THE USE OF REAL TIME CONTROL IN A DYNAMIC MODEL TO CONTROL TUNNEL INFLOW AND DEWATERING

The Northeast Ohio Regional Sewer District (NEORSD), as part of the Easterly CSO Phase II Advanced Facilities Plan, used Real Time Control (RTC) to evaluate filling and dewatering of approximately 10 miles of large diameter storage tunnels. The tunnels were developed to store and convey Combined Sewer Overflow (CSO) to the Easterly Wastewater Treatment Plant (WWTP). Filling and dewatering the tunnel presented many hydraulic modeling challenges. These challenges included using RTC to control surge and to systematically dewater portions of the system after a storm event. This paper presents the development, testing and problems that were resolved using RTC in a dynamic hydraulic model.The large volumes and high flow rates generated a considerable amount of surge, which was analyzed using a separate surge model to determine how it could effectively be minimized. The surge model only simulated the tunnel system and not the full collection system. It was decided that limiting flow to the tunnel using automated sluice gates was the best way to minimize the surge. The specific locations for the gates were chosen based on the volume and flow rate that was entering at that point. The operational procedure for the gates was initially developed using surge modeling. The full collection systems model incorporated the recommendations from the surge model.RTC was used as part of a strategy for the dewatering of the tunnels. Three individual tunnels comprise the storage tunnel system. Two of the tunnels are connected at the downstream end where they enter a tunnel dewatering pump station. The third is at a higher elevation, and is hydraulically disconnected from the other two tunnels. The tunnels store flow until treatment capacity is available. As the tunnel is dewatered, WWTP flow increases. Fully automated control of the dewatering pump station required monitoring water level in the headworks at the Easterly WWTP that are influenced by the dewatering, which caused instabilities in the model.When the tunnel has filled the inflow control gates close to avoid surge and degradation of water quality downstream. Then as the tunnel is dewatered, the inflow control gates open and send a large volume of water to the tunnel due to the head built up behind the sluice gates. The water level would rise past the set point in the tunnel causing the inflow control gates close. Sluice gate opening and closing set points had initially been determined using the surge model; however, it was not until it was simulated in the full collection system model that it was determined it would not function as intended. Additionally, the shallower tunnel empties by gravity into the deep tunnel system only after the deep tunnels are completely dewatered. This required the use of Proportional Integral Differential (PID) control because the monitoring of downstream water levels caused gate oscillations due to fluctuating water levels.The results of this study provided a dynamic model simulation of a dewatering strategy that allows the NEORSD to protect the tunnel system from the formation of damaging surge waves, optimize monitoring locations and define operational strategies. The study also identified limitations in the capabilities of current RTC functions in hydraulic models. By incorporating these findings, the NEORSD will be able to efficiently evaluate the optimum combination of manual and automated controls in the Easterly CSO tunnel system.