Physical Model Tests for Seawater Intake Pumping Station

This paper presents the result of the physical model testing for seawater intake pump station located in the North coast of the Persian Gulf. The facilities were designed to provide 20,000m³/hour seawater through gas plant as cooling water of process units. The seawater intake consisted of two screening channels, each have 4.75m height and 4.0 m wide aperture. Each channel is outfitted with stop logs, bar screen and band screen. Screening channels are connected to the 15.8m long by average 18.0m wide trapezoidal shape forebay which delivers seawater to eight (8) cooling water pumps with rated capacity of 2500m³/hour each.

This is very essential to prevent a very uneven flow approach which may lead to significant pre-rotation in the pumps and therefore increases the likelihood of establishing surface and/or subsurface vortices entering the pumps. Hence, the design of the screen area and pump bay performance were evaluated through physical model testing by using a scaled model built in a Froude scaled model applying a linear scale 1:9.26.The seawater intake model was constructed mainly in plywood, however, the downstream wall is built in Plexiglas in order to facilitate observations of pre-rotation and vortex action close to the pumps. The model pumps and bellmouths were built in Plexiglas. The flow through the model pumps were established by means of siphons which allowed an accurate representation of the complex flow patterns caused by the physical geometry of the approach bay and pump bays.

The general system flow patterns and stream lines were observed and water levels in the intake pump house and flow through the individual pump intake were measured. In addition, pre-rotation in the individual pump intakes and current velocity profiles through horizontal and vertical sections andof the approaching flow in one of the pump chambers for selected load cases were monitored. The surface and subsurface vortices close to the pumps and flow conditions in the distribution chamber were observed. Some modifications were proposed to optimize the flow conditions in the intake by considering the head-losses, uniformity of the approach flow, evenness of pump throat velocity distribution, and free and subsurface vortex formation and they were physical model tested.