Fitting probability distributions to component water quality data from a treatment plant

Abstract

The treatment of water is carried out to make the available water meet the standards for its intended use. Such use may be for drinking and other household needs, industries,livestock rearing or fisheries etc. poor quality water is commonly treated to ensure potability. Potable water should be free from unpleasant tastes and odour, have a good appearance and be suitable for domestic as well as a wide range of industrial purposes. Continuous statistical distributions are usually applied to engineering situations. A goodness-of-fit test is usually necessary to determine the fitness of a distribution to specific data. The Kolmogorov-Smirnov test which is a widely used goodness-of-fit measure was used in the work. A total of four continuous distributions namely normal, log normal, gamma and weibuill were employed for data obtained from effluents of coagulation/clarifying process, (settled water), filtration (filtered water) as well as chlorination process (final water). For the settled water, colour and turbidity are lognormally distributed while pH is normally distributed; filtered water has colour, turbidity, free chlorine and total chlorine log normally distributed while pH has gamma distribution for best fit. The log normal distribution fitted all the parameters undertaken for the final water.

Keywords: water, treatment plant, clarifier, filter, distributions

Global Journal of Environmental Sciences Vol. 4(2) 2005: 151-154