View Full Version : Fixed runoff depression storage or runoff coefficient reduction

Andy King
March 5, 2013, 01:42 AM

We are using the fixed runoff model to represent paved and roof surfaces defined in OS Mastermap, I would like to know the current best practice for using fixed runoff. The model will not be verified against flow survey and the 2D model will be used to determine flood risk to properties.

For impervious areas such as paved car parks, roads etc; is it better to make an assessment of depression storage for different types of paved area and adjust the Initial Loss Value appropriately and apply 100% runoff from that surface? This must be the case once all puddles and cracks etc have been filled?

Or should the default values be used for depression storage and the runoff coefficient adjusted between 70-90% as per the help guidance?

Following the rainfall events of last year, surfaces that were previously assumed to be permeable were observed to contribute 100% runoff. With the use of 2D models, we would like to see where runoff that is not collected by the sewer network ends up. Hence the concern about adjusting the percentage contribution to be less than 100%.

Is there any recommendation for runoff from ‘pervious areas’, which in many cases last year were observed to contribute large volumes of runoff? The help simply suggests a value between 0 and 10% of the total area. Presumably different uses of pervious area and soil type will have different saturation depths and therefore are to irregular to model with any degree of confidence and would have to be modelled on a case by case basis rather than by rule of thumb?

Thanks, your suggestions will be much appreciated.


Kristian Ravnkilde
March 6, 2013, 02:18 AM

Andy King
March 6, 2013, 05:00 AM
We are using IWCS

Kristian Ravnkilde
March 6, 2013, 05:31 AM
In that case, you're limited as to what you can do with rainfall direct to the mesh (only a percentage is available), and I'd go with 100% on the assumption that depression storage has been taken up and permeable surfaces are saturated. This is a worst case. You can get around the initial loss issue with rainfall direct to mesh by creating a specific rainfall profile for the mesh only, with the first Xmm of rainfall "chopped off" (a bit of basic arithmetic before editing the first few timesteps) OR with an evaporation profile to achieve the same thing (NB the evaporation profile is in mm/day, so watch your units).
ICM is much more sophisticated, with even more ways to confuse yourself!
However, it sounds like you are also using traditional subcatchments, in which case you can either go with Initial Loss Type = Slope and use the default values, OR = Absolute and put in values in m (I find an initial loss of e.g. 0.1mm = 0.0001m far easier to understand, and easier to understand the impact if I change it). You can then use 100% fixed runoff.
For permeable surfaces, you could have a higher initial loss, and possibly reduced% runoff on the basis that some will infiltrate - depends how much of a worst case you want to create.
Whatever you decide, put it as clearly as you can in the report so thet the client/ others can understand what you have done and the uncertainty/ risks involved. I generally find that a degree of conservatism is preferable.

Andy King
March 6, 2013, 06:41 AM
Thanks for your detailed reply. We had been using the default wallingford runoff model which I think applies 100% runoff. So I think when we use the fixed PR model we will go with 100% runoff and adjust the initial loss. As you suggest it is easier to understand when set to Absolute. Thanks again