This Is AuburnElectronic Theses and Dissertations

Numerical modeling of rapidly varying flow conditions in collection systems




Pachaly, Robson

Type of Degree

PhD Dissertation


Civil and Environmental Engineering


The design, operation and maintenance of urban water infrastructure depends on the urban runoff flow characteristics. Many modeling tools are being applied for predicting the flow characteristics and their accuracy are essential for more resilient, cost-effective, and safer operation of urban water infrastructures. Engineers and practitioners around the world face difficulties in applying such modeling tools due to the large number of models currently available, the necessary set up parameters, and the required precision to achieve the modeling goals. This research focused in applying well-known models in the context of urban drainage, aiming for improvements in their hydraulic accuracy and in more efficient applications of these models. The Stormwater Management Model (SWMM) is one of the most used tools to simulate different components of urban water systems. Typical unsteady flow conditions are well represented by SWMM, but its capability to precisely simulate more complex phenomena such as regime transition, mixed flows, closed pipe transients, and surges were unknown. The introduction of artificial spatial discretization in SWMM, by increasing the number of computational cells in each link, and the addition of the Preissmann slot pressurization algorithm have the potential to expand SWMM's applications. Hence, artificial spatial discretization and pressurization algorithms were systematically investigated using the conditions presented in the SWMM 5 Quality Assurance Program report. General improvements were achieved in terms of continuity error and numerical stability when artificial spatial discretization was introduced along with the Preissmann slot pressurization algorithm. The rapid filling of collection systems can lead to the development of fast transients, specially caused by unexpected situations such as pump failure or sudden flow blockage. Significant pressure and velocity variations may occur during these events. It was unknown whether SWMM could accurately represent such situations. For this reason, a modification for the new Preissmann slot pressurization algorithm that enforces a celerity value close to the ones anticipated in transient flows was proposed along with artificial spatial discretization. An analytical solution of a hydraulic transient and a model comparison of a real-world situation where a hydraulic transient is expected were used to assess the potential benefits of these modifications. The results demonstrated that SWMM is capable to represent certain types of hydraulic transients when set up accordingly. Stormwater tunnels under rapid filling conditions caused by intense rain events might face operational problems, such as surging. The SWMM capability to represent such situation was never investigated and the addition of artificial spatial discretization as well modifications on the Preissmann slot algorithm are expected to improve SWMM's representation of surging. Using part of the Chicago's TARP tunnel system, a combination of artificial spatial discretization and pressurization algorithms in SWMM was compared to the HAST model, which was specifically designed to represent surges in stormwater tunnels. It was shown that, with adequate model set up, SWMM can represent surging in stormwater tunnels more precisely. Urban areas tend to experience flooding events, especially during intense heavy rain events and/or when the drainage system has limited hydraulic conveyance. Combining a 1D model to represent the key hydrological aspects of the watershed and a 2D model to simulate the flooding extent would enable a better representation of flooding in urban areas as well as faster model set up. Therefore, a 1D PCSWMM was used to represent the surface hydrology and a 2D HEC RAS model was used to simulate the flooding extent based on 1D PCSWMM results. Field data was collected for calibration purposes and possible conceptual approaches that could mitigate the extent of flooding were assessed. This modeling framework predicted the flooded areas according to reported flooding events and it demonstrated that flooding depth and duration was reduced when the conceptual approaches were employed. In large stormwater tunnels, rapid filling conditions may lead to the formation of air pockets and its discharge through vertical structures can cause damages to the system. The pressure variation of uncontrolled air release in complex dropshaft structures was little known. Hence, an investigation of a multiphase rapid filling condition in a tunnel system in Columbus, OH was performed. The methodology coupled a 1D and a 3D model to determine the magnitude of surges, possibility of air pocket entrapment, air–water surging, and the consequences of uncontrolled air pocket release. Results demonstrated that proper ventilation is required to reduce the growth of air phase pressure to safe levels since the air compressibility can cause damages to the dropshaft top slab. Finally, the methodologies proposed in this dissertation improved the accuracy of flow simulation in a range of dynamic, transient, and multiphase flow conditions. We hope that the findings of this research will aid in future applications of simulating flows in collection systems, leading to better operational conditions and greater resiliency.