The new tool does not alter the way in which flood calculations are undertaken, it is instead a way of representing flood depth at manholes within a flood compartment.
Three elements are needed for flood mapping:
Flood points, which represent manhole locations;
Flood compartments (either one representing the entire area or a number of smaller polygons such that each manhole has its own compartment);
A ground model, which can be either a grid or a TIN - the finer the resolution, the better this will be. It is possible for a grid to be copied into a TIN, to enable a smoother representation.
A new toolbar contains elements that allow flood compartments to be generated automatically – this enables compartments to be merged, added manually and edited.
The compartments can be generated from selected sub-catchments and the nodes within them are selected based on the system type – a drop-down menu allows the user to specify whether this is a foul or storm sewer. Compartments can also be created within a selected boundary using Thiessen polygons in a similar way – an outer boundary polygon is created, and the solution will generate a flood compartment for each flood point/manhole.
There is also an option to manually digitize, for areas where the user is aware that flooding is likely. It is recommended that a ground theme is created on the ground model, so the low points and areas likely to flood can be identified more easily.
When flood compartments are created, particularly from existing sub-catchments, it is possible that there may be some overlap between them. Instead of manually sifting through and removing overlaps, InfoWorks CS v7.5 contains a function that is able to remove them automatically.
The new version of InfoWorks CS can also highlight flood compartments that do not contain flood points. To correct this problem, the user can select a new flood grid view, which contains a flood compartment tab and flood point, allowing the node grid to be accessed and the node ID and x and y coordinates copied and pasted into the flood point tab.
The results will depend on whether there is a single or multiple flood points within a flood compartment. Single flood points will have one flood level depicted throughout the compartment and there will be no impact on adjacent flood compartments – in other words, a boundary is created.
One key point is that the shape of the flood cone within a diagram does not represent the ground model data. It is important not to simply use default flood cones, but to consider the implications of the ground model – the flood level within a compartment will probably be higher than would be expected in reality.
It is possible to derive flood levels from a digital terrain model (DTM) so that flooding can be represented at ground level for the node or the node flood level. Overland flow routes and gully manholes should also be taken into consideration.
Testing enables theory to be matched with reality, in particular with overland flow issues. If a modeler creates a flood compartment that appears correct, it is also necessary to query whether water will actually pond at the locations where it is represented. It will in reality probably travel via an overland flow route, and the locations where flood compartments should be created would be those areas where pooling occurs in reality.
As indicated, having one flood point means that one water level is represented throughout the flood compartment, and this is compared with the ground model to provide a flood depth. One option is to create a theme using the new flood tab, so that when the model is run, the level is extended throughout the model. The key point is that if there is just one flood point, that will be extended as a constant level throughout a compartment.
Another useful feature is the solution’s ability to show ground level and flood depth at the cursor point. If a negative number is shown, this does not indicate a problem - the model only shows a flood level where this is greater than the height of the manhole.
Issues may also arise when compartments are represented incorrectly. For example, if a compartment contains only one flood point, the results will show one level extended throughout even if when examined in 3D a number of depressions can be seen within its boundary. It is necessary in such instances to question whether this would happen in reality.
For those planning to utilize the solution’s overland flow routing capability, there is an option to insert gully manholes. With these, users can specify the rate at which flow goes into the manhole – both positive and negative flows can be limited via a head:discharge curve.
This facility is useful because in reality not all flows enter gully manholes. Some floodwaters may continue downstream, so the head:discharge curve can be used to restrict the flow entering the manhole, rather than the more time-consuming method of creating a two-tier flood cone. It is also possible to define a flood storage area, which is similar to a traditional storage manhole but provides more scope to define the flood area.
For example, if a gully manhole has an assigned sub-catchment where the head:discharge relationship does not allow all of the flow from the sub-catchment to enter the manhole, surface ponding will occur. The level taken for the flooding will be that surface level, and not the water level within the sewer. If there is an overland flow link, it will be necessary to identify the downstream flow path.
Multiple flood points
Where there are multiple flood points within a flood compartment, flood levels can be predicted within it by two methods: inverse distance weighting (extrapolation) or TIN (interpolation), which creates a triangle between the points and uses the vertices of the flood compartment to extend the flood levels.
The new tool is an impressive visual way of representing flooding, without interfering with the essential hydraulic elements of the model. It is sensitive to the way in which flood cones are defined, which means that users should delineate these as thoroughly as possible.
Using this capability enables modelers to represent flood depths and set overland flood routes. Errors can also be easily identified by the validation process, which allows modelers to avoid tedious manual checking for such problems as polygon overlaps and compartments that do not contain flood points.
When the flow is the limiting factor, meaning the demand surpasses the spring flows, the valve will close, setting the flow to the desired value equal to the supply of the spring. This will provide adequate pressure in the system, corresponding to the pressure that would be created by the smaller flow at a Transfer Node.
The only limitation of this approach is that it ignores the local storage at the spring and does not calculate the amount of water that overflows.
As always when Transfer Nodes are used, the user should be careful to balance the supply to the demand in the system. This can be achieved by ensuring sufficient storage in the system or by providing alternate sources of supply from Fixed Head Nodes.