AbstractThe study reported in this thesis aims to further knowledge of sediment behaviour and provides engineers with appropriate tools and methods to facilitate proactive sediment management. The techniques developed during this investigation are therefore able to predict the locations and quantities of sediment deposits through improved sediment transport methods and offer a strategy of sediment control using sediment traps where appropriate. To facilitate these methods, a number of the previous shortcomings of drainage sediment modelling required to be addressed. Most notably these included: single particle size limitation, impractical processing times; the use of purely granular analysis; and no feedback between sediment and hydraulic models.
A programme of data collection was devised to develop and test the techniques of this investigation, focussing on sediment processes in sewer pipes and sediment traps. The data collected included, suspended and near bed transport rates, long-term sediment bed development and sediment trap filling patterns. These data and previous data sets were then used in the development of analytical methods for the prediction of sediment behaviour. Field tests were also undertaken to assess the performance of an alternative form of sediment trap using partial covers. These tests highlighted the importance of good site selection for sediment traps as no improvement in trap performance could be achieved at a poor trap location through the alteration of trap form.
The resulting sediment transport models were typically developed using a combination of the historical and new short-term data sets, and were then verified using newly collected long- term data. Models were developed for rapid hydraulic simulation, sediment location prediction, sediment erosion, sediment movement, sediment deposition and trap filling. The current limitations of sediment modelling were addressed through a number of innovations which have built on the experience of CIRIA report 141. Following the development of each component model, an overall sediment prediction model was created in the SIMULINK programming environment. This combined model took the interdependency of the various submodels into account and represented interacting processes where appropriate. The full model was then tested against a new long-term data set and was then applied to a new catchment to demonstrate suitability as a sediment management tool. The implementation of the various sediment transport methods developed during this study have not only allowed the identified limitations to be overcome but have also led to the development of a model with reduced likelihood of inappropriate calibration.
In verification tests, the combined sediment prediction model was found to be able to predict the location of sediment deposits to a level of accuracy suitable for allowing operational decisions to be made. Under quantitative tests, the deposition model was found to predict depths of sediment deposits to within -5 and +10%. A high level of accuracy was also achieved in the prediction of sediment trap fill rates at three test sites. However, in each case some local calibration of the model was required.
Analysis of the behaviour of the sediment deposition model and observed data has informed the identification of influencing factors in bed development. These influencing factors of pipe gradient, initial deposit location, bed gradient reduction, increased roughness and then finally increased bed gradient, result in the formation of an “S” curve reaching an equilibrium sediment level. The erosion model tests have indicated that for practical reasons, the concept of critical erosion criteria should be re-assessed as a result of the deposition that typically follows an erosion event. It is proposed that a new critical erosion criteria should be applied that allows for a net erosion during rainfall events.
These analyses have furthered knowledge of sediment behaviour and combined with the modelling techniques devised during this investigation, can be used to confidently predict sediment behaviour in drainage systems and enable proactive sediment management.
|Date of Award||Aug 2005|
|Sponsors||Scottish Water & Engineering and Physical Sciences Research Council|