| The purpose of this paper is to develope models to estimate coagulant feeding rate under the consideration of water quality variation in water treatment plant. Two year of water quality data are analysed statistically to find the con-elations among them and it was found that it is not significant to derive functional relationships between coagulant feeding rate and single wafer quality data. Accordingly, temperature, pH, alkalinity and turbidity, which represent the water qulity of influent to water treatment plant, are considered to develope the models. Five multiple linear regression models are employed and the parameters of the models are estimated by statistical analysis. The results are presented as followings ; Model I : LAS=0.2538×temperature+2.1833×pH+0.3428×alkalinity+0.2583×urbidity Model II : PAC=0.2038×temperature+2.4965×pH+0.1922×alkalinity+0.130×turbidity Model III : LAS=0.4543×temperature+0.6216×alkalinity+0.3144×turbidity Model IV : PAC=0.4622×temperature+0.5162×alkalinity+0.3187×turbidity The different water quality data are used for the verification of proposed models and the fittness of the models can be explained by the determination coefficient, which is over 0.98 between estimated and observed in modelling process and over 0.81 in verification process. Also, the models are evaluated by normality test as residual analysis. From the results, it is concluded that the developed models can be reasonable to improve the control process of coagulant feeding to get better water quality in water treatment plant. |
|
PDF
