3. WATER SUPPLY OPTIONS
 
 

The options available for water supply in the arsenic affected areas can be brought into two major categories:

1. alternative arsenic-safe water source, and
2. Treatment of arsenic contaminated water.

The type of handpump technology suitable for a particular area depends on the groundwater level, water quality and hydrogeological conditions. Arsenic safe groundwater is generally found in shallow aquifers in north-western region, and in pockets/strata in arsenic contaminated areas where conventional shallow tubewells are producing arsenic-safe water. Deep aquifers separated from shallow contaminated aquifers by impermeable layers can be a dependable source of arsenic safe water. The deeper aquifers without any separating aquiclude/clay layer may initially produce arsenic safe water but vulnerable to contamination. The important alternative water supply technologies include:

• Shallow shrouded tubewell (SST) and Very shallow shrouded tubewell (VSST)
• Deep Tubewell;
• Dug Well; and
• Infiltration galley

In many areas, groundwater with low arsenic content is available in shallow aquifers composed of fine sand at shallow depth. This may be due to accumulation of rainwater in the topmost aquifer or dilution of arsenic contaminated groundwater by fresh water recharging each year by surface and rain waters. However, the particle size of soil and the depth of the aquifer are not suitable for installing a normal tubewell. To get water through these very fine-grained aquifers, an artificial sand packing is required around the screen of the tubewell. This artificial sand packing, called shrouding, increases the yield of the tubewell and prevents entry of fine sand into the screen.

These low-cost handpump tubewell technologies have been designed and installed in the coastal areas to collect water from very shallow aquifers formed by displacement of saline water by fresh water. The SST/VSSTs can be convenient methods for withdrawal of fresh water in limited quantities. Over-pumping may yield contaminated water. Installation of low capacity pumps may prevent over exploitation of shallow aquifers. The systems may be considered suitable for drinking water supply for small settlements where water demand is low. A shallow/Very shallow tubewell is shown in Figure 3.1. The depatment of Public Health Engineering has sunk a total of 5,904 VSST/SST to provide water to 0.44 million people in coastal areas (DPHE, 2000).

Deep Tubewell

The deep aquifers in Bangladesh have been found to be relatively free from arsenic contamination. The aquifers in Bangladesh are stratified and in some places the aquifers are separated by relatively impermeable strata as shown in Figure 2.4. In Bangladesh two types of deep tubewells as shown in Figure 2.4 are constructed, manually operated small diameter tubewell similar to shallow tubewells and large diameter power operated tubewells called production well. Deep tubewells installed in those protected deeper aquifers are producing arsenic safe water. The BGS and DPHE study has shown that only about 1% deep tubewell having depth greater than 150 m are contaminated with arsenic higher than 50m g/L and 5% tubewell have arsenic content above 10m g/L (BGS and DPHE, 2001). Sinking of deep tubewells in arsenic affected areas can provide safe drinking water but replacement of existing shallow tubewells by deep tubewells involves huge cost. Some of the deep tubewells installed in acute arsenic problem areas have been found to produce water with increasing arsenic content. Post-construction analysis shows that arsenic contaminated water could rapidly percolate through shrouded materials to produce elevated levels of arsenic in deep tubewell water. Experimentation by sealing the borehole at the level of impermeable layer is yet to be conducted to draw conclusions.

However, there are many areas where the separating impermeable layers are absent and aquifers are formed by stratified layers of silt and medium sand. The deep tubewells in those areas may yield arsenic safe water initially but likely to increase arsenic content of water with time due to mixing of contaminated and uncontaminated waters. Again the possibility of contamination of deep aquifer by inter-layer movement of large quantity of groundwater cannot be ignored. If the deep aquifer is mainly recharged by vertical percolation of contaminated water from the shallow aquifer above, the deep aquifer is likely to be soon contaminated with arsenic. However, recharge of deep aquifer by infiltration through coarse media and replenishment by horizontal movement of water are likely to keep the aquifer arsenic free even after prolong water abstraction. Since many people in the rural area still use surface water for cooking, installation of deep tubewell in an area can be a source of drinking water supply for a large number of people.

In general, permeability, specific storage capacity and specific yield usually increase with depth because of the increase in the size of aquifer materials. Experience in the design and installation of tubewells shows that reddish sand produces best quality water in respect of dissolved iron and arsenic. The reddish colour of sand is produced by oxidation of iron on sand grains to ferric form. Which will not release arsenic or iron in groundwater, rather ferric iron coated sand will adsorb arsenic from ground water. Dhaka water supply, in spite of arsenic contamination around is probably protected by its red coloured soil. Hence, installation of tubewell in reddish sand, if available, should be safe from arsenic contamination.

Some areas of the coastal region of Bangladesh is very suitable for construction of deep tubewell. Department of Public Health Engineering has sunk a total of 81,384 deep tubewell mainly in the coastal area to provide safe water to 8.2 million people (DPHE,2000). The identification of areas having suitable deep aquifers and a clear understanding about the mechanism of recharge of these aquifers are needed to develop deep tubewell based water supply systems in Bangladesh.

Dug Well

Dug well is the oldest method of groundwater withdrawal for water supplies. The water of the dug well has been found to be free from dissolved arsenic and iron even in locations where tubewells are contaminated. The mechanism of producing water of low arsenic and other dissolved minerals concentration by dug wells are not fully known. The following explanations may be attributed to the low arsenic content of dug well water:

• The oxidation of dug well water due to its exposure to open air and agitation during water withdrawal can cause precipitation of dissolved arsenic and iron.
• Dug wells accumulates groundwater from top layer of a water table which is replenished each year by arsenic safe rain and surface waters by percolation through aerated zone of the soil. The fresh recharges also have diluting effects on contaminated groundwater.
• The presence of air and aerated water in well can oxidize the soils around dug wells and infiltration of water into wells through this oxidized soil can significantly reduce the concentration of arsenic in well water.

It is very difficult to protect the water of the dug well from bacterial contamination. Percolation of contaminated surface water is the most common route of pollution of well water. The upper part of the well lining and the space between the wall and soil require proper sealing. The construction of an apron around the well can prevents entry of contaminated used water at the well site by seepage into the well. Water in a dug well is very easily contaminated if the well is open and the water is drawn using bucket and rope. Satisfactory protection against bacteriological contamination is possible by sealing the well top with a watertight concrete slab. Water may be withdrawn by installation of a manually operated handpump. Water in the well should be chlorinated for disinfection after construction. Disinfecion of well water may be continued during operation by pot chlorination. A conventional dug well and a dug well with sanitary protection sunk in most common soil strata in Bangladesh are shown in figure 3.2.

In a completely closed dug well, the inflow of water is actuated by suction created due to withdrawal of water from the well. If aeration is controlling process of decontamination of well water, sanitary protection may affect the quality of well water. Extensive research to understand the mechanism of dearsination of well water and the effects of sanitary protection of well on chemical quality of water is needed. An open dug well and a closed dugwell with a tubewell for water collection are shown in Figures 3.3 and 3.4 respectively.

In the Chittagong hilly areas, Sylhet and northern parts of Bangladesh, construction of handpump tubewells is not always possible due to adverse hydrogeological and stony soil conditions. Construction of protected dug wells can be a good option for water supply in these areas. A large number of dug wells are found operating in those areas. Dug wells are not successful in many areas of Bangladesh having thick impermeable surface layer. In areas with thick clayey soil, dug wells do not produce enough water to meet the requirements. Again in areas having very low water table, there may be difficulty in construction as well as withdrawal of water. Although tubewells in Bangladesh have replaced traditional dug wells in most places, it appears from Table 2.1 that about 1.3 million people in both urban and rural areas are still dependent on dug well for drinking water supply in Bangladesh.

Infiltration Gallery / Well

Infiltration Galleries (IG) or wells can be constructed near perennial rivers or ponds to collect infiltrated surface waters for all domestic purposes. Since the water infiltrate through a layer of soil/sand, it is significantly free from suspended impurities including microorganisms usually present in surface water. Again, surface water being the main source of water in the gallery/well, it is free from arsenic. If the soil is impermeable, well graded sand may be placed in between the gallery and surface water source for rapid flow of water as shown in Figure 3.5.
 
 

Experimental units constructed in the coastal area to harvest low saline surface waters show that water of the open infiltration galleries is readily contaminated. The accumulated water requires good sanitary protection or disinfection by pot chlorination. Sedimentation of clayey soils or organic matters near the bank of the surface water source interfere with the infiltration process and require regular cleaning by scrapping a layer of deposited materials.
 
 

A protected pond in a community can provide water for drinking purpose with minimal treatment and for other domestic uses without treatment. Traditionally, rural water supply, to a great extent, was based on protected ponds before and during early stage of installation of tubewell. Sinking of tubewells under community water supply schemes in rural Bangladesh began in 1928. There are about 1,288,222 nos. of ponds in Bangladesh having an area of 0.114 ha per ponds and 21.5 ponds per mauza (BBS, 1997). About 17% of these ponds are derelict and probably dry up in the dry season. The biological quality of water of these ponds is extremely poor due to unhygienic sanitary practices and absence of any sanitary protection. Many of these ponds are made chemically and bio-chemically contaminated for fish culture. In order to maintain good quality water, a protected ponds shall not receive surface discharges or polluting substances and should only be replenished by rain and groundwater infiltration.

Pond Sand Filters

A prospective option for development of surface water based water supply system is the construction of community type Slow Sand Filters (SSFs) commonly known as Pond Sand Filters (PSFs). It is a package type slow sand filter unit developed to treat surface waters, usually low-saline pond water, for domestic water supply in the coastal areas. Slow sand filters are installed near or on the bank of a pond, which does not dry up in the dry season. The water from the pond is pumped by a manually operated hand tubewell to feed the filter bed, which is raised from the ground, and the treated water is collected through tap(s). It has been tested and found that the treated water from a PSF is normally bacteriologically safe or within tolerable limits. On average the operating period of a PSF between cleaning is usually two months, after which the sand in the bed needs to be cleaned and replaced. The drawing of a typical PSF is shown in figure 3.6.

The program initiated by DPHE with the construction of 20 experimental units in 1984 to utilize low saline pond water for water supply in the coastal area. Pond Sand Filters serve 200-500 people per unit. The PSF is being promoted as a option for water supply in arsenic affected areas.

The problems encountered are low discharge and difficulties in washing the filter beds. Since these are small units, community involvement in operation and maintenance is absolutely essential to keep the system operational. A formal institutional arrangement cannot be installed for running such a unit, community involvement in operation and maintenance is the key issue in making the system work. By June 2000, DPHE has installed 3,710 units of PSF, a significant proportion of which remains out of operation for poor maintenance.
 
 

The PSF is a low-cost technology with very high efficiency in turbidity and bacterial removal. It has received preference as an alternative water supply system for medium size settlements in arsenic affected areas and areas. Although PSF has a very high bacterial removal efficiency, it may not remove 100% of the pathogens from heavily contaminated surface water. In such cases, the treated water may require chlorination to meet drinking water standards.

The costs of construction, advantages and limitations of Pond Sand Filters as described by different organization involved in arsenic safe water supplies are summarized in Appendix-1. The major limitations mentioned are as follows:

• Operation and maintenance are difficult;
• Not suitable for heavily contaminated ponds;
• People complained of foul taste in pond water and many resorted to using it for cooking only;
• Conflicts with fish culture;
• It is difficult to find an appropriate/reserve pond for installation of PSF;
• Many ponds dry up in the dry season in some parts of the country;
• Secondary contamination takes place due to lack of proper maintenance.

Collection of safe water by installation of PSF on the bank of a pond is shown in Figure 3.7.

Combined Filters

A combined filter consists of roughing filters and a slow sand filter. It is introduced to overcome some of the difficulties encountered in PSF. The PSF cannot operate effectively when the turbidity of surface water exceeds 30 mg/l. The low discharge and requirement of frequent washing of the filter beds are common in Bangladesh. This is due to high turbidity and seasonal algal bloom in pond water. The situation can greatly be improved by design modifications particularly by the construction of roughing filters for pretreatment of water. The roughing filters remove turbidity and colour to acceptable level for efficient operation of the slow sand filter installed in sequential order following roughing filters. A diagram of a combined filtration unit is shown in Figure 3.8.

The roughing and slow sand filter units have been constructed in many parts of the world with success in reduction of very high turbidities and coliform counts. Operation and maintenance are relatively easy and less frequent attention is needed for longer duration of operation between cleaning. A combined unit of such filter has been constructed in Samta Village in Jessore district by Asian Arsenic Network (AAN) and it functions well in rural condition but remains idle in dry season when there is no water in the pond (AAN, RGAG and NIPSOM, 1999).

 Household Filters

Surface water containing impurities can be clarified by a pitcher filter unit or a small sand filter at the household level. It is an old method of water purification, once widely used in rural areas of Bangladesh. These processes of water treatment at household level have been phased out with the introduction of tubewells for village water supply. Pitcher filters are constructed by stacking a number pitchers (Kalshis), one above the other, containing different filter media as shown in Figure 3.9. Raw water is poured in the top Kalshi and filtered water is collected from the bottom one. In this process, water is mainly clarified by the mechanical straining and adsorption depending of the type of filter media used.

 Small household filters can be constructed by stacking about 300-450 mm thick well graded sand on a 150-225 mm thick coarse aggregate in a cylindrical container as shown in figure 3.10. The container is filled with water and the filtered water is collected from the bottom. It is essential to avoid drying up of the filter bed. Full effectiveness of the filtration process is obtained if the media remain in water all the time. The pitcher and other small household filters cannot completely remove micro-organisms if these are present in large numbers in raw water. Experimental units constructed in Bangladesh and in other countries show that the residual coliform bacteria present in the filtered water may vary from a few to several hundred. However, improvement of water quality by household filters is remarkable.

The important characteristics of household filters are:

• suitable for surface water treatment;
• remove turbidity, colour and micro-organisms;
• complete removal of pathogenic micro-organisms is not guaranteed;
• not suitable for high-turbid water;
• difficulty in cleaning and keeping the system operational.

Bangladesh is a tropical country and receives heavy rainfall during the rainy season. In the coastal districts, particularly in the offshore islands of Bangladesh, rainwater harvesting for drinking purposes is a common practice in a limited scale for long time (Chowdhury et al, 1987). In some areas of the coastal region with high salinity problem, about 36 percent households have been found to practice rainwater harvesting in the rainy season for drinking purpose (Hussain and Ziauddin, 1989). In the present context, rainwater harvesting is being seriously considered as an alternative option for water supply in Bangladesh in the arsenic affected areas. The main advantages and disadvantages of a rainwater system are shown in Table 3.1

Table 3.1: Advantages and disadvantages of rainwater collection system
 

Advantages

Disadvantages

The quality of rainwater is comparatively good. The system is independent and therefore suitable for scattered settlements. Local materials and craftsmanship can be used in construction of rainwater system. No energy costs are incurred in running the system. Ease in maintenance by the owner/user The system can be located very close to the consumption points.

The initial cost may prevent a family from installing a rainwater harvesting system. The water availability is limited by the rainfall intensity and available roof area. Mineral-free rainwater has a flat taste, which may not be liked by many. Mineral-free water may cause nutrition deficiencies in people who are on mineral deficient diets. The poorer segment of the population may not have a roof suitable for rainwater harvesting.

A rainwater based water supply system requires determination of the capacity of storage tank and catchment area for rainwater collection in relation to water requirement, rainfall intensity and distribution. The availability of rainwater is limited by the rainfall intensity and availability of suitable catchment area. The mineral free rainwater may not be liked by many and the poorer section of the people may not have a roof/catchment area suitable for rainwater harvesting.

Availability of Rainwater

The mean rainfall intensity recorded in 28 stations for the period from 1987 to 1998 is shown in Figure 2.2 ( BBS, 1999). It appears that the average yearly rainfall in the country varies from 1900 to 2900 mm. However, there are some losses in the collection system and a part of the rainwater is used for washing of the catchment area.. The available rainwater can be estimated by the equation:

Q = C I A … (1)

Where Q is the total quantity of rainwater available in m³/ year, C is coefficient of available runoff, I is the rainfall intensity in m/year and A is the catchment area in m². Spatial variation of rainfall in Bangladesh is quite high. Lowest rainfall less than 1500 mm occurs in the western part of the country and the highest rainfall exceeding 4000 mm occurs in the north-east. Therefore, the requirements for rainwater harvesting would vary from place to place depending on the total rainfall and its distribution over the whole year.

Rainwater Catchment

The catchment area for rainwater collection is usually the roof, which is connected with a gutter system to lead rainwater to the storage tank. A rainwater collection system from roof has been shown in Figure 3.11. Rainwater can be collected from any types of roof but concrete, tiles and metal roofs give cleanest water. The C.I. sheet roofs commonly used in Bangladesh perform well as catchment areas. In Bangladesh 48% of the households have C.I sheet, tiles and pucca roofs suitable for the collection of rainwater (BBS, 1997).

The minimum catchment area A, required for the collection of rainwater for N number of people supplied with q litres per capita per day (lpcd) of water can derived from Equation (1) as:

A = 0.365 q N / C I … (2)

About 25% of the rainwater may be assumed to be lost by evaporation and for washing the catchment area using first rain that produces inferior quality rainwater (Ahmed, 1999). The Equation (2) can be written for an average annual rainfall of 2.46 m/yr., as indicated in Figure 2.2 and a coefficient of runoff of 0.70 in the following form :

A = 0.212 q N … (3)
 
 

The poorer section of the people is in disadvantageous position in respect of utilization of rainwater as a source of water supply. This section of people has smaller size thatched roof or no roof at all, to be used as catchment for rainwater collection. A thatched roof can also be used as catchment area by covering it with polyethylene but it requires good skills to guide water to the storage tank. In coastal areas of Bangladesh, cloths fixed at four corners with a pitcher underneath is used during rainfall for rainwater collection. A plastic sheet as shown in Figure 3.12 has been tried as a catchment for rainwater harvesting for the people who do not have a roof suitable for rainwater collection. The use of land surface as catchment area and underground gravel/sand packed reservoir as storage tank can be an alternative system of rainwater collection and storage. In this case, the water has to be channeled towards the reservoir and allowed to pass through a sand bed before entering into underground reservoirs. This process is analogous to recharge of underground aquifer by rainwater during rainy season for utilization in the dry season.

Storage Tank

The unequal distribution of rainfall over the year requires storage of rainwater during rainy season for use in the dry season. The minimum volume of the storage rainwater tank V, required for rainwater can be computed by the equation:

V = O.365 f q N … (4)

Where f is the fraction of water required to be stored for consumption of total available rainwater at a constant rate throughout the year. The total annual rainfall in 1996 as shown in Figure 2.2 is approximately equal to the average annual rainfall of the last 12 years. The monthly distribution of average rainfall in 1996 shown in Figure 2.3 is assumed to represent the average condition. The rainwater availability mass curve assuming and cumulative consumption/demand of total available rainwater at constant rate are also shown in Figure 3.13.
 
 

The mass curve has been prepared considering the fact that 75% of the rainwater would be available. It may be observed that there is a shortfall of 0.48 m³ in the dry periods and an excess of 0.24 m³ during rainy season. For full utilization of rainwater potential, a storage tank of capacity 0.72 m³ that is 40% of the available rainwater is required for uninterrupted water supply at a constant rate throughout the year. However, if only drinking and cooking water is harvested, the sizes of the storage tank and catchment area would be smaller and within affordable range a family. Substituting f = 0.4 in Equation 4 for representative rainfall distribution of 1996, the minimum volume of the storage tank required for rainwater becomes:

V = 0.146 q N … (5)

However, the above simple design based on average rainfall will provide about 50% reliability of year around water supply. A design with higher reliability will require bigger catchment area and larger size of storage tank. There are several methods available for the design of a rainwater system with desired reliability and computer programs based on these methods are also available for design.
 
 

Quality of Rainwater

The quality of rainwater is relatively good but it is not free from all impurities. Analysis of stored rainwater has shown some bacteriological contamination. Cleanliness of roof and storage tank is critical in maintaining good quality of rainwater. The first run off from the roof should be discarded to prevent entry of impurities from the roof. If the storage tank is clean, the bacteria or parasites carried with the flowing rainwater will tend to die off. Some devices and good practices have been suggested to store or divert the first foul flush away from the storage tank. In case of difficulties in the rejection of first flow, cleaning of the roof and gutter at the beginning of the rainy season and their regular maintenance are very important to ensure better quality of rainwater. The storage tank requires cleaning and disinfection when the tank is empty or at least once in a year. The rainwater is essentially lacking in minerals, the presence of which is considered essential in appropriate proportions. The mineral salts in natural ground and surface waters sometimes impart pleasing taste to water.
 
 

3.2.4 Solar Distillation

Solar energy available in Bangladesh can be used for solar distillation of contaminated water in crisis areas. Experimental units based on conventional evaporation-condensation facilities have been found to produce 0.6 - 2.4 Um2/d. The water produced by solar distillation in free from all chemicals including arsenic but cannot produce enough water at a reasonable cost. The system requires further development for cost effective use in water supply in rural areas.
 
 

3.2.5 Solar Disinfection

Presence of pathogenic organisms even in apparently clear arsenic safe surface water is a hindrance for use as drinking water. These organisms can be destroyed or inactivated by solar Disinfection. This is a natural process of elimination of disease producing microorganisms using solar energy and can be applied to disinfect small quantity of water for drinking purpose. If solar radiation allowed to penetrate in water in a thin layer. the water is disinfected by the combined action of ultraviolet ray and temperature. It has been shown that if water in a transparent bottle is exposed to full sunlight for about 5 hours the water is completely disinfected (EAWAG-SANDEC. 1998). The method is not suitable for treatment of large volumes of water containing high turbidity.
 
 
 
 

3.3 TREATMENT OF ARSENIC CONTAMINATED WATER
 
 

3.3.1 General

Treatment of arsenic contaminated well water is an alternative option to make use of a huge number of tubewells likely to be declared abandoned for yielding water with high arsenic content. There are several methods available for removal of arsenic from water in large conventional treatment plats. The most commonly used technologies include oxidation, co-precipitation and adsorption onto coagulated flocs, lime treatment, adsorption onto sorptive media, ion exchange resin and membrane techniques (Cheng et al., 1994; Hering et al., 1996, 1997; Kartinen and Martin, 1995; Shen, 1973; Joshi and Chaudhuri, 1996). A detailed review of arsenic removal technologies is presented by Sorg and Logsdon (1978). Jackel (1994) has documented several advances in arsenic removal technologies. In view of the lowering the drinking water standards by USEPA, a review of arsenic removal technologies was made to consider the economic factors involved in implementing lower drinking water standards for arsenic (Chen et al., 1999). Many of the arsenic removal rtechnologies have been discuswses in details in AWWA reference book (Pontious, 1990). A comprehensive review of low-cost, well-water treatment technologies for arsenic removal with the list of companies and organizations involved in arsenic removal technologies has been compiled by Murcott (2000) with contact detail.

Some of these technologies can be reduced in scale and conveniently be applied at household and community levels for the removal of arsenic from contaminated tubewell water. During the last 2-3 years many small scale arsenic removal technologies have been developed, field tested and used under action research programs in Bangladesh and India. This sub-section presents a short review of these technologies with the intention to update the technological development in arsenic removal, understand the problems, prospects and limitations of different treatment processes as alternative water supply options for water supply.
 
 

3.3.2 Oxidation

Arsenic is present in groundwater in As(III) and AS(V) forms in different proportions. Most treatment methods are effective in removing arsenic in pentavalent form and hence include an oxidation step as preteatment to convert arsenite to arsenate. Arsenite can be oxidized by oxygen, ozone, free chlorine, hypochlorite, permanganate, hydrogen peroxide and fulton's reagent but Atmospheric oxygen, hypochloride and permanganate are commonly used for oxidation in developing countries. Air oxidation of arsenic is very slow and can take weeks for oxidation (Pierce and Moore, 1982) but chemicals like chlorine and permanganate can rapidly oxidize arsenite to arsenate under wide range of conditions.

Passive Sedimentation

Passive sedimentation received considerable attention because of rural people's habit of drinking stored water from pitchers. Oxidation of water during collection and subsequent storage in houses may cause a reduction in arsenic concentration in stored water(Bashi Pani). Experiments conducted in Bangladesh showed no to high reduction in arsenic content by passive sedimentation. Arsenic reduction by plain sedimentation appears to be dependent on water quality particularly the presence of precipitating iron in water. Ahmed et al.(2000) showed that more than 50% reduction in arsenic content is possible by sedimentation of tubewell water containing 380-480 mg/l of alkalinity as CaCO₃ and 8-12 mg/L of iron but cannot be relied to reduce arsenic to desired level. Most studies showed a reduction of zero to 25% of the initial concentration of arsenic in groundwater. In rapid assessment of technologies passive sedimentation failed to reduce arsenic to the desired level of 50 µg/L in any well (BAMWSP, DFID, WaterAid , 2001).

In-situ Oxidation

In-situ oxidation of arsenic and iron in the aquifer has been tried under DPHE-Danida Arsenic Mitigation Pilot Project. The aerated tubewell water is stored in a tank and released back into the aquifers through the tubewell by opening a valve in a pipe connecting the water tank to the tubewell pipe under the pump head. The dissolved oxygen in water oxidizes arsenite to less mobile arsenate and also the ferrous iron in the aquifer to ferric iron, resulting a reduction in arsenic content in tubewell water. Experimental results shows that arsenic in the tubewell water following in-situ oxidation is reduced to about half due to underground precipitation and adsorption on ferric iron.

Solar Oxidation

SORAS is a simple method of solar oxidation of arsenic in transparent bottles to reduce arsenic content of drinking water (Wegelin et al., 2000). Ultraviolet radiation can catalyze the process of oxidation of arsenite in presence of other oxidants like oxygen (Young, 1996). Experiments in Bangladesh show that the process on average can reduce arsenic content of water to about one-third.
 
 

3.3.3 Co-Precipitation and Adsorption Processes

Water treatment with coagulants such as aluminium alum, Al₂(SO₄)₃.18H₂O, ferric chloride , FeCl₃ and ferric sulfate Fe₂(SO₄)_3.7H₂O are effective in removing arsenic from water. Ferric salts have been found to be more effective in removing arsenic than alum on a weight basis and effective over a wider range of pH. In both cases pentavalent arsenic can be more effectively removed than trivalent arsenic.

In the coagulation-flocculation process aluminium sulfate, or ferric chloride, or ferric sulfate is added and dissolved in water under efficient stirring for one to few minutes. Aluminium or ferric hydroxide micro-flocs are formed rapidly. The water is then gently stirred for few minutes for agglomeration of micro-flocs into larger easily settable flocs. During this flocculation process all kinds of micro-particles and negatively charged ions are attached to the flocs by electrostatic attachment. Arsenic is also adsorbed onto coagualted flocs. As trivalent arsenic occurs in non-ionized form, it is not subject to significant removal. Oxidation of As(III) to As(V) is thus required as a pretreatment for efficient removal. This can be achieved by addition of bleaching powder (chlorine) or potassium permanganate. Arsenic removal is dependent on pH. In alum coagulation, the removal is most effective in the pH range 7.2-7.5 and in iron coagulation, efficient removal is achieved in a wider pH range usually between 6.0 and 8.5 (Ahmed and Rahaman, 2000).
 
 

Bucket Treatment Unit

The Bucket Treatment Unit (BTU), developed by DPHE-Danida Project is based on the principles of coagulation, co-precipitation, and adsorption processes. The unit consists of two bucket, each 20 liter capacity, placed one above the other. Chemicals are mixed manually with arsenic contaminated water in the upper red bucket by vigorous stirring with a wooden stick for 30 to 60 seconds and then flocculated by gentle stirring for about 90 second. The mixed water is then allowed to settle for 1- 2 hours. The water from the top red bucket is then allowed to flow into the lower green bucket via plastic pipe and a sand filter installed in the lower bucket. The flow is initiated by opening a valve fitted slightly above the bottom of the red bucket to avoid inflow of settled sludge in the upper bucket. The lower green bucket is practically a treated water container.

The DPHE-Danida project in Bangladesh distributed several thousands BTU units in rural areas of Bangladesh. These unit are based on chemical doses of 200 mg/L aluminum sulfate and 2 mg/L of potassium permanganate supplied in crushed powder form. The units were reported to have very good performance in arsenic removal in both field and laboratory conditions (Sarkar et al., 2000 and Kohnhorst and Paul, 2000). Extensive study of DPHE-Danida BTU under BAMWSP, DFID, WaterAid (2001) rapid assessment program showed mixed results. In many cases, the units under rural operating conditions fails to remove arsenic to the desired level of 0.05 mg/L in Bangladesh. Poor mixing and variable water quality particularly pH of groundwater in different locations of Bangladesh appeared to be the cause of poor performance in rapid assessment.

Bangladesh University of Engineering and Technology (BUET) modified the BTU and obtained better results by using 100 mg/L of ferric chloride and 1.4 mg/L of potassium permanganate in modified BTU units. The arsenic contents of treated water were mostly below 20 ppb and never exceeded 37 ppb while arsenic concentrations of tubewell water varied between 375 to 640 ppb. The BTU is a promising technology for arsenic removal at household level at low cost. It can be build by locally available materials and is effective in removing arsenic if operated properly.

Stevens Institute Technology

This technology also uses two buckets, one to mix chemicals (reported to be iron sulphate and calcium hypochloride) supplied in packets and the other to separate flocs by the processes of sedimentation and filtration. The second bucket has a second inner bucket with slits on the sides as shown in Figure 3.14 to help sedimentation and keeping the filter sand bed in place. The chemicals form visible large flocs on mixing by stirring with stick. Rapid assessment showed that the technology was effective in reducing arsenic levels to less than 0.05 mg/L in case of 80 to 95% of the samples tested (BAMWSP, DFID, WaterAid, 2001). The sand bed used for filtration is quickly clogged by flocs and requires washing at least twice a week.

Bangladesh Council of Scientific and Industrial Research (BCSIR) has developed an arsenic removal system, which uses the process of coagulation/co-precipitation with a iron based chemical followed by sand filtration. The unit did not take part in a comprehensive evaluation process.
 
 

DPHE-Danida Fill and Draw Units

It is a community type treatment unit designed and installed under DPHE-Danida Arsenic Mitigation Pilot Project. It is 600L capacity (effective) tank with slightly tapered bottom for collection and withdraw of settled sludge. The tank is fitted with a manually operated mixer with flat-blade impellers. The tank is filled with arsenic contaminated water and required quantity of oxidant and coagulant are added to the water. The water is then mixed for 30 seconds by rotating the mixing device at the rate of 60 rpm and left overnight for sedimentation. The water takes some times to become completely still which helps flocculation. The floc formation is caused by the hydraulic gradient of the rotating water in the tank. The settled water is then drawn through a pipe fitted at a level few inches above the bottom of the tank and passed through a sand bed and finally collected through a tap for drinking purpose as shown in Figure 3.15. The mixing and flocculation processes in this unit are better controlled to effect higher removal of arsenic. The experimental units installed by DPHE-Danida project are serving the clusters of families and educational institutions.

The principles of arsenic removal by alum coagulation, sedimentation and filtration have been employed in a compact unit for water treatment in the village level in West Bengal, India. The arsenic removal plant attached to hand tubewell as shown in Figure 3.16 has been found effective in removing 90 percent arsenic from tubewell water having initial arsenic concentration of 300 m g/l. The treatment process involves addition of sodium hypochloride (Cl₂), and aluminium alum in diluted form, mixing, flocculation, sedimentation and up flow filtration in a compact unit.

The use of naturally occurring iron precipitates in groundwater in Bangladesh is a promising method of removing arsenic by adsorption. It has been found that hand tubewell water in 65% of the area in Bangladesh contains iron in excess of 2 mg/l and in many acute iron problem areas, the concentration of dissolved iron is higher than 15 mg/l. Although no good correlation between concentrations of iron and arsenic has been derived, iron and arsenic have been found to co-exist in groundwater. Most of the Tubewell water samples satisfying Bangladesh Drinking Water Standard for Iron (1 mg/l) also satisfy the standard for Arsenic (50 m g/l). Only about 50% of the samples having iron content 1 - 5 mg/l satisfy the standard for arsenic while 75% of the samples having iron content > 5 mg/l are unsafe for having high concentration of arsenic.

The iron precipitates [Fe (OH)₃] formed by oxidation of dissolved iron [Fe(OH)₂] present in groundwater, as discussed above, have the affinity for the adsorption of arsenic. Only aeration and sedimentation of tubewell water rich in dissolved iron has been found to remove arsenic. The Iron Removal Plants (IRPs) in Bangladesh constructed on the principles of aeration, sedimentation and filtration in a small units have been found to remove arsenic without any added chemicals. The conventional community type IRPs, depending on the operating principles, more or less work as Arsenic Removal Plants (ARPs) as well. A study suggests that As(III) is oxidized to As(V) in the IRPs to facilitate higher efficiency in arsenic removal in IRPs constructed in Noakhali (Dahi and Liang, 1998). The Fe-As removal relationship with good correlation in some operating IRPs has been plotted in Figure 3.17. Results shows that most IRPs can lower arsenic content of tubewell water to half to one-fifth of the original concentrations. The efficiency of these community type Fe-As removal plants can be increased by increasing the contact time between arsenic species and iron flocs. Community participation in operation and maintenance in the local level is absolutely essential for effective use of these plants.

Some medium scale Fe-As removal plants of capacities 2000-3000 m³/d have been constructed for water supplies in district towns based on the same principle. The treatment processes involved include aeration, sedimentation and rapid sand filtration with provision for addition of chemical, if required. These plants are working well except that treated water requirement for washing the filter beds is very high. Operations of small and medium size IRP-cum-ARPs in Bangladesh suggest that arsenic removal by co-precipitation and adsorption on natural iron flocs has good potential.
 
 

3.3.4 Sorptive Filtration Media

Several sorptive media have been reported to remove arsenic from water. These are activated alumina, activated carbon, iron and manganese coated sand. kaolinite clay, hydrated ferric oxide, activated bauxite, titanium oxide, silicium oxide and many natural and synthetic media. The efficiency of all some sorptive media depend on the use of oxidizing agent as aids to sorption of arsenic. Saturation of media by different contaminants and components of water takes place at different times of operation depending on the specific sorption affinity of the medium to the given component.Saturation means that the efficiency in removing the desired impurities becomes zero.

Activated Alumina

Activated alumina, Al₂O₃, having good sorptive surface is an effective medium for arsenic removal. When water passes through a packed column of activated alumina, the impurities including arsenic present in water are adsorbed on the surfaces of activated alumina grains. Eventually the column becomes saturated, first at its upstream zone and later the saturated zone moves downstream towards the bottom end and finally the column gets totally saturated.

Regeneration of saturated alumina is carried out by exposing the medium to 4% caustic soda, NaOH, either in batch or by flow through the column resulting in a high arsenic contaminated caustic waste water. The residual caustic soda is then washed out and the medium is neutralized with a 2% solution of sulfuric acid rinse. During the process about 5-10% alumina is lost and the capacity of the regenerated medium is reduced by 30-40%. The activated alumina needs replacement after 3-4 regeneration. Like coagulation process, pre-chlorination improves the column capacity dramatically. Some of the activated alumina based sorptive media used in Bangladesh include:

• BUET Activated Alumina
• Alcan Enhanced Activated Alumina
• ARU of Project Earth Industries Inc.,USA
• Apyron Arsenic Treatment Unit

Granular Ferric Hydroxide

M/S Pal Trockner (P) Ltd, India and Sidko Limited, Bangladesh installed several Granular Ferric Hydroxide based arrsenic removal units in India and Bangladesh. The Granular Ferric Hydroxide (AdsorpAs^®) is arsenic selective abdsorbent developed by Technical University, Berlin, Germany. The unit requires iron removal as pre-treatment to avoid clogging of filter bed. The proponents of the unit claims to have very high arsenic removal capacity and produces non-toxic spent granular ferric hydroxide

Read-F Arsenic Removal Unit

Read-F is an adsorbent produced and promoted by Shin Nihon Salt Co. Ltd, Japan for arsenic removal in Bangladesh. Read-F displays high selectivity for arsenic ions under a broad range of conditions and effectively adsorbs both arsenite and arsenate without the need for pretreatment. The Read-F is Ethylene-Vinyl Alcohol Copolymer (EVOH) -borne hydrous cerium oxide in which hydrous cerium oxide ( CeO₂ • n H₂O), is the adsorbent. The material contains no organic solvent or other volatile substance and is not classified as hazardous material. Laboratory test at BUET and field testing of the materials at 4 sites under the supervision of BAMWSP showed that the adsorbent is highly efficient in removing arsenic from groundwater (SNSCL, 2000).

Iron Coated Sand

BUET has constructed and tested iron coated sand based small scale unit for the removal of arsenic from groundwater. Iron coated sand has been prepared following the procedure similar to that adopted by Joshi and Choudhuri ( 1996). The iron content of the iron coated sand was found to be 25 mg/g of sand. Raw water having 300 m g/L of arsenic when filtered through iron coated sand becomes significantly arsenic-free. It was found that the number of bed volume that can be treated satisfying the Bangladesh drinking water standard of 50 ppb arsenic was around 350. The saturated medium is regenerated by passing 0.2N sodium hydroxide through the column or soaking the sand in 0.2N sodium hydroxide followed by washing with distilled water. No significant change in bed volume (BV) in arsenic removal was found after 5 regeneration cycles. It was interesting to note that iron coated sand is equally effective in removing both As(III) and As(V).

Shapla Filter

Shapla filter, a household arsenic removal unit, has been designed with iron coated brick dust as an adsorption medium and works on the same principles as iron coated sand described above. The unit is effective in removing arsenic from drinking water.

Indigenous Filters

There are several filters available in Bangladesh that use indigenous material as arsenic adsorbent. Red soil rich in oxidized iron, clay minerals, iron ore, iron scrap or fillings, processed cellulose materials are known to have capacity for arsenic adsorption. Some of the filters manufactured using these material include:

• Sono 3-Kolshi Filter,
• Granet Home-made Filter,
• Chari Filter,
• Adarsha Filter,
• Shafi Filter, and
• Bijoypur Clay/Processed Cellulose filter.

The Garnet home-made filter contains relatively inert materials like brick chips and sand as filtering media. No chemical is added to the system. Air oxidation and adsorption on iron-rich brick chips and flocs of naturally present iron in groundwater could be the reason for arsenic removal from groundwater. The unit produced inadequate quantity of water and did not show reliable results in different areas of Bangladesh and under different operating conditions. The Chari filter also uses brick chips and inert aggregates in different Charis as filter media. The effectiveness of this filter in arsenic removal is not known.

The Shafi and Adarsh filters use calyey material as filter media in the form of candle. The Shafi filter was reported to have good arsenic removal capacity but suffered from clogging of filter media. The Adarsha filter participated in the rapid assessment program but failed to meet the technical criterion of reducing arsenic to acceptable level (BAMWSP, DFID and WaterAid, 2000). Bijoypur clay and treated cellulose were also found to adsorb arsenic from water (Khair, 2000).

Cartridge Filters

Filter units with cartridges filled with soptive media or ion-exchange resins are readily available in the market. These unit remove arsenic like any other dissolved ions present in water. These units are not suitable for water having high impurities and iron in water. Presence of ions having higher affinity than arsenic can quickly saturate the media requiring regeneration or replacement. Two household filters were tested at BUET laboratories, These are:

• Chiyoda Arsenic Removal Unit, Japan
• Coolmart Water Purifier, Korea.

3.3.5 Ion Exchange

The process is similar to that of activated alumina, just the medium is a synthetic resin of more well defined ion exchange capacity. The process is normally used for removal of specific undesirable cation or anion from water. As the resin becomes exhausted, it needs to be regenerated.

The arsenic removal capacity is dependent on sulfate and nitrate contents of raw water as sufate and nitrate are exchanged before arsenic. The ion exchange process is less dependent on pH of water. The efficiency of ion exchange process is radically improved by pre-oxidation of As(III) to As(V) but the excess of oxidant often needs to be removed before the ion exchange in order to avoid the damage of sensitive resins. Development of ion specific resin for exclusive removal of arsenic can make the process very attractive.

Tetrahedron ion exchange resin filter tested under rapid assessment program in Bangladesh (BAMWSP, DFID and WaterAid, 2001) showed promising results in arsenic removal. The system needs pre-oxidation of arsenite by sodium hypochloride. The residual chlorine helps to minimize bacterial growth in the media. The saturated resin requires regeneration by recirculating NaCl solution. The liquid wastes rich in salt and arsenic produced during regeneration require special treatment. Some other ion exchange resins were demonstrated in Bangladesh but sufficient field test results are not available on the performance of those resins.
 
 

3.3.6 Membrane Techniques

Membrane techniques like reverse osmosis, nonofiltration and electrodialysis are capable of removing all kinds of dissolved solids including arsenic from water. In this process water is allowed to pass through special filter media which physically retain the impurities present in water. The water, for treatment by membrane techniques, shall be free from suspended solids and the arsenic in water shall be in pentavalent form. Most membranes, however, can not withstand oxidizing agent.

MRT-1000 and Reid System Ltd.

Jago Corporation Limited promoted a household reverse osmosis water dispenser MRT-1000 manufactured by B & T Science Co. Limited, Taiwan. This system was tested at BUET and showed a arsenic (III) removal efficiency more than 80%. A wider spectrum reverse osmosis system named Reid System Limited was also promoted in Bangladesh. Experimental results showed that the system could effectively reduce arsenic content along with other impurities in water. The capital and operational costs of the reverse osmosis system would be relatively high.

Nanofiltration and Reverse Osmosis

The reverse osmosis (R/O) and nanofiltration (N/F) technologies can separate 95-98% of total dissolved solids including arsenic but it is relatively costly. In recent years, a new generation of R/O and N/F membranes have been introduced by Techno-food in Bangladesh which is less expensive and is being commercially used in industry, hotel and public water supplies. Techno-food membrane technology can remove arsenic and all other impurities present in water including bacteria at a pressure of 50-150 psi. This method of arsenic removal does not require any chemicals and Operation and maintenance requirements are minimum. The Techno-food water technologies section has marketed several models of R/O and N/F units of various water treatment capacities.

Oh et al.(2000) applied reverse osmosis and nanofiltration membrane processes for the treatment of arsenic contaminated water applying low pressure by bicycle pump. A nanofiltration membrane process coupled with a bicycle pump could be operated under condition of low recovery and low pressure range from 0.2 to 0.7 MPa. Arsenite was found to have lower rejection than arsenate in ionized forms and hence water containing higher arsenite requires pre-oxidation for reduction of total arsenic acceptable level. In tubewell water in Bangladesh the average ratio of arsenite to total arsenic was found to be 0.25. However, the reverse osmosis process coupled with a bicycle pump system operating at 4 MPa can be used for arsenic removal because of its high arsenite rejection. The study concluded that low-pressure nanofiltration with pre-oxidation or reverse osmosis with a bicycle pump device could be used for the treatment of arsenic contaminated groundwater in rural areas (Oh et al., 2000).
 
 

3.3.7 Summary

A remarkable technological development in arsenic removal from rural water supply based on conventional arsenic removal processes has taken place during last 2-3 years. A comparison of different arsenic removal processes is shown in Table 3.2.

Table 3.2 A Comparison of Main Arsenic Removal Technologies
 

Technologies	Advantages	Disadvantages
Oxidation/ Precipitation Air Oxidation  Chemical oxidation	Relatively simple, low-cost but slow process  Relatively simple and rapid process  Oxidizes other impurities and kills microbes	The processes remove only a part of arsenic
Coagulation Coprecipitation : Alum Coagulation  Iron Coagulation	Relatively low capital cost,  Relatively simple operation  Common Chemicals available	Produces toxic sludges  Low removal of As(III)  Preoxidation may be required
Sorption Techniques Actvated Alumina  Iron Coated Sand  Ion Exchange Resin  Other Sorbents	Relatively well known and commercially available  Well defined technique  Plenty possibilities and scope of development	Produces toxic solid waste  Replacement/regeneration required  High tech operation and maintenance  Relatively high cost
Membrane Techniques Nanofiltration  Reverse osmosis  Electrodialysis	Well defined and high removal efficiency  No toxic solid wastes produced  Capable of removal of other contaminants	Very high capital and running cost  High tech operation and maintenance  Toxic wastewater produced

 

A rapid assessment of 9 household level arsenic removal technologies has been completed recently (BAMWSP, DFID and Wateraid, 2000). On the basis of this study the Technical Advisory Group (TAG) of Bangladesh Arsenic Mitigation Water Supply Project (BAWSP) has recently recommended the following household arsenic removal technologies for experimental use in arsenic affected areas:

• Alcan Enhanced Activated Alumina
• BUET Activated Alumina
• Sono 3-Kolshi Method
• Stevens Institute technology.

All the technologies described in this paper have their merits and demerits and are being refined to make suitable in rural condition. The modifications based on the pilot-scale implementation of the technologies are in progress with the objectives to:

• improve effectiveness in arsenic removal,
• reduce the capital and operation cost of the systems,
• make the technology user friendly,
• overcome maintenance problems, and
• resolve sludge and arsenic concentrates management problems.

3.4 PIPED WATER SUPPLY

Piped water supply is the ultimate goal of safe water supply to the consumer because:

• Water can be delivered to the close proximity of the consumers
• Piped water is protected from external contamination
• Better quality control is possible
• Water of required quantity can be collected at ease.

It can be a feasible option for clustered rural settlements and urban fringes. Water can be made available through house connection, yard connection or standpost depending on the financial condition of the consumers. The water can be produced as per demand by sinking deep tubewell in arsenic-safe aquifer or treatment of surface or even arsenic contaminated tubewell water by community type treatment plants. A rural piped water supply system with provision for supplying water for irrigation installed by DPHE, UNICEF, BRAC and RDA, Bogra at Pakunda in Sonargaon Upazila and inaugurated by the Hon’able Minister, Ministry of Local Government, Rural Development and Cooperatives on 6 January, 2002 is shown in Figure 3.18. The system constructed at a cost of Taka 1 944 880 provides arsenic safe water to 419 households for all domestic purposes.
 
 

3.5 SCREENING AND MONITORING

3.5.1 Screening

The arsenic content of water of tubewells within short distances varies widely in many palces. This is probably due to variation in the depth of tubewells and geoenvironmental conditions of the strata of aquifers from which the tubewells abstract water. As a result, the levels of contamination in an area cannot be accurately predicted by testing of water of sample tubewells. Screening of all tubewells in the country is needed to identify the contaminated tubewell. Government of Bangladesh has decided to test arsenic content of water produced by all tubewells to identify the safe and unsafe tubewells. Bangladesh Arsenic Mitigation Water Supply Project and UNICEF have so far completed screening of all tubewells of 41 Upazillas and 5 Upazilas respectively. The estimated cost of field test kit only for the screening of estimated 7.5 million tubewells in the country is given below:

Number of Tubewell : 7.5 Million ( Estimated)

Average number of tubewels can be tested by a Kit (100 test capacity) : 80

Number of Kits Required : 7 500 000/80 = 93 750

Average Cost of an Arsenic Test Kit : Tk 2 500 ( Assumed)

Cost of Test Kit : Tk. 2 500 x 93 750 = Tk. 234 375 000 @ Tk 234 Million

The number of contaminated tubewells estimated on the basis of sample survey conducted by BGS and DPHE (2001) is 1.875 million which is 25% of the total estimated 7.5 million tubewells in Bangladesh. The present cost of this 1.875 million tubewells is Taka 8.44 billion. The significant deviations in intensity of contaminated tubewells by total screening from BGS/DPHE values justify the national screening program.
 
 

2. Monitoring

Monitoring Frequency : 1 Sample/year/Arsenic Safe Tubewell

Number of Uncontaminated Tubewell : 7 500 000 x 0.73 = 5 625 000

Number of Kits Required : 5 475 000/80 = 70 313

Cost of Test Kit : Tk. 70 313 x 2 500 =Tk. 175 782 500 @Tk. 176 Million
 
 
 
 

Since estimated 87% tubewells are likely to be privately owned, testing of water for arsenic should be the responsibility of the owner. The testing of tubewell water once in a year should be made mandatory and test facility should be available locally, preferably at the lowest level of the Local Government body.
 
 

5. COSTS

The quality and quantity of water, reliability, cost and convenience of collection of water of the different alternative options vary widely. Among the cheaper options providing water at a cost of 4 to 5 billion Taka to 29 million arsenic-exposed population, the deep tubewell can provide water at nominal operation and maintenance cost. But deep tubewells are not feasible, nor able to provide arsenic free water at all places in Bangladesh. Dug/ring well is the next option, which can provide water at moderate installation and nominal O & M cost. It is not yet fully known whether the quality of water can be maintained at desired level and arsenic content remains at safe level under conditions of proper sanitary protection. Piped water supply can be provided at a higher cost and relatively higher O & M costs but the convenience and health benefits would be enormous. Because water of adequate quantity and relatively superior quality for all domestic purposes including sanitation will be available at residences or close proximity of the residences. The increase in the number of household reduces costs but it would be difficult to get clustered houses in most places in rural areas. The installation costs of arsenic removal varies from lowest to moderate but O & M costs would be a constant burden. It may be observed that cost of installation and operation of rainwater harvesting system at household level with about only 50% reliability are very high. Installation of community rainwater harvesting system may be cheaper but management of such a system may be difficult.