Leaping the knowledge gap
Victorian CSO technology gave rise to some interesting hydraulic structures. Alan Wisdish from Atkins Water reports on how a novel approach was developed to represent CSOs in a hydraulic model
In a sewerage system, there are many different types of structure. These include pumping stations, storage tanks, siphons, and the ever-present combined sewer overflow (CSO). For those involved in producing drainage area plans and assessing the performance of sewers, these ancillaries provide some of the more interesting elements – and areas where modellers can demonstrate their expertise.
CSOs act as pressure relief valves, diverting excess storm flows to watercourses. Because sewer systems for many major urban areas were built in Victorian times, these structures often fall short of current standards. Early engineers were more concerned with quantities of flow rather than quality issues, and the structures reflect this. Early CSOs ranged from simple ‘hole in the wall’ systems to more complex arrangements with a low or a high-sided weir. Over time methods were introduced to address some of the quality-related issues by attempting to pass more flow forward for treatment under low-flow conditions than in storm. One of these was the leaping weir.
The leaping weir works on the principle that storm flows have a higher velocity than those in dry weather, and will therefore leap over a gap. Figure 1 shows how, under dry weather conditions, the relatively low flow falls down the opening and is passed forward for treatment. With a storm, the flows increase and, once a certain velocity is achieved, most of the flow leaps over the weir to the overflow pipe. Although the function is simple, it is difficult to represent in most standard hydraulic software.
Figure 2 shows a typical arrangement, and in showing it demonstrates that the potential exists for operators to adjust the ‘weir’ distance if the need arises. We can see too that while the hydraulics may be appropriate, the ability of the leaping weir to deal with any floating matter is less than desirable – though of course the original engineers could not have foreseen the changing nature of materials discharged into sewers.
To enable water companies to assess CSO performance in relation to specific rainfall events, mathematical models are used to represent the sewer system behaviour. The models allow the companies to replicate the current system performance and then develop outline schemes to address specific issues. These schemes can then feed into a capital maintenance plan.
Severn Trent Water appointed Atkins to prepare drainage area plans (DAPs) for five of the 20 drainage catchments serving Birmingham. Within these five catchments there were more than 50 overflows, two of which were leaping weir CSOs. All the CSOs were to be modelled. The leaping weir CSOs were among those designated as unsatisfactory by the Environment Agency.
The studies being completed for Birmingham represented only a small part of the overall DAP programme for Severn Trent. However, it was of strategic interest, given the number of overflows, and given the existence of an associated study into the impact of the overflows on a local watercourse.
For most CSOs, the performance is based on the flow depth and can be readily represented by industry standard software. For leaping weirs, however, this is not the case.
The modelling of this type of structure had not been reported elsewhere and a novel approach was required so that the performance of the ancillary could be reliably assessed with design rainfall and not just the verification events. This was particularly important, as Severn Trent Water wanted to understand the performance of the CSOs as part of their capital investment programme and for input to the Urban Pollution Management study being completed on the River Tame. The leaping weir overflows had an added complication, as the incoming pipes were egg-shaped.
The approach taken, in agreement with Severn Trent, saw the overflows being represented as the last section of the incoming sewer. The outgoing pipes were modelled so that the invert levels were the same as the incoming, but with the overflow partially blocked with silt. A sluice was modelled on the continuation pipe with control rules to reduce the area available as the flow depth in the incoming pipe increased. The dry weather flow and initial part of the storm was then allowed down the continuation pipe. As storm flows increased the depth, the area available for the continuation pipe was reduced, increasing flow in the overflow pipe. This provided a good representation of the actual performance.
The model results were calibrated and verified by observed data from flow monitors upstream and downstream of the overflows. The model predicted observed depths and flows well and hence gave Severn Trent more confidence in the model for the performance assessment under design scenarios. The approach is robust and can be repeated for similar leaping weir overflows to gain a better understanding of their performance.
According to David Terry of Severn Trent, “There are many overflows of this type in the Severn Trent region and it is good to see an innovative approach to simulating their performance”.
The modelling and research work was completed by an Atkins graduate, Guo-song Zhong, in addition to the overall model verification. With the other complex ancillaries in the catchment (vortex drops), it was this leap in the understanding of the CSO performance under verification and design storm conditions that helped to complete the project.