View Full Version : Flood type of storage nodes

Eva Choudalaki
March 21, 2017, 07:42 AM
Hi all,

I am wondering about the flood type of storage nodes since there is no option to define it. Is it automatically set to stored or lost? If it is lost do I need to manually include a flood cone in the storage array? And if it is stored how can I model it as lost? Keep in mind I am talking about a 1D model of course.

Apologies for the fairly simple question but I cannot find the answer in the help files.

Thank you for any help,


Duncan Kitts
March 22, 2017, 01:45 AM
The following blog post should cover this:-


Andy Bolden
March 23, 2017, 06:05 AM
The relevant help in ICM is in the Node Data Field page at "Flood type" and, yes, it is always difficult to remember where the help about how storage nodes behave is located.

"Note for Strage Nodes :
Unlike manholes, storage nodes do not include any alteration of the flood cone or stage / area characteristics to account for transition. If the last stage value is below ground level, the storage node is treated as sealed. If the last value is the same as ground level then it is lost. If it is above then it is stored, in which case the flood cone rises vertically and has an area equal to the last defined plan area."

The key point to note is the last sentence, i.e. storage nodes don't use flood cones but use a cylinder. So if you have storage node with 100m2 plan area within the underground tank but it has an 2m deep access shaft with a plan area of 1m2, this means that reported flooding would only be 1m3 for each 1m of depth above ground/cover level. If you have a troublesome storage tank which is liable to flood you could include a dummy manhole node alongside so as to be able to represent a flood cone on this in the usual way (based on a stated floodable area, or notional dummy subcatchment area).

If I recall correctly, back in HydroWorks days, storage nodes were treated the same as ordinary nodes in relation to mobilisation of flood cones. Whether a storage node was sealed, lost or stored was set by a flag, just as it is for nodes in ICM. It has always seemed a retrograde step that ICM disgarded this simple direct approach and required precise levels in order to set the flood type behaviour. It is so easy to change the array but to forget to change the ground level and vice versa. The result is the help has to be consulted each time it is needed because one can never remember which level determines which flood type behaviour.