Physical Model Study of Isolation Gate Hydraulics
for Deep Tunnel Sewerage System (DTSS) Phase 2, Singapore

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. 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 required for detailed inspection and repair works on 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 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 undertaken the CFD modelling which 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 to assist in the design evaluation. The prototype to model length scale was set to be 31.5. Particle Image Velocity (PIV) was used for the velocity measurements. The results showed that with a rising gate and filled tunnel in the design scenario, the peak velocity would reach a maximum of ~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. Finally, recommendations were made to increase the downstream head during the lifting operation for safety against possible surges in the upstream water level.