Research for this project was commissioned following flooding in 1994 and 1999. The project takes a strategic look at urban flood risk management on a large scale, examining the flood risk for entire towns in terms of spatial planning. The project also involves construction of integrated urban drainage models to incorporate existing Drainage Area Plan (DAP) models with surface water systems and small urban watercourses.
The project has been running since 2004, looking at catchment management for small, often culverted watercourses. Because the watercourses are integrated with the sewer system, they have been modeled in InfoWorks CS.
One recent question that the project team set out to answer was whether gully modeling should be included when assessing flood risk. Renfrewshire Council was concerned that flooding might occur if floodwaters were unable to enter the sewer system due to constraints at the gullies.
To resolve the issue, the consultancy built a detailed model of a very small area, including the gully connections. This test area was compared with a model that contained no gully connections, to see if there was an increase in flooding due to the presence of the gullies. The detailed model was built based on guidance provided in the CIRIA (the Construction Industry Research and Information Association) guide ’Designing for Exceedance.’
The sample area focused on a few streets, where a number of sporadically-spaced gullies were part of the sewer system. An area draining to the gullies had to be assigned, which required a great deal of local knowledge of details such as the location of kerbs. A stage area curve of the over ground topography had to be defined to replace the flood cones, and a stage discharge relationship had to be generated for each gully.
An overland flow path for bypassed gullies also had to be defined. These flows were planned such that if flows could not enter a gully for some reason, there was a defined flow route. This very detailed design was very time and data intensive, according to Fiona Dow.
The detailed model was represented with all manholes completely sealed, so that all waters would have to flow into and out of the gullies. The other model had no restrictions on the manholes.
The results were surprising, Ms Dow noted. For a 200-year extreme event the model without gullies represented showed greater flooding than the test model, which did predict some floodwaters, but generally suggested less flooding.
The reason, the modelers found, was that as the sewer system surcharged and flooded the floodwaters took time to exit the system. The gully gratings did not cause much of an obstruction, compared to the gully connection pipes. The modeling found that clean gullies have a much greater capacity for floodwaters than the connection pipes.
The flood volume increased by 15% because the water was not restricted leaving the system, a welcome factor of safety.
The event that would overwhelm the gully depends on the size of the area draining to it; larger areas require less intensive rainfall to overwhelm the gullies than smaller ones. Typically intensities of greater than 100 mm/hr would be required for the gully connections to be a constraint.
Looking back through the historical data it was realized that flooding was probably related to blockages, which are particularly hard to identify, clear and maintain within gully connection pipes.
The ability to provide a fully integrated model of pipe and watercourse systems has enabled the council to gain an overview of the interaction between the systems and identify the areas of highest flood risk, to ensure improvement carried out gains the most benefit.
Current work is being carried out to develop a method for 2D modeling of rainfall to identify flood risk in urban areas, which will remove the need to model the sewer system for strategic studies.
This article is based on a paper presented at Wallingford Software’s International User Conference in September 2007 by Fiona Dow of JBA Consulting.