Structure Equation Model for Costs and Benefits of Distribution Pipe Rehabilitation in Reactive Maintenance

Some water distribution pipe networks were constructed several decades ago and have been damaged. The old pipe network has experienced many leaks and should be rehabilitated. In old cities like Jakarta, with high leak rates (1.6 leaks/km/year), pipeline network managers must frequently repair leaks or perform reactive maintenance. Far too many pipes need to be rehabilitated compared to the available budget. This research will develop an optimization model for selecting rehabilitated pipes under reactive maintenance conditions, which has never been performed before. The selection of pipe segments to be rehabilitated must be optimal to obtain maximum benefits using minimum costs. The variables that influence rehabilitation costs and benefits and the relationship between these variables need to be known to optimize the selection. The influencing variables and their model structure were obtained from literature reviews and surveys of respondents with sufficient experience in water pipe network management. These influential variables have varying characteristics. Data processing and analysis uses the confirmatory factor analysis (CFA) method with a formative–formative measurement model. In total, 22 variables were found to be valid and significantly influenced the cost and benefits of rehabilitation. The analysis results are a structural model of the relationship between variables that influence the costs and benefits of rehabilitation in reactive maintenance with a limited budget, which can be used to optimize rehabilitation models.