Cohesive sediments

To obtain representative settling velocities for fine cohesive sediments, water/sediment samples should be collected and a settling velocity distribution should be determined from them.  Ideally, the samples should be collected during periods of high river flow, and the settling velocity distribution should be determined in-situ, rather than after the sediment in the sample has been allowed to settle and has been re-suspended in a laboratory.

Concentration and settling velocity data for cohesive sediments may be difficult and costly to obtain.  In some cases no field data will be available and values will have be to estimated, and then probably adjusted during model calibration.  In the absence of any data, the settling velocities should be set initially in the range 0.005mm/s to 3mm/s.

The number of sizes chosen to represent the cohesive sediments depends on the distribution of their settling velocities.  If there is a small range of settling velocity then a single size may be sufficient, however this would be unlikely.   It is recommended that two or three fractions are used in the initial simulations. If an acceptable calibration cannot be achieved, then further fractions can be introduced.

When cohesive sediments are being modelled use of the SORTED algorithm is strongly recommended.  Alternatively a single, relatively coarse, fraction can be taken (settling velocity above 1mm/s) or erosion can be prevented by setting very high values for threshold shear stress (>200). If none of these measures are adopted, then there will be no mechanism to prevent erosion of the finest fractions predicting a strongly degrading bed, despite there being no material in the bed of that size.

Values for the threshold shear stress in N/m2, tthresh, for erosion can be derived from the following formula given in HR Wallingford (1992)

t_thresh =   0.0012 den^1.2

Where 'den' is the density of the deposits, given as kg of dry material per m³. In cases with good consolidation, such as when a channel dries out for extended periods, values above 10 N/m² can be specified. Values around 1 N/m² would apply for poor consolidation, such as when a few metres of deposition occur in a few weeks without the channel drying out. 2 N/m² would be a typical value.