Coastal and storm water management supported through computational fluid dynamics (CFD)

Challenge

Coastal stormwater systems are an essential part of flood risk management, used during high rainfall events to divert water away from urban development. Unfortunately, the design of stormwater systems has not always considered the impact of coastal processes on their performance, in particular, the effect of sedimentation at the outfall head. The management of these structures is challenging, as they are influenced by tides, waves, and sediment movement.

For one small Australian town, its well-known sandy beaches are also the cause of ongoing sand accumulation in front of two open stormwater outfalls. This is currently managed through regular mechanical removal, shifting between 300m3 to 2,400m3 of sand annually to adjacent beaches.

Our client asked if we could simulate a novel approach to flush the sand out of the outfall by the introduction of a manual gate, minimising the time and costs spent excavating sediment.

Solution

Each stormwater channel is low lying, connected to the ocean, and regularly infilled during a high tide. The idea is that a sluice gate could be added to the lower system and shut manually during a high tide to trap a volume of water within the upstream channel.

The gate would be opened during the following low tide to release the compounded water. Water flow would be subject to relatively high velocities, which would mobilise sediment from within the channel allowing it to be transported into the nearshore zone.

Unfortunately, testing of this scenario is beyond the capabilities of standard hydraulic modelling packages like HEC-RAS, and physical modelling was considered cost-prohibitive.

JBA Idea to flush accumulated sediment
Idea – flush accumulated sediment using a new sluice gate and trapped tidal water

CFD modelling

Initial JBA computational fluid dynamics output
Initial computational fluid dynamics output

Using survey data, we constructed a three-dimensional Computational Fluid Dynamics (CFD) model of the channel, outlet and nearshore. Using tidal planes, a mean high-water spring would enable approximately 0.6m of water to be trapped within the channel after the sluice gate was closed. A number of sand accumulation profiles (sloping and flat) were introduced in the model, each with a sand grain diameter of 0.28mm.

Each model run started with the impounded water being released as if a gate was opened. The outflowing water then scoured away the accumulated sand profiles that were tested.

The images are looking upstream from the coastline towards the new gate. It shows the evolution of the accumulated sediment over time, just before the gate is opened, as scour begins, and following the release of the trapped water.

  • Image (a) shows the initial condition of the model
  • Image (b) shows the release of water, with the sand slug starting to
    change shape
  • Images (c) shows the continuing scour in the channel
  • Image (d) shows the final channel, with around 10m3 sediment having
    been cleared.

Benefit

The modelling shows the potential for new sluice gates to be installed upstream of coastal stormwater outfalls to allow manual clearing within the use of excavation plant.

Whilst each site will be different, subject to a range of sedimentation rates, tidal properties and grain sizes, the locations with the greatest potential for new ‘flush’ gates can be assessed easily through CFD.

Want to know more?

Email Andy Collier or Kate Bradbrook or visit our Computational Fluid Dynamics web page for more information on CFD modelling.



2 Comments

  • Jingchun Wang

    Hi, I wonder how do you describe the motions of sand. Do you use an additional particle tracking model or something similar?

    • Sophie Bunker

      Hi Jingchun,

      Thanks for your comment.

      We are using a few three-dimensional fluid dynamics packages, but this particular problem was modelled using Flow3D. This software has a sediment scour solver, which represents the motion of the sand (or a mix of up to 10 particle sizes) due to shear stress. This shows a realistic evolution of a river bed, coastal region, or in this case, block of accumulated sand.

      We have also performed a flume based physical test and this showed a very similar evolution of the erodible sand block.

      If you would like more information on other ways that we are applying the scour solver, or you are interested in other applications for 3D CFD modelling, we would be happy to discuss these further. Please email Andrew Collier on andrew.collier@jbaconsulting.com

Leave a Reply