Science Inventory

UNDERGRADUATE PROJECT ON VIRUS REMOVAL IN SLOW SAND FILTERS FOR RURAL MAYAN COMMUNITIES

Impact/Purpose:

Socorro is located in the Bocacosta region along the volcanic slopes of the highlands in southwestern Guatemala. Despite the village’s proximity to a major international highway, infrastructure is severely underdeveloped. A survey of the village reveals that few have latrines or other waste disposal systems, and in most cases the waste is channeled towards the Chichoy River. This river also serves as a drinking water source for approximately 75% of the village. Due to community effort, most homes now have piped untreated river water, which is used for drinking, cooking, and washing. However, this water is not potable. Water- and soil-borne pathogens cause problems such as scabies, lice, diarrhea, dysentery, hepatitis A, giardia, shigella, entamoeba, and soil transmitted helminthes. We hope to alleviate these health issues by implementing a slow sand filtration (SSF) system that will use iron oxide particles mixed in with the sand substrate.

Description:

Long-Term Removal in Columns

To simulate the normal operation of a biosand filter, 4 glass columns (Figure 1) packed with different iron orientations were charged daily with 1 PV of aquifer water containing ~108 pfu/mL of MS-2 bacteriophage, a commonly used surrogate for enteric virus, and 2.5% pasteurized primary effluent (PE). While the sand column averaged only 1-log10 (90%) removal, all three iron columns had greater than 5-log10 (99.999%) removal for the duration of the experiment.

M2-2 Removal by BSFs

Plastic biosand filters were used to study virus removal by BSFs amended with iron nails. The nail filter started with a 6.5-log10 removal (99.99997%) but retention of virus particles quickly declined as flow paths short-circuited the iron. These independent flow paths were caused by the difference in size and shape between the nails and sand media. To avoid this problem, two concrete biosand filters were constructed using a steel mold built to specifications provided by CAWST. The filters were packed and charged daily with commercially available iron particles. The sand filter achieved an average of only about 2-log10 (90%) removal (Figure 2), while the iron particle filter maintained an average of 6-log10 (99.9999%) reduction of viriron particles. The iron particles had dimensions similar to the sand and, thus, promoted near plug-flow conditions that allowed for sufficient contact time between each filter charge.

Testing for Locally Available Iron Materials

It is highly desirable to pack the filters with iron that is readily obtained in rural Guatemala. Four plastic bottles, shown in Figure 3, packed with different iron sources were charged with 1 PV aquifer water and ~108 pfu/ml MS-2. The bottles were run for a period of 2 weeks. Nails achieved an average of only 3-log10 removal, which continued to decline over the course of the experiment. Zerovalent iron particles obtained an average removal of approximately 7-log10 and steel wool an average removal of 5.5-log10 (Figure 3). The large surface area of zerovalent iron particles and steel wool facilitated higher removal.

Record Details:

Record Type:PROJECT( ABSTRACT )
Start Date:08/15/2009
Completion Date:08/14/2010
Record ID: 249508