null

Message Boards

How does DELWAQ deal with sediment concentrations in cells that wet & dry?

Ben Williams, modified 6 Years ago.

How does DELWAQ deal with sediment concentrations in cells that wet & dry?

Jedi Knight Posts: 114 Join Date: 3/23/11 Recent Posts
Hi,

I have what may seem like a silly question (below). I have had a skim through the WAQ manual and process technical reference manual and I haven't found anything that jumps out at me in terms of answering it. Maybe someone on this forum might be able to provide some assistance?

DELWAQ deals with fine suspended sediment transport as a mass balance process. For a simulation dealing with the tidal zone (such as an estuary), there will be times when some cells are either dry or will be in a very shallow depth. Therefore, where the water depth becomes shallow within a grid cell, the concentration (for a given mass) must necessarily increase.

How does DELWAQ deal with calculating IM1, IM2, IM3 (all specified in the results files as gDM/m3) in these circumstances to avoid 'divide by zero' issues or circumstances where the water depth is so shallow that it would lead to artificially high IM concentrations in the water column?

Best regards,

Ben
AM
Arjen Markus, modified 6 Years ago.

RE: How does DELWAQ deal with sediment concentrations in cells that wet & d (Answer)

Jedi Knight Posts: 223 Join Date: 1/26/11 Recent Posts
It is not that silly a question, but there are several parts to it:
- Treatment of drying and flooding in general
- Treatment of processes, notably sedimentation, in very shallow segments

To answer the first one:
If a segment dries up, the flow rates at all sides are supposed to become zero. This is an indication for DELWAQ to skip the evaluation of all transport processes. Only when the flow rates become non-zero again. the transport over such sides is calculated again. All the while the volume of the segment remains constant.

This behaviour is part of the "contract" with any hydrodynamic model that DELWAQ can be coupled to. Sometimes that is a trifle difficult to realise - for instance with some of the drying/flooding optiions in Delft3D-FLOW. There are more technical details involved and we introduced a more sophisticated algorithm to detect the drying and flooding some time ago, but this is short story.

To answer the second one:
Segments that are about to become dry are very shallow and that means that there are some numerical problems to be solved. Sediment can sink to the bottom, but within one time step
there should not be more mass settling than is actually present. This could happen because the depth may be so small that within a single time step the sediment could be deposited, that is:
depth < settling velocity * time step.

The routine that takes care of settling uses a limiting condition to avoid this problem - at most the mass that is present can settle. Also if the depth is lower than a given threshold, no settling is assumed to occur at all.

Effectively, if a segment is "dry" (depth below a given threshold), it is assumed no processes are taking place. This is not always the best solution - like with Ecoli bacteria, but it avoids unrealistic concentrations.

Regards,

Arjen