Date of Award

Spring 2005

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Environmental Engineering - (Ph.D.)

Department

Civil and Environmental Engineering

First Advisor

Mohamed E. Labib

Second Advisor

Taha F. Marhaba

Third Advisor

S. S. Dukhin

Fourth Advisor

Thomas B. Atherholt

Fifth Advisor

Robert Dresnack

Sixth Advisor

Jay N. Meegoda

Abstract

Recent monitoring studies have indicated that many ground water (GW) sources would benefit from the development of effective technologies for removing viruses. Although reverse osmosis can achieve high log removal it is not economical. The main objective of this study is to test and validate a novel adsorption filtration (AF) technology for removing viruses from GW sources for drinking water production in New Jersey.

The convective diffusion of viruses to adsorbent particle surfaces in flow packed beds (FPB) enhances their removal efficiency if the adsorbent particle size can be decreased to about 100 microns. The development of effective virus removal technologies has not been successful because of two fundamental difficulties. First, the pressure drop in FPBs increases very rapidly with decreasing adsorbent particle size, and becomes unacceptably high at about 100 microns. Second, additional difficulties in achieving high log virus removal can arise due to competition for adsorption sites between negatively-charged viruses and humic acid anions which are fractions of Natural Organic Matter (NOM).

The two main challenges to be overcome for developing an effective adsorption filtration technology for virus removal are: i) the selection of an appropriate adsorption medium, and ii) the development of means to minimize the effect of NOM on virus removal efficiency. Based on systematic experimental studies, calcite was selected as the adsorbent for virus removal. In order to overcome the adverse effects of NOM, a special prefilter, to be placed upstream of the virus filter, was developed using large calcite particle size, about 600 microns. This was possible because NOM molecules are approximately 15 times smaller than viruses, and accordingly, their diffusivity is about 15 times larger. Consequently, NOM diffusional transport is sufficiently rapid in spite of the larger calcite particles used in the prefilter. This ability to utilize such larger particles in the NOM prefilter made it possible to use a small prefilter pressure head during filtration.

In order to decrease the head loss arising from the use of small particle size dimension in the virus filter, two sets of channel arrays were introduced in the flow packed bed design similar to the concepts used in membrane technology, namely: decreasing the filtration depth and expanding the filtration surface area. In the design, the filtration surface area was increased and the filtration depth was decreased so that the same footprint and volumetric filtration velocity of the conventional FPB could be accomplished. In the AF filter, virus-contaminated water enters the feeding channels, crosses the adsorbent layer and then exits along the receiving channels. Therefore, this design allows efficient virus removal (> 6 logs) using 100-micron particles without the need to use high pressures to drive the filtration process.

The experimental studies confirmed that a model composite NOM/virus filter can remove more that 6 logs of viruses in the presence of NOM. The composite filter run was about 100 hours, compared to only 10 hours without the NOM prefilter. The design of the composite NOM/virus filter with a linear velocity of 0.1 cmlsec, P 0.1 bar, was elaborated for scaling up to an industrial scale.

Share

COinS