For the past several years, AY Consulting has been developing computer models for watertreatment plant operations. These models not only help plant operators control chemical feedrates more effectively, but also allow the operations staff to optimize process performance tomeet plants' specific treatment objectives. This paper summarizes the results of optimization projects directed towards identifying optimal coagulation practices at several watertreatment plants, including the following: Holland Water Treatment Plant, Holland, Michigan; South Haven Water Treatment Plant (South Haven, Michigan); St. Joseph Water Treatment Plant, (St. Joseph, Michigan); and, Fort Wayne Three Rivers Water Filtration Plant (Fort Wayne, Indiana). Among the coagulants evaluated, alum provided the least challenging feed control. It was not,however, the most effective coagulant in any of the treatment plants evaluated. On the otherhand, a combination of metal coagulant and cationic polymer provided very effectivecoagulation in all treatment plants. The polymer feed, however, had to be optimized for varyingraw water conditions in order to achieve maximum effectiveness. The effectiveness of metalcoagulant/cationic polymer premanufactured blends varied seasonally, depending on raw waterconditions. The design factors such as rapid mixing, type of clarifiers, filter configuration, plantsize, etc., influenced the choice of coagulant. The use of some coagulants also called for seasonaladjustments in flocculation speed in order to achieve maximum effectiveness.
