The District of Columbia Water and Sewer Authority (DC Water) was designing a large tunnel system for flood control and to capture and store combined sewage for treatment. Predicting the filling dynamics of this large tunnel system was important to prevent accidental expelling of captured sewage, ensure the safety of tunnel operators and the public, and prevent damage to sewer system and tunnel infrastructure.
LimnoTech collaborated with scholars at the University of Michigan to develop and apply an innovative model framework named SHAFT (Surge and Hydraulic Analysis for Tunnels), formulated to simulate all stages of the highly dynamic tunnel-filling process. For the DC Water system, this included simulating the formation and movement of both open-channel and pipe-filling bores during rapid filling events, as well as the locations where air can potentially be trapped and lead to geysers or other potentially hazardous phenomena. The SHAFT model was used to simulate transient hydraulic grade lines and air pocket formation for alternative tunnel geometries, using a variety of tunnel-filling and -dewatering scenarios.In the first stage of the project, LimnoTech simulated tunnel filling under critical peak flow conditions. By working interactively with the design engineering team, LimnoTech identified potential transient surge problems, evaluated a range of modifications to the tunnel geometry, and recommended a change in tunnel slope, based on SHAFT simulations that minimized transient grade line problems. SHAFT was also used to estimate air release requirements to support design.
With the first phase of the tunnel system now under construction, the SHAFT model continues to be used to evaluate surges, transients and pneumatics of other tunnel phases as their designs evolve. SHAFT model results are being used to design surge protection infrastructure for a few individual tunnel dropshafts still at risk in the tunnel-filling process, to evaluate surge and transient effects of proposed changes in the tunnel alignment, and to provide estimates of tunnel venting for filling events of various intensities.