Structure bypass flow

I am undertaking some flood modelling of a large river catchment and have a query about representing flow which bypasses structures.

I understand the general approach is to use a weir in parallel to the structure (with the same chainage). However for cases such as at a weir, there is no specific change in elevation across the floodplain area. Therefore I feel that using a weir to represent out of bank flow in this instance would result in a greater head loss than is realistic.

Does anyone use a link channel to represent flow around structures, connected to the main river immediately up and downstream of the structure? Are there any limitations with this approach?

Thanks in advance for views / suggestions.

Is it possible to add sketch to your question?

I have attached a sketch of the cross-section of an example bridge and my suggested representation using a link channel. In this case the floodplain is large and you would not expect a large headloss once water levels exceed the bridge deck, therefore I don't believe using a parallel weir is justified. Instead I would create a link channel with the level-width data of the floodplain and bridge deck and link this to the main channel immediately upstream and downstream of the structure.

Any comments on this set up would be welcome. I am primarily concerned about the discretisation along the channel, making sure the link channel is connected to prevent any double counting with the cross-sections.

Hi gillianm,

MIKE 11 has a Bridge structure type which you might want to consider using. In fact under that category you even have a submerged bridge equation which might be of use to you.

I think it's common to use a culvert and weir structure on the same chainage to model flow under and over a bridge at the same time, so you could definitely try that approach. The weir doesn't have to be "parallel" to the culvert, but rather "above" it. As for the headloss, you can maniputate the values at your discretion if you believe they are too high.

Best regards,
Ricardo Machado

As Ricardo said Modelers put weir above the culvert not parallel to it.
  I have a question. If as gillianm said we have a culvert in a flat floodplain. It is clear that in case of flooding, the flood will be spread out in ae large area of the floodplain behind the  the culvert structure. If we don't have DEM for the floodpalin to use Mike 21 and Mike flood. what is the best approach to model it in mike 11?
cheers

To try to answer you question backspace2062, I don't think it's possible to simulate a floodplain if you don't have any information on it. That being said, it's not absolutely necessary to have a DEM to simulate a floodplain, although that would be the more rigorous approach to follow by using the DEM in MIKE FLOOD (MIKE 21 terrain model + M11 main river bed model) because MIKE 11 is a uni-dimensional model and therefore doesn't accurately model floodplain flow which is typically bi-dimensional. But MIKE 21 is bi-dimensional hence it is able to better model floodplains.
Anyway, in case you are using only MIKE 11 and you have information on a few cross-sections large enough to include the floodplain, you can use those to model it, but you should remember it is an approximation as in reality the flow is likely bi-dimensional.

Best regards,
Ricardo Machado