A DEA cross-efficiency inclusive methodology for assessing water quality: A Composite Water Quality Index

Abstract

• An inclusive DEA-based approach to compute a Composite Water Quality Index (CWQI). • CWQI fully ranks water sources via observed physicochemical factors. • OWA operator incorporates practical water treatment conditions into CWQI. • k-Means clustering analysis used to derive water quality ranges. • The new methodology is applied on 47 dams located in the Tellian region, Algeria. This paper introduces a new index, identified as Composite Water Quality Index (CWQI), for assessing water quality. The novelty of CWQI is rooted in the practical significance of the methodological approach that is developed for its computation. The CWQI is computed within an inclusive framework that integrates data envelopment analysis (DEA) Cross Efficiency (CE) and the Ordered Weighted Averaging (OWA) operator, using Optimistic Closeness Values (OCVs) as input variables. The OCV, which measures the potential of a water quality parameter to reach its best quality status, sets a solid preliminary ground for the assessment process. The DEA-CE approach enables a collective evaluation of the water quality, which bestows more inclusiveness on the quality assessment process and, hence, more robustness of the CWQI. The OWA operator extends the standard role of CWQI, as solely a water quality measurement device, to incorporate the practical conditions of water treatment for future decision plans. The new methodology has been applied on a sample of 47 dams, described with 10 physicochemical parameters, located in Northern Algeria. Adopting a wide range of water treatment conditions, the results reveal “Kissir” and “Bougara” as the best and the worst water sources, respectively. Meanwhile, the ranking patterns of the dams are found almost the same. The k-means clustering identified the Oranie–Chott–Chergui (OCC) basin as the poorest water quality zone and Algerois–Hodna–Sommam (AHS) basin as the best.