Our research focuses on understanding and improving the removal of pathogens in the biosand filter, a form of point-of-use water treatment used in 300,00-500,000 homes worldwide. Currently our tests focus on using iron particles mixed throughout the sand media of the filter to improve the removal of viruses in the filter.
Under the direction of Professor Helen Nguyen and research leads Ian Bradley and Tony Straub, 18 undergraduate researchers are currently performing research related to iron-amended BioSand filters. Current experiments have been running for over 400 days, and on-site testing in Guatemala will begin in Spring 2011. Our activities are supported by EPA P3 phase 1 (SU834296) and phase 2 (SU834754) grants and NSF Career grant (0954501) to our advisor Professor Helen Nguyen. We also received support from WEF Central States and College of Engineering at University of Illinois.
The experiments have demonstrated that iron oxides, specifically small iron particles, can improve the virus removal in biosand filters from 0.5 log (~70%) to at least 5 log (99.999%) meeting WHO and EPA health standards. The research has also given evidence that the after almost year, the standard biosand filter may be able to remove viruses to meet EPA standards (4 log removal) due to biofilm development.
Concrete Biosand Filters
Experiments are now being run on five concrete biosand filters. All filters follow the designs specified by the Centre for Affordable Water and Sanitation Technology (CAWST) in Calgary. While both filters operate similarly, the version 10 filter is design to have improved pathogen removal with a larger sand column and a smaller recommended pore volume.
The V9 filters are packed with sand, iron and steel wool. The V10 filters are packed with sand and steel wool in the diffuser basin. All filter are charged daily with MS2 bacteriophage and E. Coli bacteria.
Version 9 Concrete Filters
Two version 9 filter are being run daily, both for over 400 days (since January 2010). One is sand-only, the other has 10 lbs of iron particles mixed throughout the top half of the sand media.
The iron filter has shown an average of 5-log to 7-log (99.999% to 99.99999%) removal of MS2 over the entire period experiments were conducted. The sand filter began with virus removal seen in previous literature (1-log to 2-log), but has continuously increased removal over the entire period that it has been run. Around day 250, it surpassed the EPA standard of 4-log removal. It is currently maintaining an around 5-log removal of viruses.
The third version 9 filter, packed with steel wool in the top half of the sand media, was run for approximately 250 days. At day 200, the filter began to steadily decrease removal of viruses. This is thought to be a result of complete oxidation of the iron to the point where viruses can no longer adsorb to its surface.
Version 10 Concrete Filters
In September 2010, we began experimenting with version 10 concrete biosand filters. The first filter was packed with sand. The purpose of it was not only to serve as a control for further tests, but to also test if the version 10 filter could demonstrate improved removal over the version 9 filter. The second filter was packed with steel wool above the diffuser basin to see if viruses could be effectively removed without having to disturb the sand media.
From the start, the version 10 filter demonstrated much higher removal than previous version 9 filters.
In March of 2011, four filters were constructed in Socorro, Guatemala to be used to test the iron-amended filter in the field. These filter are being run with the same source water that is used by members of the community. Over the summer they will be tested for bacteria and virus removal using the lab facilities at Universidad del Valle de Guatemala. We hope that further collaboration will allow for more test filters to be constructed throughout Guatemala.
The filters labeled 2 and 4 to the right are sand only filters. The ones labeled 1 and 3 are packed with 10 lbs of iron in the top half of the sand media.
Continuous Flow Columns with MS2 Bacteriophage
The first experiments used 50mL pore volume glass columns to determine whether iron could effectively remove viruses under continuous flow in a sand filter. Two filters were tested, the first a sand-only control and the second packed with 10% zerovalent iron particles by volume. Each column was run for 10 pore volumes using aquifer water spiked with MS2 bacteriophage (a virus commonly used to simulate human enteric viruses).
Daily Charged Columns with MS2 Bacteriophage
Further tests were conducted on four small-scale glass columns simulating the normal operation of a biosand filter. These column were charged daily with groundwater and MS2. Primary effluent attained from the local wastewater treatment plant was also added to the influent water to stimulate biofilm growth and simulate organic contaminants. In addition, the filter were used to determine if the orientation of iron can have an affect on the removal. Three different orientation were tested (shown on the right) all using 10% iron particles by volume.
Daily Charged Columns with Rotavirus
Glass column were also used to test the efficacy of removal of rotavirus by zerovalent iron particles. Rotavirus is a common human enteric virus. These experiments used the same procedure as the previous daily charged columns experiments.
Plastic Biosand Filters
Large-scale experiments were performed using plastic HydrAid filters. Two filters were packed using the standard four-layer procedure developed by Dr. Manz. The first filter was a standard sand-only control. The second filter was amended with iron by adding 5.5 kg of small steel nails mixed evenly throughout the top half of the sand media. Nails were used as the source of iron because of their availability and low cost. These filters were run for 70 days. The first 20 days, primary effluent was added to establish biofilm growth.
Alternative Materials Experiments
Four plastic columns were used to determine alternative sources of iron. The objective was to find a source of with large surface are that did not create independent flow paths. It was also important to consider materials that would be locally available and inexpensive. Three iron sources were tested: iron particles, steel wool, and smaller nails. The plastic columns were charged daily for two weeks.