Masters Theses
Date of Award
8-1986
Degree Type
Thesis
Degree Name
Master of Science
Major
Electrical Engineering
Major Professor
R. Bruce Robinson
Committee Members
Gregory Reed, Dennis Weeter
Abstract
Conventional treatment for problems associated with iron in groundwater supplies involves the physical removal of the iron from these waters. For this process, the unit operations of sedimentation and filtration are utilized. However, while this technique has been proven reliable, the associated capital, operating, and maintenance costs can be prohibitive for small low tax-based communities. Treatment by the addition of sodium silicate and sodium hypochlorite/chlorine to these iron bearing waters provides a recently exploited and relatively in expensive alternative to conventional treatment. This technique does not remove the iron from the water but sequesters or stabilizes it such that it does not settle out in the distribution system and without the usual attendant "red water" problems. However, this method has been developed largely on an empirical basis. The objective of this investigation was to establish the chemical basis for success of the silicate hypochlorite treatment and to provide associated information that would allow optimization of treatment conditions.
In this investigation, a model groundwater at pH 7.0 containing 2 mg Fe2+/L, 0 mg/L dissolved oxygen, and an alkalinity of 150 mg/L as CaCO3 added as NaHCO3, was dosed under anoxic conditions with a variety of sodium silicate solutions and sufficient 0.1% sodium hypochlorite (added separately) to oxidize all of the ferrous iron present. Where pH control was necessary, all sampling was done under a blanket of ultrapure nitrogen with the inclusion of sufficient carbon dioxide admixed to provide a pH of 7.0. The major criterion for treatment success was % iron filterability, i.e., the fraction of total iron passing a 0.1 μm membrane filter. Total iron, pH, and turbidity were also measured routinely on a set schedule of 0,1,3,5,7, and 10 days post treatment. In one experiment, filterability was based also on ultrafiltration, and samples were taken also for particulate characterization and identification by transmission electron microscope (TEM). A variety of sodium silicate types (m-ratios 3.75, 3.22, 2.5, and 1.8) and dilutions (undiluted, 1/10,1/50,1/100, and 1/500 aged 24 hours) were used for dosing (8 to 15 mg SiO2 /L). Some of these silicate solutions were characterized by reactivity coefficients, K1 and P0, based on the rate of reaction of the solutions with molybdic acid as determined from changes in optical density measured at 400 nm. Theoretically, K1 and P0 are inversely and directly related, respectively, to the level of silica polymerization in the silicate solutions.
In the preliminary phases of the investigation, the optimal sequence of chemical application was found to be silicate applied 15 to 150 seconds ahead of the hypochlorite oxidant. Also, pH control was an important experimental condition. Thus, a given silicate dosage was less effective at the lower controlled pH(7.0) than at the higher pH levels (8 to 8.3) attained in the system for which the pH was un controlled.
Analysis of the data generated from treatments based on varying m-ratios and dilutions of the dosed silicate solutions showed that for typical treatment conditions (1) higher m-ratio silicates provided more effective iron stablization and (2) higher silica concentrations in the feed solution produced better levels of iron stabilization. Thus, the t50 's, times elapsed to 50% filterability, for undiluted silicates of m-ratios 3.75, 3.22, 2.5, and 1.8 were 9.8,6.7,5.8, and 2.4 days, respectively, and the the t50 's for undiluted, 1/10,1/50,1/100, and 1/500, 3.75 ratio silicate were 9.8,8.7,6.2,5.5, and 2.1 days, respectively, for doses of 13-15 mg SiO2 /L.
The unifying interpretation of these results was that feed silicate solutions with higher levels of silica polymerization were more effective at stabilizing aqueous ferric iron. This conclusion was verified in large part by statistical analyses applied to models representing either treatment effectiveness (t50) or % iron filterability as the dependent variable in terms of the measured parameters, including the reaction coefficients P0 and to a lesser extent K1. The critical importance of time elapsed since treatment, and silicate dose was also substantiated. Detailed analysis by ultrafiItration for various treatments supported the premise that the role of the silicate in effecting iron stabilization was to disperse the "iron" particles limiting their coalescence, growth, and subsequent sedimentation. Examination of treatment particulates did not provide any consistent trends with regard to the development of crystal 1inity with time or silicate dilu tion. Where some mineral identification was possible, the best characterizations were as an unnamed iron oxide hydroxide.
The treatment mechanism was readily explained in terms of classical colloidal double layer theory. It was assumed that the treatment silicate combined with the iron particulates to produce a pHpzc lower than for the iron particulate alone, favoring the production of a net negative surface charge at treatment pH and, as a consequence, particle dispersion. This last process disfavors coagulation and sedimentation.
Recommended Citation
Browman, Michael Gordon, "Role of silica polymerization in the treatment of iron bearing groundwaters by the addition of sodium silicate and sodium hypochlorite. " Master's Thesis, University of Tennessee, 1986.
https://trace.tennessee.edu/utk_gradthes/13658