Fuzzified hedging rules for a reservoir in the Brazilian semiarid region

Abstract

This study quantitatively evaluated different operating rules that were optimized with a genetic algorithm using a water budget simulation model. The study area was a reservoir located in the Brazilian semiarid region, which has great hydrological variability and severe droughts. In this context, the performances of standard operating policy (SOP), simple hedging rule (HR), and fuzzified hedging rule (FHR) were assessed. The objective was to maximize the volumetric reliability of meeting water demands during the dry season, a period with high water demand. Indicators such as volumetric reliability, resilience, vulnerability, and dead storage timeline were analyzed. The results showed that, in decreasing order, FHR, HR, and SOP varied in their effective mitigation of water scarcity during the dry season. FHR provided the greatest volumetric reliability for priority demands in both the dry season and the wet season, a significant reduction in vulnerability, and shorter durations of dead storage. However, HR and FHR induced more frequent supply rationing than SOP. The results confirm that predicting water rationing measures is crucial to mitigating continuous and severe water scarcity.