Removal of Trihalomethanes by Dual Filtering Media (GAC-Sand) at El-Manshia Water Purification Plant

Transcription

1 The Journal of the Egyptian Public Health Association (JEPHAss.) Vol.81 N o. 3& 4, 2006 Removal of Trihalomethanes by Dual Filtering Media (GAC-Sand) at El-Manshia Water Purification Plant Manal A. Mohamed, Ahmed H. Hassan, Mamdouh A. EL-Messiry, Reham A. Hazzaa Environmental Health Department, High Institute of Public Health, Alexandria University. ABSTRACT Prechlorination is used as an initial step in water purification for public supply. One of the drawbacks of the prechlorination is the reaction between natural organic matters with chlorine forming trihalmethanes. This study aims at evaluating the performance of granular activated carbon (GAC) with sand as a dual filtering media with different depths on removal of trihalomethanes (THMs) for improving water quality. The Czeck sand filter at El-Manshia Water Purification Plant was chosen in this study in order to improve its water quality. The pilot filter was designed to work as mono medium sand filter and dual GAC-Sand media. The depths of GAC were 5 cm, 10 cm, 15 cm, 20 cm, 30 cm, and 40 cm over 115 cm, 110 cm, 105 cm, 100 cm, 90 cm, and 80 cm of sand, respectively. The six filter depths of GAC in the dual filter were studied to choose the optimum depth of GAC to improve water quality especially for THMs removal and comparing with mono-sand media and with Czeck filter. The results showed that the GAC-Sand dual media filter of 30 cm depth of GAC and 90 cm sand was the best depth for improving water quality where it was efficient in adsorbing mostly the total trihalomethanes in which its percentage of removal was 87%. The filtered water turbidity had an average of 0.3 NTU and its percentage of removal was 90%, algae removal was 95%, but it had a poor effect on bacteria removal with 27% removal due to adsorption of residual chlorine by GAC. The study recommended replacing mono media by dual media filter to improve water quality where the GAC was efficient to remove trihalomethanes in which the Correspondence to: Dr. Manal A. Mohamed Environmental Health Department, High Institute of Public Health Alexandria University manalmahdaly@hotmail.com

2 relative concentration (C/Co) was The benefit cost calculated on 30 cm depth of GAC is equal to 0.04 piaster/m 3. In addition, it resulted in longer filter run of 54 hrs compared to average filter run of 24 hr for Czech filters, as well as increased water productivity where unit filter run volume was 324 m 3 /m 2 instead of 144 m 3 /m 2 for Czech mono media. Keywords: Trihalomethanes formation, Dual filter, granular activated carbon, sand removal of trihalomethanes. INTRODUCTION Alexandria Governorate has six water purification plants; one of these plants is El-Manshia Water Purification Plant. Raw water is delivered to El-Manshia Plant from Mahmoudia Canal through a private canal (Special Drinking Water Canal) by a low lift pump station and the average daily production of treated water is approximately 500,000 m 3 /day. The sequences of water purification processes are prechlorination, coagulation using alum, sedimentation, and filtration followed by disinfection using chlorine. Prechlorination is the addition of chlorine prior to water treatment steps. Its purpose is to reduce the counts of bacteria and pathogens, destroy the algae, and oxidize ammonia-nitrogen compounds. One of the drawbacks of the prechlorination is the formation of trihalomethanes (THMs). The reaction takes place between natural organic matter in water (Algae) and chlorine forming THMs, which is known as haloform reaction. (1) Hoehn (2) reported that algae were a problem in water treatment plant because they built up on the filter media and increase the head loss which means that the filter needed to be backwashed more frequently with short filter run. In addition, they cause tastes and odors problem and increase THMs precursors. The maximum contaminant level (MCL) of THMs according to the Environmental Protection 242

3 Agency regulations (3) is 0.08 mg/l as annual average, while according to Egyptian Standards (4) is 0.1 mg/l. From the public health point of view, THMs at short-term exposure levels above the MCL have been shown to cause risky health effects. At long term exposure levels above MCL, THMs may cause liver, kidney, or central nervous system problems and may cause an increased risk of cancer. (3) Many water treatment plants currently use GAC to eliminate tastes and odors (5) and to reduce the amount of synthetic organic compounds (6), natural organic matter (NOM) (7), total organic carbon (TOC), and trihalomethanes (THMs) in their water supply. (8) Filtration with GAC columns for organic removal can occur after conventional deep-bed filtration for particle removal, second-stage GAC filtration (GAC post-filter adsorber), or GAC can be used as the filtration medium in an otherwise conventional deep-bed filter. This combines the removal of dissolved organics and of turbidity in a single filtration step. (9) This study aims at evaluating the performance of GAC with sand as a dual filtering media with different depths for removal of trihalomethanes (THMs) and for improving water quality. MATERIAL AND METHODS EL-Manshia Water Purification Plant Alex-Governorate has three types of rapid sand filters. These are Italba, Degremont, and Czeck filters. During the survey study, it was obvious that the filtration rate of Czeck filter of 6 m/h was the lowest filtration rate compared to Degramont and Italba filters. Moreover, the Czeck filter produced the lowest water quality compared to the other two filters. So, the Czeck filter was chosen in this study in order to improve its water quality. The study was performed from March 2003 to December

4 The basic concept of pilot filter operation is that the water level and the total head loss are kept constant, while filtration rate declines continuously in accordance with the increase in head loss in the media grains. The filtration rate starts at 9 m/h then gradually decreases as the filter gets dirty and head loss increases. This type of filtration is known as the declining rate filtration. The pilot filter consisted of a cylindrical column of PVC with a height of 3.25 m and inner diameter of 0.2 m, and the outer diameter of 0.21 m as shown in figure (1). The pilot filter worked as mono medium sand filter and dual media. The sand specifications were 1 mm effective size, 1.2 uniformity coefficient, and 2.6 specific density and GAC specifications were effective size 1.4 mm, uniformity coefficient 1.4, and specific density 1.2. The iodine number for GAC was 900 mg/l, 2% moisture percent, and phenol value was 3.5. The bed depth was 120 cm and the water level was 140 cm. The depths of GAC were 5 cm, 10 cm, 15 cm, 20 cm, 30 cm, and 40 cm over 115 cm, 110 cm, 105 cm, 100 cm, 90 cm, and 80 cm sand, respectively, to choose the optimum depth of GAC to improve water quality especially for THMs removal. Performance of optimum depth of GAC-Sand dual media was compared with Czeck filter. Six filter depths of GAC in the dual filter compared with mono-sand media were studied to select the optimum GAC depth for removal of THMs and its cost was estimated. Operation conditions of each filter were monitored. Samples of influent and effluent water for both pilot filter and field filter were collected during each filter run. These samples were analyzed according to the standard methods for the examination of water and wastewater (10) for turbidity, residual chlorine, free ammonia, THMs, bacteria, and algae count. THMs were measured by using gas chromatography (GC). The removals were calculated to obtain the best water quality. 244

5 water inlet Overflow Pilot filter of 20 cm diameter, 325 cm high Water level (140 cm) drainage GAC-Sand filtering media (120 cm) Filter pressure drop measuring tubes Gravel (20 cm) Drainage Filter outlet Filter backwash Figure (1): The Experimental Pilot Filter. RESULTS AND DISCUSSION The study showed that dual media GAC-Sand filter performance was better than sand filter for turbidity removal. In summer, the average effluent turbidity for dual medium (F 6 & F 7 ) was 0.3 NTU and for sand filter was 0.7 NTU. In winter, the average effluent turbidity for dual medium was 0.3 NTU (F 5 & F 6 & F 7 ) while for sand filter was 0.6 NTU as shown in table (1). This comes in agreement with several 245

6 studies (9, 11) which found that dual filter media produced better effluent quality than single medium sand filter. The results in table (2) illustrate that GAC-sand filter was better to remove ammonia, where in summer its percentage removal was > 94% (F 6 & F 7 ) and for sand filter was 9%. In winter, its percentage removal was > 83% (F 6 & F 7 ) and for sand filter was 2.8%. This was due to that GAC was a medium for bacterial growth that oxidized ammonia to nitrate. (12) Table (1): Effect of Different Depths of GAC on the Removal of Turbidity during Winter and Summer Seasons Pilot Filter Media Type Mono-Media Sand Dual-Media GAC-Sand Turbidity (NTU) Winter Summer Inf.* Eff.** Removal % Inf.* Eff.** Removal % F F F F F F F * Inf. = Influent ** Eff. = Effluent F 1 = 120 cm sand only. F 3 = (10 cm GAC cm sand) F 5 = (20 cm GAC cm sand) F 7 = (40 cm GAC + 80 cm sand) F 2 = (5 cm GAC cm sand) F 4 = (15 cm GAC cm sand) F 6 = (30 cm GAC + 90 cm sand) Table (2): Effect of Different Depths of GAC on the Removal of Free Ammonia during Winter and Summer Seasons Pilot Filter Media Type Mono-Media Sand Dual-Media GAC-Sand Free Ammonia (mg/l) Winter Summer Inf. Eff. Removal % Inf. Eff. Removal % F F F F F F F

7 Figure (2) shows the effect of different depths of GAC on adsorption of THMs specsies as chloroform (CHCl 3 ), dichlorobromomethane (CHCl 2 Br), dibromomethane (CHBr 2 ), and bromoform (CHBr 3 ). It is noticed that sand filter had no effect on the removal of THMs species (C/Co = 1) while GAC-Sand filters had a great effect on the removal of these species and by increasing the depth of GAC, the adsorption increases and relative concentration decreases. This result is in agreement with a study carried out in Greece by Koumendies (2001) (13) to verify the efficiency of GAC in removing THMs and in particular CHCl 3, CHCl 2 Br, CHClBr 2, and CHBr 3 from drinking water. The relative concentrations of CHCl 3, CHCl 2 Br, CHClBr 2, and CHBr 3 for GAC-Sand filter at depth of 30 cm of GAC were 0.22, 0.08, 0.06, and 0.0, respectively. Table (3) and figure (3) illustrate that the relative concentrations (C/Co) of total THMs for the pilot filters were 1.04, 0.48, 0.45, 0.29, 0.2, 0.16, and 0.1, respectively. The relative concentration (C/Co) of total THMs was decreased by increasing GAC depth (i.e., there is a direct relationship between the depth of GAC and the percentage removal of THMs). The relative concentration for THMs at a depth of 10 cm of GAC was 0.45 while at a depth of 40 cm was 0.1. Robertes and Summers (14) observed that the ratio of effluent concentration to influent concentration tended to be in the range of 0.1 to 0.5 immediately after the startup. On the other hand, the removal percentage of total THMs were -4.6%, 49%, 54%, 70%, 80%, 83%, and 89%, respectively. The percentage removal of total THMs for GAC-Sand filter of depth 30 cm was 83%, while for the sand filter was 4.6%. The GAC at a depth of 30 cm was efficient in removing total THMs up to a bed volume of approximately These agreed with Koumendies (2001). (13) 247

8 Relative concentration (C/Co) Types of filters F1 F2 F3 F4 F5 F6 F CHCl3 CHCl2Br CHClBr2 CHBr3 0 Figure (2): Effect of Different Depths of GAC on Adsorption of Trihalomethanes Species. 248

9 Table (3): Effect of Different Depths of GAC on the Adsorption of Total Trihalomethanes. Pilot Filter Media Type Mono-Media Sand Dual- Media GAC-sand *Co (µg/l) **C (µg/l) Total trihelomethanes (µg/l) Relative Concentration C/Co Removal % F *** F F F F F F * Co: Influent concentration (µg/l). ** C: Effluent concentration (µg/l). *** Effluent concentration of trihalomethanes was greater than the influent concentration. From the field survey of El-Manashia Water Purification Plant, it is obvious from table (4) that annual average concentration of THMs in the distribution system was 63.9 µg/l and the outlet of the plant was 52.3 µg/l. The highest concentration was observed in May (85.7 µg/l in the distribution system and 78.3 µg/l in the outlet of the plant) due to the presence of high algae blooms and the injection of chlorine with a dose ranging from 2 to 2.5 mg/l as post-chlorination into filtered water that leads to the formation of THMs. This value was higher than EPA regulation (80 µg/l) but still below the Egyptian standard (100 µg/l). Depending on the GAC depths, the empty bed contact times (EBCT) were recorded where EBCT is equal to the volume of the empty bed into which the GAC be placed divided by the flow rate through the bed or equal bed depth of GAC (m)/linear velocity (m/min). The EBCT were 0.5 min, 1 min, 1.5 min, 2 min, 3 min, and 4 min for depths 5, 10, 15, 20, 30, and 40 cm, respectively. 249

10 Total trihalomethanes (µg/l) Co C C/Co F1 F2 F3 F4 F5 F6 F7 Types of filters Relative concentration (C/Co) Figure (3): Effect of Different Depths of GAC on Adsorption of Total Trihalomethanes. 250

11 Hyde (15) studied the conversion of rapid sand filter to GAC. The sand was replaced by 60 cm GAC with an empty bed contact time of 7.2 min at filtration rate of 5 m/hr. He found that there was no significant difference between the two units for turbidity removal; the GAC filter removed more TOC than sand filter. On average, 60% longer filter run time was obtained with the GAC filter (72 hours) compared with sand filter (44 hours). Table (4): Average THMs Concentration at the Outlet of EL-Manashia Water Purification Plant and Distribution System during the Field Survey Month Outlet of The Plant (µg/l) Distribution System (µg/l) March April May June July August September October November N.D* N.D December Annual average *N.D: not determined. Tables (5, 6, 7) illustrate annual average of free residual chlorine (mg/l), total plate count (CFU/ml), and total algae count (unit/l) of seven filters. It is clear that free residual chlorine of mono-sand filter was higher than that of the GAC-Sand dual filter, while total plate count, and total algae were higher in the GAC-Sand filter than that of mono-sand filter. The GAC effectively removed chlorine in the top layer of bed by adsorbing it. This was explained by the reaction occurred between free chlorine and activated carbon as follows (16) : C + HOCl = CO + H + + Cl - 251

12 Where: C = an activated carbon site, CO = a surface oxide on the carbon, Cl - = free chlorine is converted to chloride ion. Residual chlorine in the effluent of GAC-Sand filter at a depth of 30 cm GAC was 0.1 mg/l in both seasons, for sand filter it was 1 mg/l in summer and 0.5 mg/l in winter, table (5). Table (5): Effect of Different Depths of GAC on the Consumption of Free Chlorine Residual during Winter and Summer Seasons Pilot Filter Media Type Mono-Media Sand Dual-Media GAC-Sand *Prechlorination dose mg/l. Free Chlorine Residual Cl 2 (mg/l) Winter Summer Inf. Eff. Temp. C Inf. Eff. Temp. C F F F F F F F Table (6): Effect of Different Depths of GAC on the Removal of Bacteria during Winter and Summer Seasons Pilot Filter Media Type Mono-Media Sand Dual-Media GAC-Sand Total Plate Count (CFU/ml) Winter Summer Inf. Eff. Removal % Inf. Eff. Removal % F F F F F F F Total coliform is not detected neither in influent nor effluent. Prechlorination dose ranged between mg/l. Free chlorine residual in influent ranged from 0.6 to1mg/l in winter & from 0.8 to1.2 mg/l in summer. 252

13 Table (7): Effect of Different Depths of GAC on the Removal of Algae during Winter and Summer Seasons Pilot Filter Media Type Mono-Media Sand Dual-Media GAC-Sand Algae Count (unit/l) Winter Summer Inf. Eff. Removal % Inf. Eff. Removal % F F F F F F F With respect to bacterial removal, it is evident from table (6) that in summer, the greater the depth of GAC (30 cm); the higher is the bacterial count in the filter effluent (60 CFU/ml), while the lower the depth of GAC (5 cm), the lower is the bacterial count (30 CFU/ml). In winter, the greater the depth of GAC (40 cm), the higher is the bacterial count in the filter effluent (55 CFU/ml), while the lower the depth of GAC (5 cm), the lower is the bacterial count (19 CFU/ml). In summer, the removal percentage of bacteria for GAC-Sand filter of 5 cm GAC depth was 63% while for GAC-Sand filter of 30 cm GAC depth the removal was 27%. This was due to decrease of residual chlorine in the effluent of dual filter media (GAC-Sand) as shown in table (5) that did not inhibit growth of microorganism on the GAC layer. The removal percent of algae for the pilot filters during summer season were 99.4%, 98%, 97%, 96.1%, 93%, 92%, and 92%, respectively. During winter season, the removal percent of the pilot filters were 99%, 98%, 97.5%, 96.5%, 95.2%, 94.9%, and 95.3%, respectively as presented in table (7). It is clear that both mono-medium sand filter and dual media 253

14 GAC-Sand filters had a high great effect on the removal of algae. Also, it is obvious that there is slight difference in the removal percent between the seven filters noticing that mono-medium sand filter is higher than dual media GAC-Sand filters. Similar results were also obtained by Salah (2000). (17) Table (8) represents a comparison between the filter operation and filtered water quality of GAC-Sand filter (30 cm GAC- 90 cm Sand) and Czech filter. It is clear that dual filter media resulted in longer filter run of 54 hrs compared to average filter run of 24 hr for Czech filters; as well as increased water productivity where unit filter run volume was 324 m 3 /m 2 instead 144 m 3 /m 2 for Czech mono media. Table (8): Comparison between Filtration Process of GAC-Sand Filter and Czech Filter Parameter Pilot GAC-Sand Filter Field Czech Filter Type of media Dual media Mono media Depth of Media (cm). E.S. (mm), UC. GAC: 30 cm, E.S. 1.4 mm, UC 1.4 Sand: 90 cm, E.S. 1 mm, UC 1.2 Sand: 120 cm, E.S. 1.1 mm, UC 1.2 Water Depth (cm) 140 cm 80 cm Head Loss (cm) Initial: 30 cm Terminal: 180 cm Head loss gauge is not working Type of Filtration Declining rate Constant rate Filtration Rate (m/h) Initial: 9 m/h Final: 3.5 m/h Range 6-7 m/h Filter Run Length (h) 54 hs 24 hs Unit Filter Run Volume (m 3 /m 2 ) 324 m 3 /m m 3 /m 2 Effluent Turbidity (NTU) 0.3 NTU 0.8 NTU Unit filter run volume (UFRV) m3/m2. = filtration rate m3/m2.h filter run h. E.S= effective size. UC = Uniformity Coefficient. 254

15 The settled water turbidity ranged from 2.2 to 3.5 NTU and averaged 3.2 NTU. The filtered water turbidity ranged from 0.2 to 0.5 NTU and averaged 0.3 NTU for GAC-Sand filter, table (9). The filtered water turbidity for Czech filter ranged from 0.6 to 1.4 NTU and averaged 0.8 NTU. The free ammonia in settled water ranged from 0.17 to 0.54 mg/l and averaged 0.35 mg/l. The free ammonia in filtered water for GAC-Sand filter averaged 0.05 while for Czech filter averaged 0.3 mg/l. The GAC-Sand filter had a great effect on the removal of ammonia. The free chlorine residual in filtered water for GAC-Sand filter averaged 0.1 mg/l while for the Czech filter it averaged 0.7 mg/l. The total THMs in settled water ranged from12 µg/l to 24 µg/l and averaged 18 µg/l. The total THMs in the filtered water for GAC-Sand filter ranged from 2 µg/l to 3 µg/l and averaged 2.5 µg/l, while in filtered water for Czech filter it ranged from 12.5 µg/l to 23.4 µg/l and averaged 16 µg/l. With respect to biological quality, total plate count in settled water averaged 78 CFU/ml, in filtered water for GAC-Sand filter it averaged 57 CFU/ml, while for Czech filter it averaged 30 CFU/ml. The total coliform was neither detected in settled water nor in the filtered water for both pilot and field filters. The total algae count in the settled water averaged 70,000 unit/l, in filtered water for GAC-Sand filter averaged 3475 unit/l, and for Czech filter 2000 unit/l. It is apparent that the biological filtered water quality for GAC-Sand filter is inferior than that of Czech filter. The physico-chemical water quality for GAC- Sand filter was much better than that of Czech filter. 255

16 Table (9): Comparison between Filtered Water Quality of Pilot Filter and Field filter Parameter Settled Water Filtered Water *GAC-Sand Filter Czech Filter Min. Max. Mean Min. Max. Mean Min. Max. Mean ph Turbidity (NTU) Ammonia (mg/l) Free Chlorine Residual (mg/l) THMs (µg/l) Total Plate Count (CFU/ml) at 37 C Total Coliform (MPM/100 ml) N.D N.D N.D Total Algae Count (unit/l) ,000 70, * Depth of GAC 30 cm, sand 90 cm. N.D: Not detected The cost of optimum depth for 30 cm GAC (if 120 cm sand for Czech is partially replaced by 30 cm GAC and 90 cm sand at filtration rate of 6 m/h) is calculated as follows: Volume of GAC = depth of GAC * Area of Czech filter = 0.3 * 42.9 = m 3. Weight of GAC = Volume of GAC * density = * 0.52 = 6.69 Kg. Volume of filtered water by GAC/month (where the GAC bed will be replaced after 30 days, i.e., month) = 6*42.9*24*30 = m 3 /month. Bed volume = Volume of water filtered by GAC/Volume of GAC = /12.87 = Cost of 1 ton GAC = L.E

17 Cost/m 3 of filtered water = 6.69 *10-3 * 12000/ = 0.04 piaster/m 3. RECOMMENDATIONS It is recommended that: 1- The Czech sand filter is better to be converted to dual-media filter using GAC of 30 cm depth, effective size mm, and uniformity coefficient <1.4, the sand depth below GAC is 90 cm with an effective size mm, and uniformity coefficient <1.4. It removes trihalomethanes by adsorbing them on the surface of GAC and increases water productivity by increasing filter run time and unit filter run volume. 2- Pre-chlorination is minimized as much as possible to minimize trihalomethanes formation in filtered water, while, post-chlorination is generally used to inhibit re-growth of bacteria in the drinking water distribution system. 3- Continuous monitoring of the drinking water quality to control operation process and to get the best quality should be adopted. REFERENCES 1. Pontius F. Water quality and treatment. 4 th Ed. New York: McGraw Hill Inc; p Hoehn RC. Algae as sources of trihalomethanes precursors. JAWWA. 1980; 72 (6): United States Environmental Protection Agency. National primary drinking water regulations Ministry of Health and Population. Egyptian standards for drinking water quality. Decree No. 108/1995. Cairo. 5. Hargesheimer EE, Waston SB. Drinking water treatment options for taste and odor. J Water Research. 1996; 30:

18 6. Speth TF, Miltner RJ. Technical note: Adsorption capacity of GAC for synthetic organics. JAWWA. 1998; 90: Summers RS, Hooper SM. Bench scale evaluation of GAC for NOM control. JAWWA. 1995; 87: Sakoda A. Trihalomethane adsorption on activated carbon filters. J Water Research. 1991;25: Wiesen MR, Rook JJ, Fiessinger F. Optimizing the placement of GAC filtration units. JAWWA. 1987; 79: Eaton AD, Cleseri LS, Greenberg AE. Standard methods for examination of water and wastewater. 20 th ed. Washington: AWWA; Omelia CR, Shin JY. Removal of particles using dual media filtration. Modeling and expermintal studies. J Water Supply. 2001; 1: Davis ML, Cornell DA. Introduction to environmental engineering. 3 rd ed. Boston: McGraw Hill; 1998.Ch Koumenides K. Using GAC to control THMs in drinking water. J Globel Nest. 2001; 3: Robertes PV, Summers RS. Performance of granular activated carbon for total organic removal. JAWWA. 1982; 74 (2): Hyde RA. Replacing sand with GAC in rapid gravity filters. JAWWA. 1987; 79: Faust SD, Aly OM. Chemistry of water treatment. 2 nd ed. London: Lewis Publishers; Ch Salah OY. Use of activated carbon and coagulant aids to improve drinking water quality. Dr. PH. Sc. Thesis, Dr. P.H. Alexandria: university of Alexandria, High Institute of Public Health,