Application of multivariate statistical methods for the assessment of the variability of on-site sanitation faecal sludge in Cameroon

Abstract

This study examines quality and variability of FS collected in Yaounde City. Hierarchical cluster analysis (HCA) and principal component analysis (PCA) were applied for the description of FS data obtained in the context of Yaounde (Cameroon). Almost 13 parameters for the appreciation of the FS characteristics (temperature, pH, electric conductivity, salinity, COD, BOD₅, NH₄⁺, TKN, TS, TVS, TSS, water content and total helminth eggs) were measured in 53 different samples originated from various on-sanitation systems found around the town represented by septic tanks, traditional latrines, ventilated improved pit latrines and piped equipped latrines located all around the town. FS samples were classified into 9 clusters with 72.91% of variation within the samples of the same cluster and 27.09% of variation between clusters. PCA of the whole data set indicated about 78% of the total variance with the first component accounting for 28.38% of the total variance and significantly correlated with COD, BOD₅, NH₄⁺, TKN, TS and total helminth eggs. The second component accounting for 18.54% of the total variance correlated with electric conductivity, salinity and TVS. We found that the quality of FS is significantly different (p< 0.05) between groups of latrines. The p-values obtained after the Kruskall Wallis test were 0.03, 0.02, 0.01, 0.05 and 0.002 respectively for the parameters salinity, TKN, DCO, TSS and Water content. On the base of this study, it can be concluded that PCA and HCA could be helpful for the representation and interpretation of high variable FS quality data produced in urban and rural area of developing countries.

© 2019 International Formulae Group. All rights reserved

Keywords: Faecal sludge quality, descriptive analysis, hierarchical cluster analysis, principal component analysis, on-site sanitation, developing countries