Author(s): Martin Sabarots Gerbec; Santiago Guizzardi

Linked Author(s): Martín Sabarots Gerbec

Keywords: Pump Stations; Numerical Modelling; Physical Modelling

Abstract: Water pollution is one of the main problems to the residents within the Matanza-Riachuelo Basin and its surroundings, in Buenos Aires, Argentina. Riachuelo river discharges highly contaminated water on the Río de la Plata and it is considered one of the most polluted rivers in the world. To ensure health and life quality of around 4,3 million of residents, the Riachuelo System project will provide a primary treatment to the waste waters from highly populated areas. The project is constituted by three sectors. The first sector involves a residential sewerage collector (16,3km long); the second part consists of an inlet pumping station (IPS), the pre-treatment plant (PP) and a second pumping station at the outlet (OPS). The last sector of the project considers a 50 m deep shaft, a 10,5 km long tunnel and a diffuser system anchored in the riverbed of Rio de La Plata. The present paper is focused on the study and optimization of the IPS designed for a 27m3/s discharge, with six operative pumps (2 pump as stand-by), and a total power of 25MW. The approach to the study is undertaken by joint numerical and physical modelling. The first is implemented to carry out a preliminary diagnosis of key hydrodynamic aspects and to develop an efficient evaluation of alternatives for the final optimization. The physical modelling is used in parallel to analyse the optimized geometry and to perform a vast range of simulations for different operation scenarios. A numerical model developed using Computation Fluid Dynamics (CFD) was implemented to diagnose the hydrodynamic behaviour of the IPS project under critical scenarios, where high concentrated streams, submerged or surface vortices, flow swirls entering the pump and non-uniform velocity distribution must be limited to achieve an optimal hydraulic performance. In addition to velocity fields post processed from CFD simulations, vortex core line detection algorithm and swirl angel (SA) were computed to identify undesired hydraulic phenomena. Ratios between SA and vortex detection indicators such as Lambda-2 and Q criteria, were obtained. The geometry primarily optimized to avoid velocities under m/s -to reduce potential sedimentation effects-, and limitation of the swirl angle. Both criteria were satisfied with an optimal design or the inlet chamber. Although the SA was reduced by 50% in the optimization process, in 2 suction pipes it results in approximately 8.5º, which is higher than the 7º admissible short-term SA. The swirl angle measured in the physical model was consistently lower than the computed from CFD simulations. Despite of the complexity of the system hydrodynamics due to the magnitude of the pumping stations, through a bifold approach including numerical and physical modelling, the design of the pumping station was be analysed and optimized to satisfy best practice criteria.

DOI: https://doi.org/10.3850/IAHR-39WC2521711920221864

Year: 2022