Obstruction Detection Methodology in Generating Units in the Hydroelectric Sector
Remote access, hydroelectric plants, predictive model, obstruction and sonar
This study aims to present a methodology for obtaining SONAR (Sound, Navigation and Ranging) images in conjunction with processing techniques based on Deep Learning (DL). The aim is to analyze multifrequency SONAR images and detect obstructions in the protection grids of Generating Units (GUs) in Hydroelectric Power Plants (HPP). With this, improve repository management and energy generation. The work had as its starting point the monitoring of the transport of floating and submerged debris (trunks) during the flood period. The selected study area was the Jirau HPP, built on the Madeira River. This HPP has a reservoir with an area of 361.6 km2 at its maximum volume and an energy generation capacity of 3,750 MW. The work also verified the sedimentation of the reservoir, as the water volume directly affects the energy generation capacity. Another important factor, which reduces energy generation, is the obstruction of the protection grids by debris. These obstructions cause a punctual decrease in water velocity and consequently favor sedimentation near the water intake structures of the HPP powerhouse. Solving this problem requires the dredging process near the civil structures upstream of the powerhouse. The obstruction causes the unavailability of the GU, requiring specific equipment and specialized labor to resume the power generation operation. To monitor these processes, SONAR equipment is used to find submerged materials that obstruct the passage of water in the GUs. Thus, the study demonstrated the methodological viability of obtaining, analyzing, and processing the data collected through SONAR images. With the use of DL, it was possible to recognize patterns of obstruction of the guardrails and insert them into predictive models that informed the Obstruction Rate (OR). The main objective was to provide information to improve the management of the HPP reservoirs and, at the same time, minimize environmental impacts and avoid damage to the physical structures of the GUs. And With these actions increase the efficiency of energy maintenance of the undertaking.