Author(s): Adrian Law Wing Keung

Linked Author(s):

Keywords: Hydraulic structures; Physical model; CFD modelling; Cavitation; Roller gate;

Abstract: The Deep Tunnel Sewerage System (DTSS) is a core infrastructure providing a cost-effective and sustainable solution to the long term needs for used water collection, treatment, reclamation and disposal in Singapore. Implementation of DTSS comprises 2 phases. Phase 1, serving north and east of Singapore, was completed in 2008 while Phase 2, serving south and west of Singapore is in progress and to be completed by 2025. In DTSS Phase 2, one of the enhanced features in the South and Industrial Tunnels is the ability for temporary isolation of tunnel sections for the purposes of manned entry (with provision of plant, materials and equipment) required for detailed inspection and repair works on very rare occasions. This will be achieved by the placement of large temporary roller gates at upstream and downstream shafts so as to hydraulically isolate the tunnel section. When these gates are closed, the upstream head of the upstream roller gate will be surcharged up to a maximum of ~50m. The surcharged flow will then bypass the tunnel through the connecting link sewers at the higher level to the downstream tunnel section. Following the completion of the detailed inspection/repair of the isolated section of tunnel, the upstream gate will then need to be opened against the large fully surcharged upstream head which will induce extremely high velocities (up to 25m/s) under the gate and for a distance downstream in the tunnel. This raises concern about the integrity of the concrete at the shaft base itself, the secondary concrete and HDPE linings in the tunnel downstream.
The B&V+AECOM Joint Venture (JV) had earlier undertaken the Computational Fluid Dynamics (CFD) modelling of the gate opening conditions. The CFD results showed that the high velocities are dissipated at a distance of 20 to 30m downstream of the gate, however the ability of the CFD modelling to predict cavitation in close proximity to the gate is limited. Thus, a scaled physical model study was performed in additional to the CFD modelling to assist in the design evaluation, with the objectives to: (a) quantify the velocity distribution inside the tunnel during the gate operation, (b) determine the extent of tunnel entrance region that would be subjected to high velocities during the gate lifting for design protection, and (c) evaluate the potential of cavitation through direct pressure measurements in the physical model.
The undistorted physical model included a short segment of the upstream tunnel, the isolation gate section and also the entire length of the downstream tunnel (300m was used in this study). The prototype to model length scale was set to be 31.5. Particle Image Velocimetry (PIV) was used for the velocity measurements. Various scenarios with a rising gate and empty or filled tunnel were examined. In the design scenario with a rising gate and filled tunnel with 10-m downstream head, the results showed that the peak velocity would reach a maximum of approximately 24 m/s at the entrance of the tunnel and decay to less than 5 m/s after about 30 m into the tunnel, which were consistent with the CFD predictions. Furthermore, the transient wall pressures inside the tunnel recorded by wall-mounted sensors were generally above +60 kPa, indicating that there was no risk of cavitation during the gate lifting operation which can damage the tunnel lining. The full set of velocity and pressure measurements for this sceniaro as well as others with empty or filled tunnel thus provided the comprehensive assessment required to evaluate the protection of the tunnel during the isolation gate lifting operation. Finally, recommendations were made based on the model results to increase the downstream head during the lifting operation for safety against possible surges in the upstream water level.

DOI: https://doi.org/3850/38WC092019-0953

Year: 2019