Programme de bourses "Jeunes Chercheurs"

Factors affecting arsenic removal with SONO filters

Starting in the 1980–1990ies, the installation of tube wells in Bangladesh has made a source of drinking water available, which, in contrast to surface water resources, is mostly free of microbial contamination. As unfortunately only discovered in the early 1990ies, groundwater drawn from shallow wells contains high amounts of arsenic, which partially exceeds concentrations of 500 μg/L (Bangladesh’s drinking water limit for arsenic is 50 μg/L). Especially in rural areas of Bangladesh, drinking water supply is highly decentralized and ensured only by these contaminated wells. With 57% of the in- itially affected population still exposed (roughly 28 million), cheap, easy to handle, reli- able, and low maintenance filters to remove arsenic are a promising approach to miti- gate the problem. The government of Bangladesh approved of several arsenic removal units of which the SONO filter fulfills these requirements and is the only filter unit de- veloped and produced in Bangladesh.
The SONO filter consists of two plastic buckets piled on top of each other. The water enters the top bucket, which contains the active part for arsenic removal, the so-called composite iron matrix (CIM). During the contact with the water, corrosion of the CIM yields solid phases, which are excellent sorbents for arsenic, eliminating arsenic from the water to levels below 50 μg/L. In the bottom bucket fine particles, which are pro- duced in the top bucket, are removed before the water leaves the filter. Although some of the general pathways for the removal of arsenic with this type of filter are known, the complex chemistry within the CIM is not yet fully understood. The overall aim of this research project was to extend the current understanding of the processes that oc- cur in SONO filters during arsenic removal, in particular the formation of solid phases and the sequestration of arsenic from water into solid phases. This knowledge is crucial for further optimizing the current filters, for predicting the filter performance for dif- ferent water compositions, and for determining the longtime performance and capaci- ty of the filters.
In a long-term field study in Bangladesh, we evaluated the effect of differing natural groundwater composition on filter efficiency and solid (trans-)formation. In close coop- eration with the local partner and manufacturer of SONO filters, 5 representative wells differing in calcium, iron, manganese, and phosphate concentration were chosen, fil- ters were adapted for aqueous sampling at different depths and were installed in households in addition to already present filters. Over the time period of 13 months, the filters were sampled every month, which allowed for monitoring the removal efficiency and the concentration profiles of arsenic, calcium, magnesium, iron, manganese, phos- phate, dissolved oxygen (DO), and pH over time. The collected data clearly shows that SONO filters can remove arsenic as well as manganese to safe levels, i.e., below the guidelines for arsenic and manganese in drinking water of Bangladesh and the WHO, irrespective of the investigated range of phosphate, calcium, and iron concentrations in the influent water and also after more than 8 years of continuous usage. From the con- centration profiles of arsenic, iron, and DO, many of the chemical reactions observed in previous research with zero-valent iron (ZVI) could also be traced in the SONO filters.

For example, we found that initially high iron concentrations in the well water facili- tated arsenic removal in the top sand layer, that corrosion, (co-)precipitation, and ad- sorption processes with iron play a crucial role in arsenic removal in the CIM of SONO filters, and that magnetite, maghemite, and hematite were the predominant solid phases formed in the process. Interestingly, the DO content in the influent water was at the most half saturated (less than 5 mg/L), which was nevertheless sufficient for complete arsenic removal within the top bucket.
Based on the concentration range investigated in our study, we suggest that our find- ings are applicable for more than 80% of the wells analyzed in a study by the British Geological Survey in 2000 (counting only those wells with more than 50 μg/L As), which are well within the concentration range investigated in this study, particularly with respect to phosphate. In the light of these results, further optimization of the cur- rent filter systems does not seem a pressing issue and the good filter performance for different water compositions seems to be confirmed.

Caption
Fieldwork in Bangladesh, September 2009–February 2011. Local staff from the collabo- ration partner helping with field measurements (top left), SONO filter equipped for aqueous sampling at different depths (top right), children benefiting from arsenic free drinking water (middle right), solid sample of used CIM (middle left), collection of solid samples from different depths of SONO filter (bottom left), dried and ground solid samples for XRF analysis (bottom right). (Source: Anke Neumann, 2009-2011)

Involved Persons and Institutions

Postdoctoral Researcher:
Dr. Anke Neumann
Eawag
Überlandstrasse 133
8600 Dübendorf
Email: anke.neumann@jenal.org

Supervisor:
Dr. Stephan Hug
Eawag
Überlandstrasse 133
8600 Dübendorf
Email: stephan.hug@eawag.ch

Local Research Partner:
Dr. Abul K. M. Munir
Sono Technology and Research Ltd. (STR)/ Manob Sakti Unnayan Kendro (MSUK)
77 Mahatabuddin Road
Courtpara, Kushtia
Bangladesh
Email: akmmunir2003@yahoo.com

Research Partner:
Prof. Abul Hussam
Center for Clean Water and Sustainable Technologies
George Mason University
Fairfax, VA 22030 USA
Email: abulhussam@gmail.com