Drinking Water Assessment Using Statistical Analyses of AL-Muthana Water Treatment Plant

Abstract

Water is essential for survival, and controlling water quality is one of the most basic requirements for protecting this natural wealth from pollution and extinction. Statistical analysis technique was used with the SPSS v26 software to evaluate the quality of raw water period from (2016–2020), and 14 water parameters were assessed (Ka, Na, TSS, TDS, Mg, Ca, SO₄, Alk, TH, pH, Turbid, Temp, Cl, and Ec). Five principal components have eigenvalues ​​value greater than unity and explain (76.159%) of the total variance of original data set. The first component was (28.678%) of the total variance with high loading on (TH, Ca, Mg, Cl and Ka), the second component was (16.141%) with positive loading on (TSS, Turb, and Temperature), the third component was (14.826%) with positive loading on (TDS and Ec), the fourth component was (8.929%) with positive loading in (Alk and SO₄), and the last one has (7.59%) from total variance which high positive loading in (pH). The Multiple Linear Regression (MLR) results in a strong relationship between water conductivity and total suspended solid with other water parameters which the coefficient of determination (R²) values ​​were 0.963 and 0.92. The ANN model was created to forecast river turbidity based on influent TSS, Mg, TDS, and Ca. The sum of squared errors and relative errors being (0.231, 0.101) and (0.009, 0.027) respectively, respectively, the error rate in predicting the model is low, indicating that the model is successful in predicting the turbidity of the river’s raw water.