Kempton SSF

Client: Atkins
Scale: 1/3.5


Further to view of the supporting information it was determined that filter beds 1-6 were experiencing erosion with the remaining filter beds 7-12 not experiencing any issues.  Detailed review of the arrangements determined that the butterfly valves installed within the inlet pipe on filter beds differed in orientation i.e., the leading edge of the valve either was located on the soffit of the pipe (filter beds 1-6) or conversely on the invert of the pipe (filter beds 7-12).

The testing programme concluded that the valve orientation and highly turbulent flow directed to the free surface of the inlet/outlet chamber on sand filter beds 1-6 produced localised erosion zones obliquely positioned entering the sand filter.

On discharge over the front wall, the observed high velocity jets had enough kinetic energy to expand vertically and descend the water volume with subsequent interaction with the sand media.  The jets were characterised by sufficient velocities that significant erosion occurred with localised mounding and radiating contours of the sand media. 

It was concluded that the high velocity jets and the multiple and continual eddies produced by flow entering the inlet/outlet chamber along the surface of the retained water volume were the contributing factor to the observed erosion on the physical model and onsite sand filter beds 1-6.

In order to categorially conclude that the valve orientation was the main contributing factor to the erosion issues witnessed onsite, the testing programme altered the valve orientation to be in line with filter beds 7-12. 

Observations determined that the previous prominent zones of erosion were not visible with only minor contour mounds noted along the sand media.  

Mitigation solutions

The testing programme installed three development proposals on the physical model to suppress and reduce the erosion recorded on the sand media on filter beds 1-6, these are detailed within the following commentary.  The initial proposal is currently installed on site and is considered to be successful in suppressing the evidenced erosion.  The proposal is a 300mm upstand wall along the front of the inlet/outlet chamber.  The testing programme commenced at the BWL with incremental increase in inflow from the minimum 170 l/s to the maximum 506 l/s.  Over a considered long time period the model was ceased with abstraction of the retained water volume.  The testing programme concluded that minimal erosion was observed with no discernible depressions throughout the sand media.

The model was incrementally increased in water volume to the TWL 14.25 mAOD with the same accumulative increase in inflow replicated.  Over comparable time steps the model was drained with re-assessment of the sand media.  The testing programme concluded that minimal erosion was observed with no discernible depressions throughout the sand media, hence, it was concluded that the 300 mm upstand wall was effective in suppressing significant erosion from occurring.

The client requested a secondary option to the 300mm upstand wall, as such, the testing programme proposed a horizontal plate extended 300mm into the inlet outlet chamber.  The testing programme re-evaluated the filter bed as per the above processes.  The results were comparable to the 300mm upstand wall and was considered to represent an alternative options to reducing the observed erosion onsite.

Supplementary evaluation

A further two designs were evaluated on the physical model to try and eliminate all evidenced zones of erosion throughout the sand media, these are detailed below.

The client requested that a further investigation into reducing the evidenced minor erosion evident local to the stop-log openings with the 300mm upstand wall installed.  Consequently, the testing programme studied a small increase in height within the stop-log openings with an 150mm invert insert installed.  The valve orientation was leading edge located along the soffit of the inlet pipe i.e., sand filter beds 1-6.

On drain down the testing programme concluded that minimal erosion was evident within the sand media.  As per the previous modifications, an increase in erosion occurred local to the stop log openings however, the position had slightly altered with the zone migrating closer to the rear of the filter bed.  It was considered that this is an inevitable feature of the re-directing the flow regime within the inlet/outlet chamber.

Consequently, a further alteration was undertaken with the inlet bellmouth removed and replaced with a 22 degree bend orientated to directed flow towards the invert of the chamber.  On drain down the testing programme concluded that minimal erosion was evident within the sand media.  No noticeable zones of depressions were evident within the sand media with only minor contouring of the sand observed.

It was concluded that that the 22 degree bend installed on the exit of the inlet pipe suppressed the erosion evident throughout the testing programme with no evidence of any significant depressions of the sand media throughout the model reach.

Consequently, it was considered that this development option was the preferred alteration to the structure given the ease of install and evidence of minimal erosion of the sand media within the physical model.