content.
- DOC, UV absorbance, COD and, less accurately, colour can all be used to estimate the amount of humic materials present in the water. Of these, UV measurement is recommended as being the most precise, rapid and convenient.
- Much of the organic matter can be removed by the most commonly used process of coagulation with aluminium or iron salts. There is residual organic content after this treatment, the amount of which, 1.5 to 5 mg/l DOC, is unrelated to that present in the raw water. The percentage residual is therefore generally higher in the low colour waters.
- The amount of disinfection by-products, mainly trihalomethanes, formed when the water is chlorinated is a function of both the concentration of organic matter and the chlorine dose. Generation of the maximum amount requires about 3 mg chlorine per mg of DOC. Measurement of DOC or, more simply, UV absorbance thus provides a good estimate of the potential for formation of disinfectant by-products and is an excellent tool for optimising and monitoring treatment processes.
- To minimise the formation of disinfectant by-products, treatment should aim at reducing UV absorbance to as low a level as possible.
- The required dose of coagulant is proportional to the amount of organic matter present and can be estimated from one of the measures of organic content. The preferred determinand is UV absorbance because of simplicity and rapidity of measurement, provided a UV/VIS spectrophotometer is available, and because it most accurately determines the removable humic fraction.
- A safe reliable dose for iron salts is: mg/l Fe = 30 x UV300/cm.
For aluminium an equivalent (equimolar) factor applies.
If filtered water UV254 is low, a lower dose can be considered.
- Optimum coagulation pH values are 4.6 for ferric salts and 5.6 to 5.8 for aluminium sulphate. If settled water turbidity is very high, a higher pH should be tried.
- Ferric chloride is the best coagulant for removal of humic materials, particularly with those waters having a high residual UV254 value. If the UV254 in the filtered water is low, the difference between coagulants is not very great. Aluminium sulphate gives the lowest settled water turbidity. Note that residual UV254 for a given source often varies throughout the year.
- The propensity of floc particles to stick together (sticking factor), one of the factors governing the rate of floc formation, is much the same for all the waters used for the project.
- A second floc formation factor related to water quality is floc volume. If, as mentioned above, a dose proportional to organic content is used, then the floc formation rate will be proportional to the coagulant dose. Floc thus forms more slowly in a light coloured water than in a dark one.
Based on the findings of this study, the following recommendations are made for further research on characterization and treatment of South African coloured surface waters:
- Investigate the occurrence of metals (iron, manganese and aluminium) in colour removal treatment plants, and draw up guidelines for removal and control of the metals, both at the treatment plant and in the distribution system.
- Perform desk studies on how alternative non-chemical treatment technologies can be used either together with chemical treatment or on its own to improve the quality of the final water, and be able to do this in a sustainable and affordable manner.
- Investigate the beneficial use of chemical sludges from colour removal treatment plants (research on management and use of water works sludges generally is currently being carried out by the University of Natal in Pietermaritzburg).