The treatment of potable water in Vars, ON is accomplished by filtering the colored, iron-laden groundwater through granular activated carbon (GAC) filters. When first installed, these filters unexpectedly experienced chromatographic displacements of iron into the produced water which resulted in orange-brown water at consumers’ taps. The treatment plant was later modified by adding potassium permanganate oxidation and a greensand filter prior to the GAC adsorption columns. Consequently, iron was almost completely removed and no longer caused operational problems. The main objective of this dissertation is to study the interactions between natural organic matter (NOM) and iron that caused the observed chromatographic effect. This study was divided into three main stages: a) characterization study on Vars groundwater and its treatment system; b) study of the competitive adsorption of iron with NOM in Vars groundwater; and c) evaluation of the rapid small-scale column test (RSSCT) for predicting the full-scale GAC column breakthroughs. The characterization of Vars groundwater showed that ferrous iron was found to be the dominant iron species, representing 90% of the total iron, and that 15 - 35% of the iron was complexed with NOM. It was hypothesized that the chromatographic displacement of iron from the GAC columns was caused by NOM-iron complexes; however, field mini-column experiments showed this was not the case. Thus, competitive adsorption between iron and NOM was seen as the more likely cause of the chromatographic effect. The adsorption capacity of ferrous iron in Vars raw water was less than that in organic-free water by a factor of 7 due to the competition with NOM over the GAC adsorbing sites. However, the NOM adsorption capacity was not reduced due to the presence of ferrous iron. It was hypothesized that ideal adsorption solution theory (IAST) models, which have been successful in describing competitive adsorption between target organic compounds and NOM, could model the competition between an inorganic compound such as ferrous iron and NOM. The hypothesis was proved to be correct, and the adsorption isotherm of iron in competition with NOM in Vars groundwater was simulated very well by several versions of the IAST model. However, none of the models were capable of simulating the competitive adsorption of NOM and ferrous iron simultaneously. Since the presence of iron did not significantly reduce the adsorption capacity of NOM, a simplified approach of using the single-solute NOM isotherm to represent the competitive NOM isotherm was recommended. The performance of the rapid small-scale column test (RSSCT) was evaluated in order to simulate the iron chromatographic effect observed at Vars’ full-scale GAC column. The RSSCT was not capable of predicting the iron phenomenon and the test proved to be problematic due to the oxidation and precipitation of iron within the small voids between the small-scale column’s GAC particles. The RSSCT, using constant and linear diffusivities, were applied to simulate the NOM adsorption after greensand treatment. Integrating both diffusivities, the tests predicted the onset and slope of the NOM breakthrough up to 10-L water treated/g GAC, which is equivalent to 250 days of operation time for the full-scale column. However, the NOM breakthroughs deviated beyond that point and the RSSCT using constant diffusivity underestimated the column performance greatly. On the other hand, the linear diffusivity RSSCT underestimated the performance to a lesser degree and its NOM breakthrough was quite parallel to the full-scale performance with lower NOM removals of 15%. The higher long-term NOM removal in the full-scale system may be explained by biodegradation, a phenomenon that was not considered by the short duration of RSSCT.
| Identifer | oai:union.ndltd.org:uottawa.ca/oai:ruor.uottawa.ca:10393/23349 |
| Date | January 2012 |
| Creators | Al-Attas, Omar |
| Contributors | Narbaitz, Roberto |
| Publisher | Université d'Ottawa / University of Ottawa |
| Source Sets | Université d’Ottawa |
| Language | English |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0016 seconds