In 2007, the Public Health Engineering Department (PHED) in West Bengal, India, began evaluating treatment processes to remove arsenic from drinking water supplies. The levels of arsenic in raw water supplies in this eastern India state frequently exceed the country’s drinking water limit of 50 µg/l; some supplies contained arsenic levels of 1,000 µg/l. Just as the arsenic limit for drinking water in the United States was reduced in 2006 to 10 µg/l, so, too, might the limit in India be reduced to 10 µg/l in the future.
The West Bengal PHED set up an arsenic task force (ATF) to evaluate arsenic removal treatment technologies and processes that can reduce the arsenic to acceptable concentrations. The ATF guidelines required that any treatment process be capable of treating raw water with arsenic levels of 1,000 µg/l and must produce minimum quantities of sludge containing arsenic, which must be detoxified before disposal. Several technologies had been evaluated in recent years with varying degrees of success. One recent pilot study tested a system that promised to remove both arsenic and iron from raw water with the production of minimal quantities of sludge containing arsenic.
The Arsenic Demonstration Unit
Severn Trent Services installed a containerized SORB 33® arsenic demonstration unit at Habra, West Bengal, India. The unit was designed to treat 100 U.S. gallons per minute (378 litres per minute). The plant was commissioned on March 15, 2007 and operated six days per week until June 15, 2007, comprising a 90-day trial period.
The SORB 33 plant was housed in a custom-built container measuring 39.4 feet x 9.8 feet (12 metres x 2 metres) and 9.8 feet (3 metres) high. The SORB 33 arsenic demonstration unit consisted of three painted carbon steel pressure vessels. Vessel 1 contained Omni-SORB™ media to remove precipitated iron and manganese; Vessels 2 and 3 contained Bayoxide® E33 media to remove arsenic. Raw water was dosed with a sodium hypochlorite solution to oxidize iron, manganese and any arsenic III. The dosed raw water was then treated in Vessel 1, and the flow from Vessel 1 was split equally between Vessels 2 and 3. Treated final water was discharged to the village distribution main.
Raw water was initially provided to the SORB 33 system from the tubewell via a water tower. However, during the trial a new tubewell was commissioned that pumped raw water directly to the SORB 33 system.
Treated water from the SORB 33 system was delivered to a newly constructed water main that connected to the village supply distribution main. The three pressure vessels were backwashed with raw water, and the backwash effluent from the SORB 33 system was delivered to a newly constructed rendered brick drain. This drain connected to the village drain which terminated at the village pond.
On the recommendation of the PHED, Scientific Research Laboratory (SRL) based in Kolkata, India, was appointed to obtain and analyze raw water, final treated water and backwash effluent samples on a regular basis throughout the trial.
Raw Water Quality at Habra
The raw water quality from the two tubewells available at Habra varied. The arsenic concentration ranged between 53 and 120 µg/l, and the average arsenic concentration during the trial was 82 µg/l. The raw water iron concentration also varied between 400 and 1,370 µg/l, and the average iron concentration during the trial was 860 µg/l. Table 1 summarizes the raw water sample results.
Based on the raw water quality, it was forecast that the SORB 33 system would be capable of treating 55,000 bed volumes (BVs) of water before the Bayoxide® E33 media reached exhaustion. This would cover a period of time of approximately 21 months.
SORB 33 System Operation
During construction, all three pressure vessels of the arsenic demonstration unit were pre-filled with graded garnet. The garnet provides support for the media and aids distribution of water flows.
Prior to filling the vessels with media, each vessel was half-filled with water to cushion the fall. Once filled with media, each vessel was backwashed to remove ‘fines’. The backwash water was discharged to the drain (open ditch) immediately to the rear of the unit. Samples were taken throughout the backwash process until the backwash water was almost colorless. Visual readings of the samples were used to judge the efficiency of backwashing and fines removal. Each vessel was given a final forward rinse before going into service.
Prior to connecting the final treated water from the arsenic demonstration unit to the village supply, the PHED required laboratory analysis to declare the water satisfactory. Until the approval to connect the treated water to the village supply was received on March 21, the plant was run to waste for four days.
During the four-day trial, the pressure vessels were backwashed with raw water. The trial demonstrated that arsenic continues to be removed during backwashing, as arsenic in the backwash effluent was 66 percent lower than the raw water. This level of arsenic could be further reduced, possibly to below the arsenic detectable limit, by reconfiguring the system to allow the adsorber vessels to be backwashed with final treated water.
Treated Water Quality at Habra
SRL’s arsenic sample results demonstrated that at the end of the trial the final treated water from the SORB 33 system contained arsenic below the limit of 10 µg/l and iron below the limit of 100 µg/l. Table 2 summarizes the final treated water sample results. With no days lost due to breakdowns, the arsenic demonstration plant was reliable throughout the trial period.
When the trial ended on June 15, neither the Omni-SORB nor Bayoxide media had reached exhaustion. Vessel 2 had treated approximately 4,000 BVs; Vessel 3 had treated approximately 3,900 BVs.
During the trial, backwash effluent contained only 7 mg/l of suspended solids. This quantity of solids was too small to carry out any quantitative analytical determinations. The SORB 33 process produces small quantities of sludge when adsorber vessels are backwashed, and sample results from the SORB 33 pilot trial suggested that only 41 grams of sludge were produced when the leading vessel was backwashed – about one ton of sludge every 230 years. The leading vessel is backwashed twice weekly.
An independent scientist obtained partially exhausted media samples during the trial and these samples were subjected to leachate tests. The Bayoxide media easily passed the United States Environmental Protection Agency (USEPA) Toxicity Characteristic Leaching Procedure (TCLP) leachate test and also the California Wet Test. It was concluded that exhausted Bayoxide media can be safely and economically disposed to landfill.
For more information, e-mail info@severntrentservices.com.
[ x ] Close Window