A comprehensive study of coagulation with aluminum sulfate

Coagulation is usually the first step in most conventional drinking water treatment processes. The efficiency of all downstream processes is directly dependent on the effectiveness of the coagulation stage. Coagulants such as aluminum sulfate (alum) have been used for treating drinking water for over a century now.  Since the early 1900s, researchers have been studying coagulants like alum to understand the mechanisms by which they remove contaminants from water. Despite the significant contributions and breakthroughs by many researchers, to this day we still rely on a trial and error process to optimize coagulation. The process of optimizing coagulation is typically performed using a jar tester, a simple laboratory scale apparatus with up to six identical jars equipped with overhead mixers. The jar tester was designed to replicate the conditions of the full-scale treatment process. The six jars allow the user to simultaneously vary a given parameter while keeping other parameters constant. The approach is usually to start with a rapid mix stage (to rapidly disperse the coagulant and/or chemicals) followed by a flocculation stage (to allow the contaminants to clump together) and finally a settling stage (to separate the flocs from water). The metric that is used to compare the results is typically turbidity (a measure of the “clarity” of the water). The goal is usually to achieve the lowest turbidity possible while considering the amount of chemicals being used.   The jar test has several shortcomings, 1) it is a labor-intensive process taking between 1 – 5 hours to complete, 2) it takes between 2 – 14 tests to find a true global optimum condition, 3) there is no standardized procedure, and 4) most jar test procedures optimize for settled water turbidity while the turbidity after filtration is what is regulated in the full-scale process. 

Most of the coagulation studies done thus far were either site-specific or focused on one variable and hence do not apply to most real-world conditions. Developing a universal and practical model of coagulation has been a near impossible task because 1) water is a chemically complex medium that varies spatially and temporally 2) sheer number of factors and their interactions that determine the performance of the coagulant.   For example, a simple aggregate parameter like the pH of water not only affects the hydrolysis species of the coagulant and its solubility but also the charge on the surface of the contaminants; therefore, the overall effective alum dose required. Unfortunately, research interest in coagulation has been on a decline for the past 2 – 3 decades; yet the same questions remain unanswered since the peak of coagulation research.  Today, one can argue that coagulation exists in two separate realms, the theoretical and the practical. For example, it is generally accepted in theory that there are two primary mechanisms of coagulation, however, in practice the distinction is not easily observed or defined.

The focus of this study is to develop a general model for coagulation with aluminum sulfate that has practical applications. Given a set of raw water quality characteristics, one can conceptually predict the optimum conditions for effective treatment. The goal is also to identify the parameters that control optimum coagulation conditions when considering the removal of particulate (e.g. bacteria) and dissolved (e.g. organic matter) contaminants.  Different levels of the independent variables shown in the figure above will be controlled and their effects on the dependent variables will be quantified. The figure below summarizes the approach that will be taken for each of the waters tested. The red circles indicate the points that will be tested at a given coagulant dose and pH. A detailed explanation of the figure below can be found in Johnson and Amirtharajah's landmark paper (here). 

Contour plots of each of the dependent variables can help identify zones where coagulation is effective for particle and organic removals as well as look at the effect of pH, surface charge, etc. on the results. 

© 2020 Amir Alansari. All rights reserved.