"Deals with the movement of flocs in suspensions, as they appear in biological water treatment (BWT) plants."
This article states that "Large (2.9 m) plates, however, have been shown to be preferable to shorter (1.5 m) plates," even though we have successfully shortened the plates in some of the plants built in Honduras by decreasing the spacing between them.
This article states that for plate settlers, "an effective spacing is about 0.3 m, but no optimization studies are known to have been published." It also claims that plate settlers make no difference, or disrupt the system, which is contrary to what we have seen in the lab.
Brentwood Industries explains why the effluent quality is greatly improved out of sedimentation tanks when tube settlers are present.
This article explains performance in terms of a two-layer (clarified and sludge) model. Model is generalized and dimensionless and requires knowing the densities of the two layers. Recommends a spacing of 4 cm for lamella and a 2.5 cm diameter for tube settlers, but does not explain the rationale for these values.
This article claims to use a superior model that focuses on the hydraulics of the system (as opposed to describing performance in terms of mass flux); their approach may be useful for our purposes but will have to be modified to apply to cocurrent lamella. Also they apply this model to wastewater treatment, not river water treatment.
This article describes the system in terms of "sedimentation effectiveness" which mostly depends on the relative length (L/h for lamella and L/D for tube settlers). Analyzing AguaClara clarifiers in terms of this model could yield valuable spacing guidelines.
This article is frequently referenced in literature that is of interest to us. However, since it is a Master's Thesis from an Asian university, it has been difficult to acquire. This article would be a good bet for finding spacing guidelines and possibly some analysis in terms of the velocity gradients that solid particles encounter in the clarifier (could provide comparisons for our floc rollup findings).