Bi-tech wells: an effective arsenic mitigation method

In 1995, the world first learned of the problem of arsenic in drinking water in India. Over time, studies have shown there are more than 150 million at risk people in India, Nepal and Bangladesh. Twenty years later, arsenic mitigation in these countries is still a patchwork of often temporary solutions. Pipelines, several types of arsenic removal filters, deep and shallow borewells (locally known as tubewells), traditional and improved/sanitary dugwells, rooftop rainwater harvesting, river and pond sand filters, in situ evaporation method: all of these interventions have limitations, but none of these above methods can be ‘the one method’ for mitigation in the arsenic-afflicted villages of this region (Smith 2001; WHO 2003; Howard et al. 2010; Milton et al. 2012).

In West Bengal, India, the arsenic mitigation options all have drawbacks:

Pipelines—In 2001, the government of West Bengal proposed a scheme to supply safe water through pipelines to eight districts that had arsenic contamination, at enormous construction cost (estimated in 2001 at 437.5 million USD or INR 21 crores). In 2006, the proposed scheme was replaced by a new master plan to tackle arsenic contamination of groundwater in West Bengal (Government of West Bengal 2013), which is ongoing. Pipelines are being constructed without operation or maintenance plans in place. Bagjola panchayat members of Deganga Block of North 24 Parganas district described on video (Project Well 2012) broken taps resulting in wasted water, tampered pipes, and even pipelines installed further than the water is able to flow and thus never working. There are similar complaints of irregular water supply in many parts of the country, and in certain areas, communities do not use the water due to the turbidity (Singh 2005). Pipelines have not been built to reach many remote villages.

Arsenic removal filters—some community arsenic removal filters require ongoing maintenance by trained personnel that communities are reluctant to take responsibility for; community and household filters require collection logistics and disposal of the filtered toxic sludge.

Shallow tubewells, which tap shallow aquifers (<60 m/200 feet) contained in the uppermost Katwa (Kandi) formation sediments (Bhattacharyaa & Banerjee 1979) of the Holocene age, usually deliver arsenic-contaminated water (Ravenscroft et al. 2005). Deep tubewells, which generally draw groundwater directly from aquifers (>150 m/500 feet) contained within deep sedimentary formations of Sijua & Lalgarh (Banerjee et al. 1998) of the Pleistocene age, are free of arsenic. The arsenic-contaminated groundwater from overlying Katwa formation seeps down through the gravel packs of the tubewells. Thus, numerous deep tubewells which were initially safe were subsequently vulnerable to arsenic contamination (Burgess et al. 2010).

Our shallow ‘modified’ dugwells—In 2001, Project Well installed its first shallow modified concrete dugwell, or ‘modified dugwell’ that tapped arsenic-safe water from an unconfined aquifer (Smith et al. 2003). Unlike the traditional rope-and-bucket system, water was extracted with a hand pump attached to the dugwell to prevent contamination, particularly bacterial contamination. Over eight years, Project Well constructed 72 modern dugwells and continued to monitor all the wells, including analyzing the water for arsenic and fecal coliform bacteria, and measuring the height of the water column to calculate the amount of Theoline, the locally available disinfectant containing 5–10% chlorine (Hira-Smith et al. 2007) to be used in each well. About 15% of the modified dugwells became dry during the dry seasons, since ‘sand boiling’ prevents wells from being deepened during construction. Sand boiling occurs when water under pressure wells up through a bed of sand. The water looks like it is ‘boiling’ up from a bed of sand and prevents the wells from being drilled any further down to the dugwell target depth of 7.6 m (25 feet). Well depths ranging from about 4 to 6 m (14 to 20 feet) were often drying in the summer. The average depth of all the modified wells was 5.7 m (19 feet).

To solve the issue of sand boiling, Project Well decided to change the design of its wells. In 2008, the first Project Well ‘bi-tech’ well, also known as a bore-dugwell, was constructed, and in 2009, 20 of 25 constructed wells were bi-tech wells. Since even during the driest month of May 2010 none of the bi-tech wells of the North 24 Parganas and Nadia districts became dry, we standardized the design of the well and constructed an additional 198 bi-tech wells during the next five years, from 2010 to 2014.

A bi-tech well combines two types of wells: borewell and dugwell. Prior to actual construction, a pilot test is done at the proposed site to determine the depth of the water-bearing sand layer. The sand layer is generally within 4.5 m (15 feet) below ground level (bgl) in the target area and the sites are chosen so the sand layer starts no more than 6 m (20 feet) bgl. Manual excavation of the dugwell portion is limited to 4.5 m (15 feet) bgl and involves mostly clay, with poor porosity inhibiting rapid recharge. This depth was determined from eight years of experience encountering sand boiling at different depths during construction. To overcome the sand boiling problem, a 3 m (10-foot) long pipe is inserted into the sand boiling layer as part of the bi-tech well. This design keeps the depth of all wells at a uniform 8 m (27 feet), including reducers, and prevents wells from drying all year round.

The plug is dropped with the tail upward so that it fits into the hole of the plug cutter. This seals the bottom end of the pipe by sealing the plug cutter and prevents mud/sand from entering the pipe in the future. The height of the cone is 20 cm (8 inches) and the tail is about 30 cm (12 inches) long. The long tail helps the plug go down vertically without tilting. A final cleaning of the inside of the plugged PVC pipe is done.

The annular space between the pipe and the borehole is next filled up with 50 kilograms of clean, uniform, coarse yellow sand.

Four or five local skilled well diggers are usually employed following the drilling. They expand the 18-inch hole radially to about 110 cm (3.5 feet), just the right size to fit the 100 cm (3.3 foot)-diameter concrete rings (specifications are available in the Project Well Guidelines (Smith & Liaw 2012)). Starting from the top, the diggers carefully cut around the inserted pipe up to a level exposing 30 to 60 cm (1 or 2 feet) of the PVC pipe including the reducer. The set-up, with 2.4 m (8 feet) of the larger pipe plus the reducer, extends the bi-tech well to a total depth of 8 m (27 feet). This depth may vary slightly depending on geology (depth & thickness of aquifers), but caution is taken to keep the depth within 30 feet to avoid reaching the arsenic-releasing redox zone (Wagner et al. 2005). The concentric rings are stacked on top of one another until the total height of the cylinder reaches approximately 5 m (17 feet) with 1–3 feet above ground level, or 1–3 rings above the surface, depending on whether the site is prone to flooding. To join two concrete rings, mortar (mixture of cement, sand and water) is used.

Following well construction, the water in the well is pumped out, preferably with a small irrigation pump, to remove the finer residue until the water is clear and sediment-free. Before allowing the community to use the water for drinking and cooking, the well water is treated with the chlorine-based disinfectant, Theoline, to kill any bacteria that may have been introduced into the aquifer during well construction. Materials and tools used during construction are generally contaminated with soil materials and certain types of bacteria found living in soils at the well site. Water from a well is considered bacterially safe to drink only when tests show that it contains no more than 1 coliform bacterium per 100 ml. The chlorine concentration must be high enough so that a free chlorine residual remains several hours after treatment. Enough retention time must be allowed so that the chlorine can kill the bacteria. Theoline is applied once a week for one month and the water is analyzed for arsenic concentration.

After implementing the bi-tech wells, we were interested in comparing availability of water during the hot summer season between the modified dugwells and bi-tech wells. We also measured levels of the two major contaminants of concern that cause disease in this area: arsenic and the diarrhea-causing fecal coliform.

Table 1 compares wells containing water and wells that dry in the summer season. In 2010, out of 50 modern dugwells, 54 percent contained water and 46 percent were seasonally dry. In the same year, 100 percent of the bi-tech wells contained water; none were dry. In all these five years, the dugwells dried at higher rates than the bi-tech wells; for example, in 2014, 48 percent of 40 modern dugwells were seasonally dry compared to 4 percent of 165 bi-tech wells. Overall, on average over a five-year period, 38 percent of modern dugwells and 5.6 percent of bi-tech wells became dry at some point during the dry season.

Almost half of the modern dugwells eventually became nonfunctional, leading to closure; the remaining half did contain water in the summer season, and by 2014 there was a decreasing trend compared to the bi-tech wells, for which 95 percent contained water.

Initially, from 2001 to 2009, bacteriological analysis of the newly constructed modern dugwells was done routinely before communities were allowed to use the well water. The water was then continuously treated with Theoline once a week or once every two weeks. Most of the analysis reports found fecal coliform to be ‘undetected,’ with only a few wells with detected numbers. At the same time, there were no reports of diarrhea related to any of the modern dugwells from the target villages. Nevertheless, bacteriological tests of the newly constructed wells continued to be conducted at random. In January 2011, water from 47 bi-tech wells constructed in 2010 and two in 2009 were analyzed for bacteria (total coliform – a group of closely related bacteria that are generally not harmful to humans – and E. coli) at a reputable laboratory in the nearby city of Kolkata. The samples were submitted in duplicate and blind. The method of collection and analysis is described in a previous article published in 2007 (Hira-Smith et al. 2007). All the samples showed undetected E. coli, while total coliform was absent in 20% of the wells. In August 2011, water from 11 functional sources was selected using computer-generated random numbers from 214 sources comprising both dugwells and bi-tech wells to measure total coliform and E. coli. There were samples from 7 modern dugwells and 4 bi-tech wells, all of which had ‘undetected’ levels of E. coli.

In January 2014, water samples were collected in duplicate from two bi-tech wells (PW153 and PW155) constructed in 2010 and from one pond to test for fecal coliform and E. coli. All the samples were collected on February 10, 2014, and neither of the wells was treated with disinfectant during the previous 30 days. As per protocol, within six hours of collection, the samples were submitted to two different laboratories: (1) CABC – a laboratory near the target villages of Berachampa, established by UNICEF; (2) Envirocheck in Kolkata as before. The methods of analysis followed by the two laboratories were different but the results were similar. Envirocheck followed the method 9222B (Standard Methods for the Examination of Water and Wastewater 1998); CABC followed the multiple tube fermentation method. Total coliform was detected by Envirocheck and not by CABC. CABC tested for and found no fecal coliform, while Envirocheck tested for and found no E. coli.

In March 2012 (during the summer season in this region), arsenic was analyzed in 15 of the 112 bi-tech wells. Selection of these sources was done using computer-generated random numbers. Water was collected in 100 ml plastic bottles after washing the bottles with water from the same well. The method used by the Kolkata laboratory to analyze arsenic is 3500 (Standard Methods for the Examination of Water and Wastewater 1998). All the samples were collected in duplicate and submitted blind to the laboratory. 87% of these bi-tech wells showed arsenic concentration <50 ppb, and 13% (or two wells) >50ppb; these two wells PW161 and PW166 were kept under observation, since their levels were high at 53.5 ppb and 111.5 ppb respectively.

From 2013 onwards, arsenic has been analyzed in the Project Well local field office by two trained field staff using Wagtech's Visual Color detection kit. The manufacturer product number is WAG-WE10600 (UNICEF 2010). The measuring range (color scale graduation) is: <10, 20–40, 50, 60–80, 100, 100–200, 200–300, 300–400, 400–500, 500 ppb. In 2013, the total water samples were collected from 141 sources during January to August 2013, of which 114 samples were from bi-tech wells and 27 from modern dugwells. 97% of the wells contained arsenic <10 ppb, 2% contained arsenic 10–40 ppb and 1% (PW246) had arsenic level of 60–80 ppb. PW 246 was tested again on February 2013 and had concentrations <10 ppb.

Annual arsenic measurement was not done in 2014, and in 2015 (Table 2) arsenic analysis using the Wagtech field kit was done on 191 wells in the summer months of March and April. The reports show 96 percent of the wells with arsenic concentration <10 ppb, 3 percent with arsenic concentration 10 to 40 ppb and 1 percent with arsenic more than 40 ppb.

Of the 15 wells measured in 2012, 8 were also measured in 2013 and 2015 (Table 3) using the Wagtech field kit. There is a decreasing trend of arsenic concentration in three years. In fact, PW166 showed a fourfold reduction in two years.

Dug-cum-bore wells are used for irrigation in many parts of Peninsula India (Singh 2005). The Project Well-designed bi-tech well, based on these dug-cum-bore wells, has been introduced to increase water availability in the summer season. Over eight years (2001 to 2009) of observation of water levels and assessment of having more than 15% of modern dugwells being dry in the summer months, Project Well came up with a well design that overcame this problem while being appropriate for the local conditions of remote Indian villages unsuitable for large construction equipment.

Historically, drinking water from surface water and traditional dugwells was used by people in West Bengal and many parts of India that caused cholera and diarrheal diseases (WHO). These traditional rope-and-bucket dugwells were replaced by tubewells in the sixties to reduce exposure to pathogenic bacteria, but these borewells introduced the new problem of arsenic contamination. Our modified dugwells, and later our bi-tech wells, replaced the rope-and-buckets of traditional dugwells with hand pumps, thus reducing the source of contamination. Care is taken in treating the water with a chlorine-based disinfectant weekly. The user community is trained to apply the disinfectant. Some user communities use their own discretion in administering the disinfectant after getting the training. For instance, Theoline was applied to PW153 and PW155 34 days and 28 days prior to the collection of water for bacteriological analysis (and not weekly as recommended by our program), yet E. coli and fecal coliform were undetected in these samples. The users of these wells also had no complaints of diarrhea.

Bi-tech wells were introduced in the arsenic-affected areas to reduce the exposure of the people and since then, 97% of these sources supplying water from the unconfined aquifer contain arsenic concentration less than 10 ppb while 1% contains more than 40 ppb. This report is of water collected from different bi-tech wells in different months of the year. Project Well continues to monitor and research the fluctuation in arsenic concentrations of the water in the bi-tech wells during different seasons.

Because some of our modern dugwells were drying during the summer, Project Well introduced bi-tech wells. Our observations from 2010 to 2014 indicated that on average, 38% of the shallow modern dugwells became dry in summer compared to fewer than 5.6% of bi-tech wells; the popularity of the bi-tech wells is growing, since water is available in 95% of the wells during the summer season.

According to the monthly survey record of May 2014, only 5% of the bi-tech wells became dry and 95% contained water. In the driest month, May, the water table falls to approximately 5 m (16 feet) bgl in the target districts of North 24 Parganas and Nadia; the water tables gets recharged by the onset of the monsoon in mid-June and rain falls until mid-October (Mukherjee et al. 2007). The annual rainfall in this region is about 150 cm (60 inches), and this amount of rainfall is enough to make water available in the bi-tech wells throughout the year.

A major component of dugwell maintenance is dredging – our modified dugwells average 19 feet deep, and every year during heavy monsoon rains, the pressure of the water pushes 2–3 rings of the dugwell into the mud. As a consequence, many dugwells require dredging after the monsoons. In contrast, the dugwell portion of bi-tech wells is only 15 feet deep and so bi-tech wells do not encounter this problem. This cuts maintenance costs considerably.

Bi-tech wells have a simple design and ongoing maintenance that is easy to adopt and adapt – the skills needed to construct and maintain the wells and the raw materials are available locally, thus enhancing the economy of rural India. These wells provide water to more than 100 persons in some areas and 40 in others. The construction cost (not including training, education and monitoring program) of each bi-tech well is approximately US$500 (INR 30,000 in 2014) and the maintenance cost is approximately US$10 (INR600) annually to purchase Theoline and repair the valve and handle of the hand pump if needed; these maintenance costs are paid for by most of the communities.

From the design and construction point of view, bi-tech wells are easily adaptable and feasible, especially in remote areas where transport and delivery of large equipment is difficult. Water is available all year round in more than 96 percent of the bi-tech wells and these wells rarely need to be dredged, making maintenance simple – mainly the regular application of disinfectant to prevent bacterial growth. Random water tests for E. coli bacteria have indicated ‘undetected’ levels, and we have not received any reports of diarrhea from users of our wells. With its low construction costs, easy long-term maintenance, and consistent safe-water output, bi-tech wells are proving an effective water solution in rural India.

We would like to acknowledge the advisors of Aqua Welfare Society (Indian partner to Project Well) and the field team for their dedication in continually improving the water program. We appreciate the role of Blue Planet Network and private donors, without whom this program could not exist. In addition to receiving funds from Blue Planet Networks since 2005, their partnership enables us to use their crowdsourced website platform and connect with other water programs around the world. We also use this platform to upload data, update reports with pictures and track the projects on the publicly-available website www.blueplanetnetwork.org. We also partner with Peer Water Exchange, based in the high-tech capital of India, Bengaluru.