Calibração de modelo hidráulico para otimizar o abastecimento de água industrial

Abstract

Water Supply Systems (WSS) are essential for factory operations, and understanding hydraulic variables such as pressure and flow is crucial for optimizing the distribution network. This study focuses on the efficient management of a flat glass factory operating 24 hours a day, where water ensures product quality. However, the lack of monitoring sensors makes a detailed system analysis challenging. In this context, hydraulic modeling becomes an essential tool for characterizing the supply network and improving its operation. The objective was to calibrate a hydraulic model of the factory, reducing the difference between simulated and observed flows. The modeling was conducted using specific software, EPANET, and calibration was performed using Evolutionary Algorithms (EA) to improve accuracy. The adopted criterion was to maintain flow errors below 3\%. The results indicated that EA were effective, with errors below 2\% for flow, 1\% for pressure, and 5\% for head loss. The research demonstrates that calibration using metaheuristic techniques optimizes system operation, enhances water management, and contributes to strategic decision-making in the supply network.