Mechanisms for Surges in Vertical Shafts in Stormwater Tunnels
Auburn University, USA

Abstract
Transient analyses of filling stormwater or combined sewer systems are often conducted to determine the potential for water returning to grade through vertical manhole or ventilation shafts. Although several studies have investigated interactions with air trapped during the filling process (Li and McCorquodale, 1999; Vasconcelos and Wright, 2005, 2006 and 2009; Wright et al, 2008; Zhou et al, 2002a and 2002b;), numerical models to simulate filling transients almost exclusively consider only the water phase (Capart et al, 1997; Cardle and Song, 1988; Politano et al, 2007; Vasconcelos et al, 2006). With these models, the mechanism for the return of water to grade is an inertial surge associated with the filling process, as discussed by Guo and Song (1990). The existence of air interactions allows for the possibility of other mechanisms for large water rises in vertical shafts. Previous research by the authors has indicated the ability of trapped air pockets entering vertical shafts to eject water ahead of the rising air (Lewis et al, 2010; Wright et al, 2007). This chapter presents the results of field measurements during geyser events in a stormwater tunnel system that indicate that the hydraulic grade line remained over 20 m below grade during events in which water was ejected 15 m to 20 m into the air.
This chapter also presents the results of laboratory experiments that were performed in a setup that involved the release of air pockets but was intended to minimize the development of inertial surges. These experiments were intended to investigate the effectiveness in mitigating surges of a configuration involving an expansion in the riser diameter. Observations of the water level in the riser attached to a horizontal pipeline with initial stagnant water and air injection were analyzed to determine the maximum water level in the riser. The observations indicated a fluctuating water level within the riser, with additional rise that was characterized as splash as the air burst through the water surface. Subsequent investigation indicated that the fluctuating water levels were created by inertial oscillations set up by pressure drops within the riser as the air was expelled.
Considering these laboratory observations, a relevant question is whether these inertial oscillations would also be relevant in prototype applications. This question was addressed using the numerical model by Vasconcelos et al. (2006) in an idealized scenario that was intended to reproduce initial conditions similar to those associated with the release of a large mass of air through a vertical shaft. The simulation results suggest that the occurrence of inertial oscillations driven by air release is relevant to full scale systems.
This paper is only available in PDF Format: View full text PDF