On-line control of polymer addition to prevent massive sludge washout

An experimental method that quantifies the effect of polymer dosing on slud ge settling characteristics is proposed. This method consists of recording batch settling curves at a grid of sludge and polymer concentrations. The e ffect of the polymer was found to depend on the mixing time between the dos ing of the polymer and the start of the batch sedimentation. The recorded e ffects could be successfully implemented in a 1D dynamic settler model. Fro m the literature it was concluded that: keeping the sludge blanket below a certain critical height is an effective way of controlling the effluent sus pended solids. From a model-based analysis this strategy appeared to be sou nd. Different control strategies using, respectively, the sludge blanket he ight [feedback (FB)], the hydraulic loading [feedforward (FF)], the solids loading (FF), or the sludge volume loading (FF) were tested for their abili ty to keep the sludge blanket below the critical height. The control strate gy based on the hydraulic loading was the least efficient with respect to m inimizing polymer dosage. The others appeared equally effective provided th at they were properly tuned. Using the excess of the critical sludge blanke t height as a measure of effectiveness, strategies based on more than one m easured variable appeared to be less sensitive to suboptimal tuning.