An enhanced model for sustainability index of water operations using risk management approach and adaptive neuro fuzzy inference system in Tangerang Indonesia [EMBARGOED]

Abstract

Water is a necessary component for all living things. Sustainability in water operations is therefore critical to providing uninterrupted water supply services. Assessment tools like water sustainability indices are important in assessing the sustainability of water operations. All existing water sustainability indices like the Canadian Water Sustainability Index (CWSI), Water Supply Sustainability Index (WSSI), Water Poverty Index (WPI), West Java Water Sustainability Index (WJWSI), and Watershed Sustainability Index (WSI) have major flaws including exclusion of important parameters and lack of parameter weightage to accurately model real situations resulting in unreliable assessments and serious implications. An example of this can be seen in the water footprint assessment. Although water footprint assessment is usable to provide a simple numerical indicator of water use, a variety of methodologies of water footprint assessment often create confusion in its estimation. This research aims to develop a model to generate more reliable water sustainability indices, especially for water operations. This model will have two different methods which are risk management methods and the ANFIS fitting model. Using this combined method a better, more structured classification of sustainability index parameters related to risks in water operations, may be developed resulting in different weightage for each sustainability index parameter. These parameters are systematically grouped in a detailed review, and the compiled results were administered as questionnaires to key experts in the water industry to identify the relative importance of each parameter in sustainability risk classification. Questionnaires of this survey were then applied to the ANFIS fitting model to create weightage for each sustainability index parameter in a newly proposed water operations sustainability index (WOSI). The proposed WOSI will then be used as a method to assess the sustainability of water service government-owned companies in Tangerang District, Indonesia, and its effectiveness compared with the existing indices. The results of the questionnaires obtained the validity tested aimed to determine the accuracy of each instrument item with a confidence level of 95 percent or alpha of 0.05 and the reliability test aims to determine whether the questionnaire used has consistency as a measuring tool or not. The combined method resulted in a more accurate measure of each criterion; environmental 0.301, technical 0.214, economic 0.280, and social 0.205, unlike previous practice with the same weightage for all criteria in the water sustainability index. Evaluating water service in 2020 of government-owned companies in Tangerang District, Indonesia using WOSI, CWSI, and WSSI resulted in sustainable levels of 86.14%, 86.79%, and 100% respectively. This value means that WOSI is more reliable because there are different weights for each criterion compared to existing water sustainability indices such as CWSI and WSSI which have the same weight value for each criterion when assessing operational water services. The study has shown that using the proposed risk classification and combining it with the ANFIS approach has resulted in a more comprehensive and reliable model to generate a water operations sustainability index, WOSI which proved the effectiveness and reliability of water operations sustainability assessment.