WATER PURIFICATION THROUGH NATURAL FILTRATION

Abstract

Natural filtration is a process wherein surface water is filtered through the bank or bed of a river or lake. The process is referred as river/ lake bank filtration (RBF/ LBF). The quality of surface water changes as it flows through aquifer due to natural processes, such as filtration, sorption, acid-base reaction, oxidation, reduction, hydrolysis, biochemical reactions etc. The purification is mainly dependent on the microbial activity and chemical transformations that commonly are enhanced in the colmation layer within the river bed in contrast to those that take place in surface or groundwater. Bank filtration has been shown to be effective in attenuating a wide variety of materials including turbidity, natural organic matter, pesticides, pharmaceuticals, taste- and odor-causing compounds and microorganisms etc. Filtrate is abstracted using a wide variety of devices. The RBF thus could be classified as a low-cost treatment (or pretreatment) method. With the policy of Government of India to provide at least one safe source of water to each village, number of piped water supply schemes has recently been constructed in the State of Uttarakhand. In most of the schemes, ground water which is in direct connectivity with the surface water, is tapped with collecting devices such as infiltration wells, tube-wells, infiltration galleries and Uttaranchal koops. The effectiveness ofthese water supply schemes has not been assessed. Considering this, a necessity was felt to investigate water supply schemes in Uttarakhand, where bank/ bed filtration is being practiced. With this objective field studies were carried for the following schemes: a) Infiltration wells at Haridwar (longitude 78°13' E and latitude 29°58' N) b) Tube-wells at Nainital (longitude 79° 28' and latitude 29° 23') in c) Infiltration gallery at Bhagwanpur (longitude 77°55' E and latitude 30°20' N) d) Uttaranchal koops at 19 places in 3 districts of Uttarakhand In addition, laboratory studies were also carried out to study the effect of filtration depth, detention period (flow rate), soil characteristics, sorting of filter material, acclimatization on removal efficiency of coliform and turbidity. The thesis has been organized into following seven chapters as follows: Chapter 1: Introduction Chapter 2: Literature Review Chapter 3: Infiltration Wells at Haridwar At Haridwar number of infiltration wells (IWs) (each of 10 m diameter), located at a distance of 4 - 250 m from the bank of either river or canal, are being used to abstract mixture of ground and river/canal and ground water after passage through the soil. Hydrogeology of the study site was investigated and water samples from river/canal and sixteen IWs were analyzed in monsoon and non-monsoon periods in the first phase. Total dissolved solids (TDS), electrical conductivity (EC), cations and anions in well water were found to be more than the river water whereas organics and coliforms were observed to reduce in filtrate collected from IWs compared to river/ canal water. Average removal of some ofthe parameters is tabulated below: Removal Parameters Monsoon period Non-monsoon period Turbidity 2.9 log (167 - 256 NTU) 0.4 log (1.2 -3.4 NTU) Total colifrom 2.6 log (5000 - 16000MPN/100 mL) 4.2 log (4300 - 930000 MPN/100 mL) Fecal coliform 3.8 log (1400 - 16000 MPN/100 mL) 3.9 log (2400 - 24000 MPN/100 mL) UV 0.4log(0.151-0.256 cm-1) absorbance Values of river/ canal water are given in parentheses IV For additional study, in the second phase IW 18 situated on the Pant Dweep Island along with monitoring well 1 and 2 was selected. The average removal of turbidity, coliform and organics observed in monsoon and non-monsoon periods is tabulated below: Removal Parameters Monsoon period Non-monsoon period Turbidity 2.5 log (82 - 200 NTU) 0.7 log (1.0-3.2 NTU) Total 4.7 log (9300 - 230000 MPN/100 mL) 2.5 log (4300 - 15000 MPN/100 mL) colifrom Fecal 4.4 log (1500 - 93000 MPN/ 100 mL) 3.5 log (2100 - 15000 MPN/100 mL) coliform UV 1.0 log (0.132 - 0.224 cm"1) absorbance Values of river/ canal water are given in parentheses Water quality data have been correlated with the distance between source and production wells and travel time. It was observed that water in most of the wells was free from coliform irrespective of the distance from the source. On the basis of the investigation, six new IWs were installed in 2010 at shorter distances (5 to 30 m) from the river/ canal. Chapter 4: Lake Bank Filtration at Nainital, India: Water Quality Evaluation At Nainital, twelve tube-wells of different depths (22.6 to 36.7 m) located at a distance of < 100 m from the lake bank are being used to abstract mixture of (i) lake water after passage through the soil and (ii) subsurface water/groundwater flowing towards the lake from higher elevation. Their proportion varies with season. Geology of the tube-well site is discussed and an effort has been made to asses the travel time based on sieve analysis for aquifer soil upto the depth of 36m. Water samples from lake and five tube-wells were analyzed in monsoon and non-monsoon months. A comparison of water quality clearly differentiated tube-well water from the lake water. TDS, EC, alkalinity and hardness were found to be marginally more in tube-well waters. Results clearly indicate that lake water as such was not potable as it contained unacceptable levels of organic matter in terms of COD (~ 44 mg/L), coliforms (~ 15.6 x 104 MPN/100 mL) and nutrients. On the other hand coliform bacteria and COD have not been detected in any of the tube-well water samples over the years. The water collected from the tube-wells does not require disinfection. However, disinfection is practiced to keep the water free from contamination in the distribution network. Lake water, treated by sand filters did not conform to drinking water standards. The average removal of turbidity, coliform and organics observed in monsoon and non-monsoon periods in filtrate compared to lake water is tabulated below: Parameters Removal Monsoon period Non-monsoon period Turbidity 1.5 log (6.7 - 7.3 NTU) 1.4 log (3.7-6.1 NTU) Total 5.2 log (0.5xl02 - 50xl04 MPN/100 mL) 5.1 log (4.6xl04 - 24xl04 MPN/100 mL) colifrom Fecal 4.3 log (0.5xl02 - 50xl03 MPN/100 mL) 4.2 log (5xl03 - 24xl03 MPN/100 mL) coliform UV 0.9 log (0.056-0.061 cm"1) absorbance COD 1.5 log(19-42 mg/L) 1.8 log(20-80 mg/L) Values of lake water are given in parentheses These investigations have led to the closure of water treatment facility and installation of seven additional tube-wells. Chapter 5: Infiltration Gallery at Bhagwanpur At Bhagwanpur, a rectangular horizontal infiltration gallery of 9 m long, 1.05 m wide and 2.4 m deep with open bottom has been in use. It collects sub-surface flow VI of river Swarna which is supplied to number of villages. Water samples collected from the river Swarna and gallery were analyzed. Turbidity was found within acceptable limits as per Indian Standards for drinking water. Organics were not found in measurable concentration in the bank filtrate. Bacteriological water quality also supported the natural purification ability of the river bank/ bed. However, disinfection is practiced to keep the water free from contamination in the distribution network. Chapter 6: Uttaranchal Koops at different places in Uttarakhand Large numbers (>1000) of small water supply schemes exist in different districts of Uttarakhand mainly in hilly part to tap sub-surface water of streams through natural bed filtration. Hilly streams even when have little or no flow on surface, normally have continuous sub-surface flow through the permeable strata saturated with water. Uttaranchal Koop (a typical koop shown below) has been developed specially to tap sub-surface flow in direct connectivity with the surface water above it. Studies pertaining to Uttaranchal Koops have been discussed under three headings namely field data collection, continuous column operation and bench scale physical model studies of Uttaranchal Koop. Closed top 50-80 mm <t> Outlet socket Epoxy paint (Inside/outside) Open bottom 0.25—0.35 m * Mild steel pipe 5mm thick 1—1.5 m long Radial laterals (Perforated) 4—8 in no. " 0.6 m long (3—3.5 m from bed) Uttaranchal koop is a small device having 1-1.5 m long hollow cylindrical mild steel pipe (or a non metallic pipe) open at the bottom and closed at the top with VII radial perforated pipes (generally 4-8) screwed/welded just above the bottom of the pipe. It is connected with a welded outlet socket at the middle of the vertical cylinder. Water can enter either through radials or from bottom. Device is placed below the stream bed such that bottom of main pipe is 3 to 3.5 m below the bed of the stream. Pit is filled with either excavated material or graded gravel layer. Depending upon requirement of water, series of koops are used. Field Studies: Nineteen Uttaranchal Koops belonging to different water supply schemes located in three districts of Dehradun, Tehri, and Pauri of Uttarakhand were selected for studies in monsoon and non monsoon seasons to assess the quantum of improvement in the source water quality. Samples collected from the all the nineteen koops were found to be clearer, free from any suspended particles and of acceptable turbidity as per Indian Standards. Some of the stream samples were found to contain organics (measured as COD), however, none of the filtrate samples collected from the koops had COD in detectable range. Due to very short travel time EC, TDS and major ions in the koop waters were found to be almost of the same order as in the source waters. Concentration of total coliform in stream water at different places was found to range widely from as low as 23 most probable number (MPN) / 100 mL to as high as 93000 MPN / 100 mL over the period of study. Bacterial counts of water samples collected from Uttaranchal Koops were found to be much less compared to those of the source water samples. Highest removal of total coliforms observed was 3.4 log for monsoon and 2.6 log for non-monsoon season. For fecal coliforms it was found to be 2.8 log for monsoon and 1.4 log during non-monsoon season. Vlll The important factors found to be affecting the filtration process are (i) type of filling material used to fill up the dugged space during installation of the koop and (ii) characterises of aquifer material. Due to the presence of indicator organisms in the filtrate, it cannot be classified as potable or safe and wholesome. Further treatment i.e. disinfection is must to ensure the potability of filtrate. At small places in hills, it is difficult to maintain effective disinfection. A need was felt to improve the bacterial removal efficiency of the system by redesigning the filter media around the koop. Continuous Column Studies: During the field studies at different locations inadequate removal of coliform and a large variation in efficiencies of removal were observed. Factors such as filtration depth, detention period, soil characteristics affect bacterial removal. Column filled with aquifer materials from Bharatwala (dio = 0.12 mm, fines = 8% and fine sand = 26%) and Jajred (dio = 0.4 mm, fines = 0.5% and fine sand = 8.5%) were continuously operated to quantitatively asses the effect of turbidity, TDS and variables mentioned above. From the column studies it was found that: • Removal efficiency of coliform and turbidity with respect to depth was not linear and was higher at the initial depths. • Removal of total and fecal coliform with respect to detention period displayed logarithmic decay • Removal efficiency of coliform with respect to detention period was found to be dependent upon the effective size of the aquifer material. It was found to be higher for smaller effective grain size diameter (dio). • Initial concentration of turbidity, EC and coliform did not have any significant effect on the removal efficiency of coliform. Physical Model Studies: In order to study the performance of Uttaranchal Koop for different grading of filter/ packing material and travel time, a physical model of rectangular shape was continuously operated by filling it with different filter materials in different patterns. Effect of sorting and type of sorting of filter material, detention period (flow rate), acclimatization (i.e. formation of filter cake) on removal efficiency of coliform and turbidity were studied in detail. Fourteen different gradations were used for the study. Some of the gradations are shown below: • Aquifer material _ Small grain size | sand(0.212- 1mm) • Large grain size sand I I Fine gravel (1-4.75 mm) (4.75 - 20 mm) From the study it was found that: • Removal of coliform displayed logarithmic decay • Removal efficiency of coliform and turbidity was increased with the operation period due to clogging and formation of filter cake at the top layer of the filter. • The performance of sorted/ repacked aquifer material was found to be better than the unsorted aquifer material with similar soil properties. • From the performance of differently sorted/ repacked filter material, gradation S3F1 was found to be better than other gradations. Before installation of a new Uttaranchal koop, the characteristics of aquifer material should be determined. For the water supply schemes, where the aquifer material has smaller effective grain size (dio) i.e. between 0.1 - 0.4 mm, the dugged space of 16 m2 (4 m x 4 m) made during the installation, may be refilled with excavated aquifer material. To increase the rate of supply, series of koops could be provided at a particular site. For the water supply schemes, where the aquifer material has higher effective grain size (dio) i-e. more than 0.4 mm, the packing pattern of gradation S3F1 can be followed in the field to refill the dugged space. A retaining structure of local stones may be provided at the koop sites in the field to help in the formation of filter cake of fine material on the stream bed to improve the performance of the koop and also to reduce the erosion of formed filter cake by reducing the stream velocity. Chapter 7: Conclusion Comparative assessment of water quality from different collection devices in terms of log removal oftotal and fecal coliform is tabulated below: Infiltration Tube-wells Infiltration Uttaranchal wells gallery koops Total coliform (log 2.6-4.2 log 5.1-5.2 log 1.1-2.0 log 0-3.4 log removal) Fecal coliform (log 3.8-3.9 log 4.2-4.3 log 0.8-2.0 log 0-2.8 log removal) The results of this research have demonstrated the benefits of natural filtration to improve the quality of source water in terms of removal of bacteria, organics and turbidity for drinking water supply. As natural filtration does not require much routine maintenance, it is a cost effective alternative for water supply schemes which relies on surface water supplies to provide high quality source water with minimal subsequent treatment needs. With increase in detention time and proper packing at site of Uttaranchal Koop site, further improvement in the filtered water quality can be achieved. XI