Loading...
HomeMy WebLinkAbout1993 03 19 - TrihalomethaneCity of Sebastian POST OFFICE BOX 780127 o SEBASTIAN, FLORIDA 32978 TELEPHONE (407) 589-5330 o FAX (407) 589-5570 M E M O R A N D U M DATE March 19, 1993 FROM : Richard B. Votapka, Utilities Director TO : Trihalomethane File SUBJECT : Meeting with Glenn Schuessler, Asst. Director, Indian River County Environmental Health Dept. Glenn told me that in his experience with trihalomethane testing (THM), he found that the longer the water remains in the distribution system, the higher the THM's seem to be. He plotted this on a map of the GDU system where he had taken tests. xv Genual Development Utilities, Inc. RECEIVED AUG J 161993 (-T-Ooc ITY CLERK:S OFFICE AN ATLANTIC GULF COMMUNITY CORPORATION SUBSIDIARY 2601 SOUTH BAYSHORE DRIVE MIAMI, FL 33133-5461 (305) 859-4331 Ms. Kathryn M. O'Halloran Acting City Manager City of Sebastian 1225 Main Street Sebastian, Florida 32958 Re: Trihalomethanes (THMs) Dear Ms. O'Halloran: Gr7U Charles E. Fancher, Jr. PRESIDENT August 13, 1993 GDU makes the following proposals to the City of Sebastian to establish a framework for addressing issues raised about allegations concerning the level of trihalomethanes CHWs) in its treated water. As the City staff is aware, GDU's water system in the City of Sebastian is not subject to any regulatory requirement regarding the level of THMs in its treated water and GDU is not currently required to take any actions concerning such levels. However, we understand that there is nevertheless a public perception of a concern about this issue. Consistent with our prior conversations with the City staff, the parties acknowledge that GDU is entitled to recover from the customers of its Sebastian Highlands system all costs incurred by GDU in addressing the THM issue. These include, but are not limited to, the cost of developing a plan for addressing the issue, any increased operating expenses incurred to address the issue, any capital expenditures incurred to address the issue, and a reasonable return on any required investment by GDU. The initial concept is that these costs will be recovered over a reasonable period of time in the form of a monthly "THM surcharge" on customers' bills. Based on these understandings, GDU is willing to work with the City to address these concerns in the following manner: 1) Within 15 days after a final order is entered on GDU's pending application for a rate increase, the City will appoint two representatives to work with GDU on addressing the THM concerns. 'IN -- Kathryn M. O'Halloran August 13, 1993 Page 2 2) Within 30 days after this proposal has been formally accepted by the City, GDU will meet with the designated representatives of the City to establish the details of a method for GDU to recover the costs of addressing the THM concerns over a reasonable recovery period. As part of this step, the City will authorize GDU to implement a surcharge or other mechanism to recover the project costs involved in Step 3 over a reasonable period of time. 3) Within 45 days after the City has authorized GDU to implement such a cost recovery mechanism, GDU will deliver an outline of a report prepared on possible methods of addressing THM levels in its Sebastian Highlands system to the City's designated representatives. 4) Within approximately 15 days after the delivery of that outline, GDU will meet with the City's representatives to discuss the outline and to attempt to identify the next step or steps to be taken in addressing the issue and the associated costs of such step or steps. 5) GDU will have no obligation to proceed beyond Step 2 until the City has taken official action to approve a surcharge mechanism and initial rates that will enable GDU to recover from its customers over a reasonable recovery period the costs incurred to that date. Similarly, GDU will have no obligation to proceed with any additional step identified as a result of Step 4 unless and until adequate provision has been made by the City for GDU's recovery of the cost of such step from its customers over a reasonable recovery period. GDU is looking forward to a cooperative effort with the City to address the public perception of a concern about THM levels in a way that is cost effective to its customers. Sincerely, Ches E. Fancher, Jr. CEF:gi cc: Marilyn Swichkow, Finance Director Richard Votapka, Utilities Director Charles Ian Nash, City Attorney CITY OF SEBASTIAN UTILITIES DEPARTMENT DATE: July 21, 1993 TIME: 10:05 A.M. [ ] CONFERENCE WITH [x] TELEPHONE CONFERENCE WITH Scott Wheeler, 190 Empress Avenue, Sebastian SUBJECT: GDU Rate Increase RESUME OF CONVERSATION: Scott called me to ask how he could register a complaint in regard to the exorbitant rate increase request by GDU. I told him that he'll have the opportunity to sign up to speak at the public hearing held on August 18, 1993. I told him that I could place him in contact with the City Clerk's office and possibly they could pre -register him to speak at the hearing. Scott told me that he doesn't drink the water ever since he read about the carcinogens in it. I explained to him that -his chances of getting cancer from THM's were 1 in 10,000 people if he drank 2 liters of water every day for the next 70 years. He also told me that his neighbors won't drink the water because of carcinogens, even though they smoke. He told me that he had done some checking and GDU water rates were the highest in Florida. I told him that Indian River County's rates were higher, and asked Scott if he had read the article in the Press Journal today. He said he hadn't. I referred him to the article which had rate comparisons of GDU now and in the future, and Indian River County. He mentioned that he had to pay for flood insurance on his lot and that his lot did not f lood out during the March 14 and March 21 100 Year rainstorms. I told him that there was a waiver or exemption form one could obtain and gave him Nancy Errett and Michelle Gentile's phone number at the County Administration Building. Nancy is at the Emergency Management Dept. and Michelle is in Engineering. He thanked me and said he'Al wait until the 18th of August. ORIGINATED BY: z FILE 1+1WOVES R CITY OF SEBASTIAN UTILITIES DEPARTMENT DATE: July 22, 1993 TIME: 10:00 a.m. [x] CONFERENCE WITH Mr. and Mrs. Newton Young from Miami, FL [ ] TELEPHONE CONFERENCE WITH SUBJECT: THM's and Utility Rates re: The GDU System RESUME OF CONVERSATION: Mr. Young and his wife came into my office. They said they were from Miami and were interested in buying an existing house located on Del Monte Road east of East Street. They said they were concerned about two things regarding the City water: 1) Is the drinking water safe to drink because of the THM's in it? 2) What about the utility rates? I explained to them that General Development Corporation serves that particular area with water, not the City. GDU has a franchise from the City to serve customers in that area with potable water. I then explained to them about trihalomethanes (THM's). I told the Newton's that THM's consist of four organic compounds. The limit set by the EPA is 100 ppb. However, I informed them that through my research I found that the reason EPA doesn't require GDU to control THM's is because their system serves only 1,200 people. Only systems with more than 10,000 people are required to provide THM control. The risk factor is such that if a person drank 2 liters of water per day for 365 days for the next 70 years, that person would have a 1 in 10,000 chance in getting cancer. I told them that they would have a better chance of being struck by lightning (1 in 9,100) in the State of Florida. I told them that I was on the GDU system and drink the water. I think it is good water. I retrieved the THM file and let them look at the 1992 results. The 1993 results were not in the file for some reason. Basically, I told them that there was not much difference between the this year's and last year's test results. They then asked me about utility rates, specifically,will they go UP? I told them that they will. I informed the Youngs that GDU had filed for a rate increase. They wanted to know if the rates would double. I told them that my bill for 7,000 gallons on the GDU proposed rates would jump from $49.17 to $103.59. This Page 2 of 2 July 22, 1993 Conference with Mr. & Mrs. Young proposed increase was more than double. I showed Mr. Young my bill where I had calculated the proposed GDU rate increase and the current County billing. It showed a 111% increase. I did tell them that the City has scheduled a public hearing in regard to the GDU rate increase on August 18, 1993. I told the Youngs that our financial consultant, Rachlin & Cohen, was reviewing the GDU rate increase. According to our Finance Department, GDU would not be able to justify the type of increase they are seeking. Most probably, an increase in rates would be granted buy certainly not as much when compared with the increase GDU has proposed. I told them that the City is trying to buy the GDU system, but we cannot conclude the agreement because GDU wants one half million dollars more. If the City had bought the GDU system at our offering price, our rates would have been lower than Indian River County's rates. However, the City's rates would be higher than the current GDU rates. The Young's thanked me and left about 10:30 a.m. ORIGINATED BY: � /'• COPY T0: 1174(*11-9�ltd A;C� M E M O R A N D U M DATE: July 1, 1993 FROM: Richard Votapka, Utilities Director TO: Trihalomethane (THM) File SUBJECT: New EPA Guidelines for the THM Maximum Contaminant Level (MCL) As instructed by EPA on March 16, 1993, I dutifully called EPA on June 1, 1993 to ask about the new EPA guidelines for the THM maximum contaminant level and effective date for mandatory requirement for THM control for all systems. I was told that the guidelines were delayed until December 1993. I am to call 1-800- 426-4791 the EPA Drinking Water Hotline on December 30, 1993 for further information. EPA feels that the guidelines will be finalized by this date. Also, the guidelines will not only involve THM's but other DBP's (Disinfection By Products) as well. CITY OF SEBASTIAN UTILITIES DEPARTMENT DATE: March 16, 1993 TIME: 4:30 PM [ ] CONFERENCE WITH: [ X ----I TELEPHONE CONVERSATION WITH: EPA Drying Water Hotline SUBJECT: Tri -amines RESUME OF CONVERSATION: I called the EPA Drinking Water Hotline (1-800-426-4791). I asked if EPA had any practical consumption comparison as to equating how much water would an individual have to drink with a certain level of THM' s in it to cause cancer. I was told that only the numerical thresholds are the acceptable guidelines now. EPA does not have any comparison to the 100 UG/L (micrograms per liter) for total trihalomethanes. I was also told to call back in the beginning of June, 1993. By then the proposed rule changes for THM' s will be established by EPA. ORIGINATED BY: f IS. V vie COPY TO T RESEARCH &TECHNOLOGY Evaluating Alternative Disinfectants for THM Control in Small Systems Anthony G. Myers A pilot -plant study was conducted at two water utilities to evaluate alternative disinfectants— ozone, hydrogen peroxide, chlorine dioxide, chloramines, and potassium permanganate—for trihalomethane (THM) control and to determine the impact of their use on small water systems. Air stripping and activated carbon were also evaluated. Combinations of alternative disinfectants reduced THMs to a range of levels from approximately 160.ugA using chlorine to 80 pg/L (chlorine-monochloramine) or <54A (ozone-monochloramine). The costs of incorporating various process modifications into small water systems were estimated. Many cities are assessing their ability to meet an anticipated lower maximum contaminant level (MCL) for trihalo- methanes (THMs). Others are motivated by professional standards or consumer pressure to provide water with the lowest practical THM concentrations as a result of concerns about possible health effects. When techniques for lowering THMs are being evaluated, other water quality concerns such as adequate disinfection and lower lead levels must also be considered. This article presents the results of a pilot -plant study conducted to determine THM reduction options for two small (2-4 mgd) water utilities. The cities of Macon and Moberly, Mo., obtain their water supply from surface - water reservoirs. The conventional pro- cess flow schemes for their treatment plants are shown in Figures 1 and 2. Raw water characteristics are summa- rized in Table 1. The practice of chlo- rinating raw water at these plants re- sulted in THM concentrations in the distribution system much greater than the current MCL of 100 jug/L (>200 µg/L), but both utilities have taken steps to reduce THM levels in their drinking water. Changing the point of chlorina- tion, adding potassium permanganate to the raw water, and optimizing the coagulation process for THM precursor removal produce drinking water that meets the current THM standard. Dis- tribution system THM concentrations were reduced by >50 percent using these techniques. Similar THM precursor removals with coagulation have been JUNE 1990 observed elsewhere.' However, if the THM standard was lowered to <50 µg/L or a water utility set a lower water quality goal, the current THM control program would not consistently produce water that would be in compliance. A pilot -plant testing program was conducted to evaluate additional means of THM reduction. The program was designed to screen a variety of treatment alternatives in a relatively short time. Optimization of each alternative was not attempted; instead, the relative effectiveness and cost of each treatment were determined to help water utilities producing 2-4 mgd anticipate require- ments to meet a range of potential THM standards. Conducting the study before new regulations are in place has several advantages. Water treatment processes such as granular activated carbon (GAC) adsorption, ozonation, and monochlor- amine require a relatively long period of time for evaluation of their effectiveness and impact on the water system. The current compliance schedule for THMs may not allow adequate time for testing and implementation of a treatment alternative if studies are delayed. In addition, cost and performance data from the pilot studies can be considered by regulatory agencies when they are de- veloping regulations. The pilot study evaluated various combinations of ozone (O�, hydrogen peroxide (H2O2), monochloramine (NH2C1), chlorine (C12), potassium per- manganate (KMnO4), and chlorine diox- ide (C1O2) for their effectiveness in reducing THMs when integrated into the existing treatment processes. The effectiveness of GAC and powdered activated carbon (PAC) in the removal of THMs and THM precursors was also Extended tests are recommended before a new treatment method is implemented. evaluated. Equipment requirements and cost estimates were prepared for the treatment alternatives investigated. Methods and materials Pilot plant. Two-week pilot -plant studies were conducted at the Macon water treatment plant (plant 1) and the Moberly water treatment plant (plant 2) in August and September 1988. The pilot -plant processes are shown in Figure 3. The basic components and design criteria are listed in Table 2. The mobile pilot plant is housed in an 8- x 34 -ft (2.4- x 10.4-m) trailer. Water from the treatment plant's raw -water supply was piped to the pilot plant's ozone contact chamber where it flowed by gravity through the static mixer, flocculation basins, primary settling basin, and into the secondary settling basin. Two peristaltic pumps in parallel transferred water from the secondary settling basin into dual -media (plant 1) or sand (plant 2) filters and GAC filters. The final'disinfectant was added to the GAC effluent to match the disinfectant residual in the dual -media or sand filter effluent. Filter effluent turbidity was monitored by a turbidimeter.* Both filter effluent turbidity and head loss data were sent to a computer for automatic collection. The hydraulic characteristics of some processes in the pilot plant were similar to those of the full-scale plants at design flow rate. However, mixing, flocculation, and settling could not be matched exactly because of physical constraints in the pilot plant. Water plant process infor- mation is shown in Table 3. Pilot -plant settled -water quality from each stage was similar to that of the full-scale treatment plants. Chemical addition. All chemicals except ,03 were fed in liquid form through peristaltic or diaphragm metering pumps. Chemical dosages (Table 4) were - adjusted by varying the pump speed or dilution of chemical stock solutions. All dosages were set so that adequate disin- fectant concentration and contact time (CT), as defined by the Surface Water Treatment Rule, would be achieved. Liquid alum was diluted from 44 per- cent to approximately 1 percent solution. Alum was added to the raw water for particle coagulation and THM precursor removal. The dosage matched that re- quired in the full-scale plant and was kept constant throughout each treatment alternative. Sodium hydroxide (NaOH) was added to the primary basin effluent to adjust the pH. The finished -water pH was 8.0 in plant 1 and 7.5 in plant 2. The pilot - plant pH was similar to that required in the full-scale plant to produce stable, noncorrosive water. The finished -water *Ratio X/R, Hach Co., Loveland, Colo. second upttow Polalum CIMIRer (stebU Wlo (PH �usttn em Kmno cottectlon sump Cls I Raw- un water pumps First upnow resereotr etuffler with mixlna wen Figure 1. Process schematic for plant 1 TABLE 1 Characteristics of raw water e a.a,w.n � storeae 'Anthrectte, sHlalf- arJce sand Pumps Parameter Plant 1 Plant 2 Total hardness -mg as CaCO3/L 95-100 120-130 Total alkalinity-mgas CaCO3/L 80-90 110-120 pH 8.0-8.2 7.6-7.8 Temperature -°C 25 28 Turbidity-ntu 10-12 2-4 254 -nm UV absorbance 0.11-0.13 0.11-0.13 TABLE 2 Characteristics of pilot plant Process Flow Rate gpm Contact Time mitt Hydraulic Loading Rate gpm/sq ft (m/min) Ozonation 1 10 5(13.1) Static mixing 1 2s 2 Flocculation 1 44 2 Primary settling* 1 49 0.8 (2.3) Secondary settling 1 500 U.08 (0.2) Filtrationt 0.15 11 1.7 (3.9) GAC filtration$ 1 0.15 13 1.7 (3.9) *36 in. of Calgon F300,.3 in. of gravel tLamella plate settler $Plant 1 used 30 in. of 0.45 -mm sand and 4 in. of gravel; plant 2 used 12 in. of 0.8 -mm anthracite, 12 in. of 0.4 -mm sand, and 3 in. of gravel. TABLE 3 Characteristics of water treatment plants at design flow rate Characteristic Plant 1 Plant 2 Design flow rate-mgd (m3/min) 4.3 (11.3) 5(13.1) Rapid mix time -min 2 Flocculation time -h 2 2 Primary settling detention time -h 3.5 4.5 Secondary settling detention time -h 3.5 4.5 Filtration gpm/sq ft (m/min) 1.9 (4.3) 2.0 (4.6) Clearwell detention time -h 3 4 78 RESEARCH AND TECHNOLOGY JOURNAL AWWA Alum Polymer Ftoccutatbn Rapid basin 1 KMtn04 mix 1 Cpumps Sugar Creak I.ake Lime Cl2 F seating babn 2 Filters Settling basin 1 Floxulatlar Rapid burin 2 mix 2 Cl, Ctear- Wet well well Iigh-serrbce I umps Distribution system Figure 2. Process schematic for plant 2 OAC Figure 3. Pilot -plant processes filtration 150 120 Ia A 90 X i; 60 G2 clog KUnOrPAC 03 O3-H2O2 Raw -Water Disinfectant Figure 4. Terminal THM results for plant 1 pH was kept relatively constant for each treatment alternative and was measured with a pH meter.* The source of C12 was bleach. In alter- natives that used free C12, the C12 dosage produced a 2-3-mg/L residual in the filtered water; this was consistent with full-scale plant operation. The applied C12 dose was 3-10 mg/L and varied with the treatment alternative and the loca- tion of C12 addition. Solid ammonium chloride was used as the source of ammonia. Ammonia dosage was based on a weight ratio of C12 to ammonia -N of approximately 3:1 to produce NH2C1. C102 was prepared from a stabilized C102 solutionj hydrochloric acid, and bleach. It was added as liquid at approx- imately 600 mg/L stock. Applied C102 doses were limited to 1.5 mg/L to mini- mize concentrations of C102 and by- products in the finished water. 03 was generated in the pilot trailer and added through a porous diffuser stone to the bottom of a cylindrical contact column. Raw water was fed to the top of the column and flowed counter- current to the 03 for efficient mixing and contact. The 03 concentration in the air was approximately 1 percent. The trans- ferred 03 dose of 2 mg/L was based on providing CT values for adequate disin- fection. Ozone transfer efficiency was approximately 70 percent. Optimization of the 03 dosage for enhanced coagulation was beyond the scope of the project. H2O2 dosage was based on a weight ratio of H2O2 to 03 of 0.5:1. The H2O2 solution was injected into the raw -water line a few feet upstream of the 03 con- tactor. H2O2 solution was stored in an opaque container to reduce degradation. KMn04 was not added to the plant raw -water line when other disinfectants were being tested but was added when pilot tests required it. The 2-mg/L dose was based on laboratory KMnO4-demand tests and plant operation experience. PAC$ was added as a slurry at 0.7 percent solution. The 20-mg/L dose was based on previous jar tests for 254 -nm UV absorbance reduction and practical limits for operational costs. Sampling and analysis. The main pa- rameter for determining the effectiveness of a treatment alternative was three-day terminal THM (term -TW. Filter and GAC effluent samples from each treat- ment alternative were collected in glass bottles with PTFE seals, stored in the dark, and held in contact with the final disinfectant for three days at 25°C. Sample pH was 8.0 at plant 1 and 7.5 at plant 2, which matched actual plant conditions. System detention time was estimated at three days based on plant operational data. After three days, so - *Model 44700. Hach Co., Loveland, Colo. tlnternational Dioxide Inc., Clark, NJ. $WPL, Calgon Corp., Pittsburgh, Pa. JUNE 1990 ANTHONY G. MYERS 79 dium thiosulfate was added to each sample. Samples were analyzed for THM by liquid -liquid extraction (method 501.2, US Environmental Protection Agency).2 Total hardness was measured by EDTA titration according to method 314B in Standard Methods,3 and total alkalinity was measured by acid titration (method 403 in Standard Methods3). Ultraviolet 254 absorbance (254 nm) was measured through a 1 -cm cell by a spectrophotom- eter* after the sample was filtered through 0.8 -mm filter paper. Chlorine dioxide was analyzed by the DPD method (Standard Methods, 401C3) and 03 by the indigo blue method. Free and total chlorine were analyzed by the DPD colorimetric method (method 408E in Standard Methods). Treatment alternatives. Treatment al- ternatives were chosen to meet a range of potential THM standards. Capital and operating costs vary with the required THM level and the treatment alternative selected. Eleven separate treatment al- ternatives were evaluated (Table 5). Alternative disinfectants -oxidants were added at various stages of the conven- tional treatment process. A strong pre- disinfectant-preoxidant was desired in the full-scale plants to reduce biological growths in the open clarifiers and to improve filter run times. The common elements in each treatment alternative were process equipment, alum dosage, sodium hydroxide dosage, and flow rate. The variables were the disinfectant chemicals used. A GACt filter-adsorber was operated in parallel with a dual -media filter at plant 1 and a sand filter at plant 2 for each treatment alternative. Virgin car- bon was installed in the GAC filter. After the pilot tests were completed, the GAC had been in use for the equivalent of five to six days of continuous operation (180 gal/ib [1.5 m3/kg] GAC). The GAC column and filter columns received the same influent. No estimates of the time to breakthrough and operational costs for GAC could be determined during pilot -plant studies because of the short testing period. Subsequent tests were conducted with a GAC column to determine THM and THM precursor removal effectiveness. In these tests, Cl2 was added to the secondary effluent of plant 1 at a dose of 2 mg/L, which reduced biological activ- ity. After approximately 10 s of contact time, the water passed through a 4 -in. - (10.2 -cm-) diameter by 3 -ft- (0.9-m-) long column of GAC. The GAC column was operated at 2 gpm/sq ft (4.6 m/min) and was backwashed once every three days. Influent and effluent samples were spiked to 6 mg chlorine/L so that a residual of at least 0.5 mg/L was present after three days. Effluent samples had higher C12 residuals at the beginning of TABLE 4 Chemical dosages Chemical Dosage—mg/L Plant 1 Plant 2 Aluminum sulfate • 18 H2O 50 30 Sodium hydroxide 15 9 Chlorine 2-8 2-10 Monochloramine 2-3 2 Chlorine dioxide 1.5 1.5 Ozone 2 2 Hydrogen peroxide 1 1 Potassium permanganate 2 2 PAC 20 20 200 160 40 CI, CIO, KMnOCPAC 03 v' -"2%J= Raw Water Disinfectant Figure S. Terminal THM results for plant 2 300 250 Q200 i 150 A a zy 9 100 10 20 30 40 Time—days 50 Figure 7. Terminal THM results for GAC pilot test 80 RESEARCH AND TECHNOLOGY JOURNAL AWWA TABLE 5 THM results 300 250 200 1S0 100 50 0 0 20 40 so so 100 120 Ttma–n Figure & Rate of formation of THMs in raw water 0.10 0.08 0.06 0.04 0.02 0 0 10 20 30 40 50 60 Time --days Figure S. Ultraviolet absorbance results for GAC pilot test the test. Samples were analyzed for THMs after three days. Resufts Filter effluent samples were collected and analyzed for three-day term-THM for each of the 11 treatment alternatives. Term-THM results for both plants are listed in Table 5 and shown in Figure 4 for plant 1 and Figure 5 for plant 2. Run 1. The objective of run 1A was to determine the THM concentration if no THM control measures were used, thus establishing a baseline for comparison with the other treatment alternatives. Chlorine was used as the preoxidant- disinfectant and final disinfectant. In alternative 1A, C12 was added to the raw water at a dose of 8-10 mg/L. A C12 residual was present throughout the treatment process, and the filter effluent contained 2-3 mg chlorine/L. The three- day term-THM concentration was 152 µg/L at plant 1 and 177 µg/L at plant 2. These were similar to full-scale plant term-THM concentrations without THM control measures. In alternative 111, ammonia was added to the filter effluent to form NH2Cl. The three-day term-THM concentrations were 88 Ag/L at plant 1 and 98 µg/L at plant 2. When free C12 is the distribution system disinfectant, >40 percent of the term-THM could be formed in the dis- tribution system, assuming that little THM formation occurs after the addition of ammonia. Ammonia could be added earlier in the treatment process to reduce the THM concentration further. However, ade- quate contact time with a disinfectant that is stronger than NH2C1 must be ensured, as outlined in the Surface Water Treatment Rule. The THM formation rates were rapid when C12 was added to the raw water of both plants (Figure 6). The THM concentration was 78 µg/L at both plants after 30 min of contact with free C12. Raw -water THM formation rate curves are shown in Figure 6 for plants 1 and 2. A better alternative would be to use a raw -water oxidant other than C12, and possibly to add C12 after settling, when THM precursors are reduced. The GAC effluent had three-day term THM concentrations of 9 µg/L at plant 1 and 20 µg/L at plant 2 with free C12 as the final disinfectant (alternative 1A). With NH2C1 as the final disinfectant (alternative 111), the GAC filter effluent had a three-day term-THM concentration of 3 µg/L at plant 1 and 2 µg/L at plant 2. Although the term-THM concentrations in water that was treated with GAC were low for all alternatives, the GAC system was not in operation long (equiv- alent to five or six days of continuous *Model 110, Hitachi, American Scientific Products, Deer• field, ill. tFittrasorb-300 (0.& 1.0 -mm e.s.), Calgon Corp., Pittsburgh, Pa. JUNE 1990 ANTHONY G. MYERS 81 Primary Three -Day Term-THM—µg/L Treatment Raw Water Basin Effluent Filter Effluent plant 1 Plant 2 Alternative Treatment Treatment Treatment GAC Filter GAC Filter IA C12 Cl2 9 152 20 177 1B C12 NH(I 3 89 2 98 2A C1O2 C12 C12 9 93 9 86 2B C1O2 C12 NH2C1 3 54 2 51 3A 03 C12 C12 11 85 9 85 3B 03 C12 NH2C1 1 59 1 45 3C 03 NH �1 NH2C1 1 3 2 10 3D 03-H2O2 CIZ C12 9 59 14 101 3E OrH2O2 C12 NH2C1 1 32 2 48 4A KMnO4 C12 C12 7 87 6 114 4B KMnO4 C12 NH2Cl 1 1 41 1 2 1 60 300 250 200 1S0 100 50 0 0 20 40 so so 100 120 Ttma–n Figure & Rate of formation of THMs in raw water 0.10 0.08 0.06 0.04 0.02 0 0 10 20 30 40 50 60 Time --days Figure S. Ultraviolet absorbance results for GAC pilot test the test. Samples were analyzed for THMs after three days. Resufts Filter effluent samples were collected and analyzed for three-day term-THM for each of the 11 treatment alternatives. Term-THM results for both plants are listed in Table 5 and shown in Figure 4 for plant 1 and Figure 5 for plant 2. Run 1. The objective of run 1A was to determine the THM concentration if no THM control measures were used, thus establishing a baseline for comparison with the other treatment alternatives. Chlorine was used as the preoxidant- disinfectant and final disinfectant. In alternative 1A, C12 was added to the raw water at a dose of 8-10 mg/L. A C12 residual was present throughout the treatment process, and the filter effluent contained 2-3 mg chlorine/L. The three- day term-THM concentration was 152 µg/L at plant 1 and 177 µg/L at plant 2. These were similar to full-scale plant term-THM concentrations without THM control measures. In alternative 111, ammonia was added to the filter effluent to form NH2Cl. The three-day term-THM concentrations were 88 Ag/L at plant 1 and 98 µg/L at plant 2. When free C12 is the distribution system disinfectant, >40 percent of the term-THM could be formed in the dis- tribution system, assuming that little THM formation occurs after the addition of ammonia. Ammonia could be added earlier in the treatment process to reduce the THM concentration further. However, ade- quate contact time with a disinfectant that is stronger than NH2C1 must be ensured, as outlined in the Surface Water Treatment Rule. The THM formation rates were rapid when C12 was added to the raw water of both plants (Figure 6). The THM concentration was 78 µg/L at both plants after 30 min of contact with free C12. Raw -water THM formation rate curves are shown in Figure 6 for plants 1 and 2. A better alternative would be to use a raw -water oxidant other than C12, and possibly to add C12 after settling, when THM precursors are reduced. The GAC effluent had three-day term THM concentrations of 9 µg/L at plant 1 and 20 µg/L at plant 2 with free C12 as the final disinfectant (alternative 1A). With NH2C1 as the final disinfectant (alternative 111), the GAC filter effluent had a three-day term-THM concentration of 3 µg/L at plant 1 and 2 µg/L at plant 2. Although the term-THM concentrations in water that was treated with GAC were low for all alternatives, the GAC system was not in operation long (equiv- alent to five or six days of continuous *Model 110, Hitachi, American Scientific Products, Deer• field, ill. tFittrasorb-300 (0.& 1.0 -mm e.s.), Calgon Corp., Pittsburgh, Pa. JUNE 1990 ANTHONY G. MYERS 81 operation) and was not anticipated to be near breakthrough. Run 2. The objective of run 2 was to determine THM levels when C102 was used as the primary oxidant -disinfectant and both C12 and NH2Cl were used as final disinfectants. In alternative 2A, C102 was added to the raw water and C12 was added to the primary basin effluent to produce a free C12 residual through secondary clarification and into the finished water. Three-day term-THM concentrations were 93 ug/L in -plant 1 and 86 µg/L in plant 2 with dual -media and sand filtration. With GAC filtration, three-day term-THM concentrations were 9 µg/L at both plants. Alternative 213 was similar to 2A except that NH2Cl was formed in the filter effluent. Three-day term-THM concen- trations were 54 µg/L with dual -media filtration at plant 1 and 51 µg/L with sand filtration at plant 2. With GAC filtration, term-THM levels were 3 µg/L at plant 1 and 2 µg/L at plant 2. Additional tests were conducted under similar conditions to those used for alternative 2A except that Cl2 was added to the filter effluent instead of to the primary basin effluent. Chlorine dioxide was still added to the raw water. Three- day term-THM concentrations were similar to those of the previous tests in which C12 was added to the primary basin effluent, indicating that little THM precursor removal occurred during sec- ondary settling and filtration. The C102 demand of the raw water left a C102 residual of only 0.1 mg/L after 0.5 h of contact time. Because C12 addition to the primary basin (versus the filter effluent) did not increase three-day term-THM when Cl2 was the final disin- fectant, a longer C12 contact time in the plant would be desirable for disinfection. Run 3. Run 3 was performed to deter- mine THM levels when 03 was used as the primary oxidant -disinfectant and both C12 and NH2Clwere used as final disinfectants. In alternative 3A, 03 was added to the raw water and C12 -was added to the primary basin effluent. Three-day term- THM concentrations in both plants were 85µg/L with dual -media and sand filtra- tion and 9-11 µg/L with GAC filtration in plants 1 and 2, respectively. Because 03 dosage was not optimized for THM reduction, lower THM levels might have been achieved at a different 03 dosage. Ozone optimization, however, was be- yond the scope of the project. Ammonia was added to the filter effluent in alternative 313 and to the primary basin effluent in alternative 3C to determine THM levels with various free Cl2 contact times. In alternative 313, three-day term-THM concentrations were 59 µg/L at plant 1 and 45 µg/L at plant 2 with dual -media and sand filtra- tion. With GAC filtration, the three-day TABLE 6 Chemical treatment costs *10 years at 8 percent Estimated Estimated Estimated §Based on three-month replacement period, no regeneration, existing filter boxes Amortized Annual Total Costs of alternative disinfectants Capital Cost* O&M Cost Cosh Chemical Dose $/1,000gal $11,OOO9al $/1,000gal Treatment mg/L ($/ms) ($/m3) ($/ms) Ammonia 1 0.007 (0.002) 0.006 (0.001) 0.013 (0.003) Potassium to filter influent, NH2Cl added to finished water permanganate 2 0.005 (0.001) 0.025 (0.006) 0.029 (0.008) Chlorine dioxide 1.5 0.010 (0.003) 0.053 (0.014) 0.063 (0.017) PAC 20 0.005 (0.001) 0.083 (0.022) 0.088 (0.023) Ozone 2 0.130 (0.034) 0.012 (0.006) 0.151(0.040) Ozone-H2O2 2/1 0.138 (0.036) 0.037 (0.010) 0.175 (0.046) Air stripping$ 0.124 (0.033) 0.046 (0.012) 0.170 (0.045) GAC§ 0.034 (0.009) 0.913 (0.242) 0.947 (0.250) *10 years at 8 percent tsum of amortized capital and operating costs *Includes new pumping and piping modifications §Based on three-month replacement period, no regeneration, existing filter boxes TABLE 7 Costs of alternative disinfectants Potential Estimated THM Standard Total Cosh 'U91L Disinfectant Alternative* $11,000 gal ($/m3) 50 Ammonia added to chlorinated finished water 0.013 (0.003) 25 KMnO4 added to raw water, C1O2 added 0.105 (0.028) to filter influent, NH2Cl added to finished water 5 03 added to raw water, NH2Cl added to settling 0.164 (0.043) basins *Combined with conventional treatment (rapid mix, coagulation -flocculation, settling, filtration) tCosts are for additional chemicals, 0&M equipment, and storage -feed equipment only. term-THM concentration was 1 µg/L at both treatment plants. In alternative 3C, NH2C1 was added to the primary basin effluent and an NH2Cl residual was maintained through filtra- tion. Ozone was added to the raw water. Three-day term-THM concentrations were 3 µg/L in plant 1 and 10 µg/L in plant 2 with dual -media and sand filtra- tion and 2 µg/L in both plants with GAC filtration. Low THM concentrations were expected because free C12 was not present. Hydrogen peroxide was added before ozonation in alternatives 3D and 3E in an attempt to enhance oxidation of THM precursors in the raw water and reduce finished -water THM concentrations. Alternative 3D was similar to alternative 3A except that H202 was added along with 03. Chlorine was added to the primary settling basin effluent, and a free chlorine residual was maintained through filtration. Results showed three- day term-THM concentrations of 59 µg/L at treatment plant 1 and 101 µg/L at plant 2 with conventional filtration. The GAC effluent term-THM concentration was 9 µg/L at plant 1 and 14 µg/L at plant 2. Alternative 3E was similar to alterna- tive 3D except ammonia was added to the filter effluent to form an NH2Cl residual. Filter effluent term-THM concentrations for both plants (32 and 48 µg/L) were approximately half that of alternative 3D in which a free C12 residual was maintained after filtration. The GAC effluent term-THM concentration was 1-2 µg/L at both plants. Run 4. Run 4 was similar to the existing water treatment plant practice of adding KMn04 (2 mg/L) to the raw water, alum coagulation, pH adjustment, and disin- fection with C12. It was different in that 20 mg PAC/L was added to the raw water to enhance removal of THM pre- cursors. The THM precursors were removed by oxidation, adsorption, and coagulation. In alternative 4A, C12 was added to the primary settling basin ef- fluent and a free C12 residual was carried through filtration. Three-day term-THM concentrations were 87 µg/L in plant 1 with dual -media filtration and 114 µg/L in plant 2 with sand filtration. These THM concentrations were not signif- icantly less than those for the full-scale water plant. Although PAC produced some term-THM reduction, significant decreases in THM levels are not antici- pated. After GAC filtration, term-THM concentrations were 7 µg/L for plant 1 and 6 µg/L for plant 2. Alternative 4B was similar to alterna- tive 4A except that ammonia was added to the filter effluent. Three-day term- THM concentrations were 41 µg/L in the dual -media effluent of plant 1 and 60 µg/L in the sand filter effluent of plant 2. Adding ammonia at the end of the treatment process reduced term-THMs 82 RESEARCH AND TECHNOLOGY JOURNAL AWWA by approximately 50 percent. The GAC effluent contained 1-2 µg THMs/L. Current plant operation does not typically use PAC, but KMn04 is added to the raw water and C12 is added to the filter influent. Previous THM results along with the pilot -study results indi- cate that a free C12 contact time of 3-4 h in the clearwell could produce THM concentrations <50 µg/L with the exist- ing water plant operation. This suggests that adding ammonia after the clearwell or to the high -service discharge line could provide adequate disinfection and THM levels <50 ug/L. Filtration. Supplementary data on head loss and turbidity were collected from the dual -media, sand, and GAC filters. Optimization of filter performance was not an objective of the pilot -plant tests. Because pilot -plant operation was in- termittent, filter runs were not carried to terminal head loss. Filter runs were typically terminated after 8-12 h of operation. During that time, effluent turbidity averaged 0.1-0.2 ntu and was similar with dual -media, sand, and GAC filtration. Filter head loss reached 1-2 ft (0.3-0.6 m) and was slightly less through the GAC filter. GAC was an effective filter medium for turbidity reduction under the conditions of the pilot test. 254 -nm UV absorbance. Ultraviolet ab- sorbance (254 nm) was measured with the raw water, filter effluent, and GAC effluent for each alternative. In some cases, 254 -nm UV absorbance indicates the level of THM precursors present in the water.' Raw -water 254 -nm UV ranged from 0.11 to 0.13 in both plants. Filter effluent 254 -nm UV ranged from 0.02 to 0.05, and that for GAC effluent was near zero. 03 pretreatment was more effective in oxidizing compounds that absorb light at a UV wavelength of 254 nm. A typical 254 -nm UV measurement for filtered water treated with 03 was 0.02. The results, however, did not indicate lower THM levels for 03 pretreated water when chlorine was used as a disinfectant, as shown in Figures 4 and 5. This suggests that, although 254 -nm UV absorbance may be useful in comparing the THM formation potentials of raw waters, oxidation or selective adsorption of organic precursors may alter 254 -nm UV absorbance so that the treated water absorbance may not be indicative of THM formation potential. GAC pilot test. The extended GAC pilot test was conducted at plant 1 from February to May 1989. Three-day term- THM results are shown in Figure 7. Influent three-day term-THMs averaged 237 µg/L, and term-THM reduction averaged 44 percent over the 90 -day test period. Because C12 was in contact with the influent for only 10 s before GAC treatment, GAC functioned mainly for THM precursor removal, not for THM removal. The GAC effluent term-THM concentrations were >100 µg/L after only three weeks of operation. The 254 -nm UV was also measured for the GAC influent and effluent, and re- sults are shown in Figure 8. Increasing effluent 254 -nm UV absorbance indicates reduced GAC capacity for 254 -nm UV - absorbing compounds and indicates, to some degree, increasing THM levels. Discussion Using C12 as the final disinfectant and 03, C102, or KMn04 as the primary oxidant -disinfectant produced three-day term-THM concentrations of 69 to 114 Ag/L. It would be difficult to meet a 50- µg/L THM standard with these disin- fectants and existing water plant pro- cesses. However, three-day term-THM concentrations were reduced by 40-60 percent with the alternative primary disinfectants compared with using C12 as the primary and final disinfectant. Pilot -plant testing was done when THM concentrations are historically highest. Average annual THM values would be lower than those given here. However, the intent was to determine THM reduction effectiveness for the most difficult -to -treat water. Using NH2C1 as the final disinfectant, adding C12 to the primary basin effluent, and using 03, C102, or KMn04 as raw water oxidants -disinfectants produced three-day term-THM concentrations of 30-60 µg/L. This represents a term- THM reduction of 60-82 percent com- pared with using chlorine as the primary and final disinfectant. A 50-Ag/L THM standard could be met using these treatment schemes and the existing water plant processes. The finished -water THM concentra- tions may be lowered further by reducing the C12 contact time. However, previous results show that even with a 1-h C12 contact time in filtered water, THM concentrations were approximately 20- 30 Ag/L. Adequate disinfection should be the first priority. Strong disinfectants like 03 can shorten C12 contact time or eliminate the need for C12 altogether and still provide adequate disinfection. To meet much lower THM standards, the use of C12 may need to be eliminated. One method would be to use 03 as the primary disinfectant and NH2Cl as the final disinfectant. This was tested in alternative 3C, for which three-day term- THM concentrations were 2-3 pg/L. Another option could be addition of KMn04 to the raw water, C102 to the filter influent, and NH2C1 to the finished water. Although this alternative was not tested, THM concentrations would be expected to be low because C12 is not present. Adequate disinfection with C102 and acceptable disinfection by-product concentrations would have to be ensured. Preozonation did not significantly reduce term-THMs compared with the use of C102 or KMn04 when both Cl2 and NH2C1 were used as final disinfectants. In another study, C102 used as a pre - oxidant was found to decrease THM precursor concentrations by more than 10 percent only when TOC concentra- tions were <4 mg/L. , Some investiga- tions found THMs to decrease with the use of 03,6,7 whereas others found in- creases in THM levels with the use of 03 and C12 or NH2C1 as final disinfectants. 8.9 Varying results may be caused by differ- ences in the organic composition of the water and the formation of THM pre- cursors after ozonation.10 In some cases the presence of bromide produces a higher concentration of brominated THMs after ozonation.11 Results from this investigation showed little difference in brominated THM species with or without ozonation. Brominated THM concentrations were between 10 and 20 percent of total THMs. A 30 percent reduction in term-THM concentration was achieved at plant 1 by using H202 with 03 compared with 03 alone. However, term-THM concentra- tion increased with H202 and 03 at plant 2. Some studies have shown THM reduction using H202 and 03,12 whereas others have shown little or no change.10 Granular activated carbon filtration produced three-day term-THM levels of 6-20,ug/L when C12 was used as the final disinfectant. The nonabsorbable fraction of THM precursors (approximately 2-10 percent of raw water precursors) would make it difficult to meet a standard of 10 µg THMs/L even with virgin GAC. If NH2C1 were used as the final disinfectant after GAC, THM levels could be very low if THMs did not form upstream of GAC or THM breakthrough did not occur. Adequate disinfection must be verified when using NH2C1 after GAC treatment. Additional GAC pilot tests were carried out to determine the capacity of GAC when a chlorinated secondary settling basin effluent is treated. Although re- sults indicate that an average of 44 percent THM reduction may occur dur- ing three months of operation, low THM levels would be difficult to achieve and costs would be high. The high costs of GAC for THM reduction have been noted by others.13 Studies have shown signif- icant breakthrough of THMs in fixed bed adsorbers after 10 weeks of opera- tion.14 Chloroform was the least strongly adsorbed of four THM species. Air stripping is another treatment alternative for removing THMs after they form. Although air stripping was not pilot tested, a simplified analysis of effectiveness can be performed from the available data. The pilot -plant results indicate that with raw -water chlorina- tion, approximately 50 percent of the three-day term-THMs would be formed during the 20- to 24-h period in the plant and the other 50 percent in the distribu- JUNE 1990 ANTHONY G. MYERS 83 tion system (Figure 6). Assuming that air stripping is the final treatment process, C12 is the final disinfectant, and 95 percent THM removal can be accom- plished with air stripping, the remaining THM formation potential could still produce 80-90 µg/L term-THMs in the two water supplies studied. Air stripping does not remove most THM precur- sors.Is.ls Other investigators found that >90 percent THM removal could be achieved with air stripping.Is Removal of THMs by air stripping decreases as the amount of brominated THM species increases. Chloroform accounted for >80 percent of the total THM concentration in the two water supplies studied. Using NH2C1 as the final disinfectant after air stripping would produce much lower term-THMs. But THM levels would depend on the removal effective- ness of air stripping, and adequate disin- fection with NH2C1 would have to be demonstrated. One last consideration is that raw -water chlorination may produce undesirable by-products not easily re- moved by air stripping. Cost of treatment alternatives Capital and operating costs were estimated for each treatment method tested. A summary of capital, operating, and total costs (amortized capital plus operating costs) is shown in Table 6. Capital cost estimates are for a water plant designed for a maximum flow rate of 4 mgd (10.5 m3/min). The operating and maintenance (0&M) cost estimates are based on an average -day demand of 1.5 mgd (3.9 m3/min) and are for addi- tional chemicals, energy, and equipment 0&M only. Combinations of the various chemical treatment methods could be grouped into a treatment alternative and the costs determined by adding the costs of the individual chemical treatment meth- ods given in Table 6. Chlorine was not included because most treatment plants have chlorination facilities. Some cost benefits of alternative disinfectants were not quantified. For example, 03 may decrease C12 and alum chemical costs." These benefits will vary with different water sources. Table 7 shows some potential treat- ment alternatives for meeting various THM standards. As indicated, costs increase steeply as the MCL for THMs is lowered. In the two water plants studied, a 50-µg/L THM standard could be met by adding ammonia to finished water at a cost of $0.013/1,000 gal ($0.003/m3). Potassium permanganate, chlorine diox- ide, and ammonia may be needed at various points in the treatment process to meet a 25-µg/L or lower THM standard at a cost of $0.105/1,000 gal ($0.028/m3). A low THM standard of 5 µg/L or less may require 03 and NH2C1 at a cost of $0.164/1,000 gal ($0.043/m3). The cost- effective alternative for a particular water plant and THM standard must be determined on a case-by-case basis. Summary and conclusions The conclusions of the pilot study are based on a limited amount of data. How- ever, the following general results of the study may help smaller water utilities prepare to produce water with lower THM concentrations. • Eliminating raw -water chlorination, using an alternative primary disinfec- tant, and optimizing coagulation for THM precursor removal are a good first step in reducing THMs. However, if C12 is used as the final disinfectant, THM precursor removal with conventional treatment may not be adequate to reach required THM levels. • Using NH2C1 instead of C12 as the final disinfectant is a relatively inex- pensive method of THM reduction, but full-scale tests should be conducted to determine whether it is an acceptable distribution system disinfectant. Tests may include using NH2Cl for one year and monitoring for coliforms, hetero- trophic plate count, and disinfectant residual in an isolated portion of the distribution system. Appropriate public notification must also be carried out. • Lower THM standards may require combinations of disinfectants besides chlorine. • Treatment costs rise steeply as lower THM levels are desired. • Granular activated carbon can be effective for removing THM and THM precursors, but costs may be prohibitive for small water systems. • Air stripping may be effective for THM removal if NH2C1 is an acceptable distribution system disinfectant. Tri- halomethanes may continue to form following air stripping if C12 is used as the final disinfectant. • Water treatment plants must be evaluated individually to determine the appropriate approach for reducing THM concentrations. Extended pilot- or full- scale tests are recommended before implementation of a new treatment method. Acknowledgment The author thanks Macon Municipal Utilities and the city of Moberly, Mo., for making this project possible and for their help and cooperation throughout the project. Analytical assistance from the St. Louis Water Department is also appreciated. Comments and suggestions from Roger Yorton, Bill Bellamy, Bob Chapman, and Alan Scrivner at CH2M HILL Inc. are appreciated. References 1. RUBEL, R.E. & EDZWALD, J.K. Removing Trihalomethane Precursors by Coagu- lation. Jour. AWWA, 79:7:98 (July 1987). 2. Identification and Analysis of Organic Pollutants in Water. Ann Arbor Sci. Publ., Ann Arbor, Mich. (1976). 3. Standard Methods for the Examination of Water and Wastewater. APHA, AWWA, and WPCF, Washington, D.C. (15th ed., 1980). 4. AMY, G.L. ET AL. A Statistical Analysis of Surrogate Parameters for Predicting Trihalomethane Formation Potential. Proc. 1984 AWWA Ann. Conf. 5. WERDEHOFF, K.S. & SINGER, P.C. Chlorine Dioxide Effects on THMFP, TOXFP, and the Formation of Inorganic By -Products. Jour. AWWA, 79:9:107 (Sept. 1987). 6. ROBERTSON, J.L. & ODA, A. Combined Application of Ozone and Chlorine or Chloramine to Reduce Production of Chlorinated Organics in Drinking Water Disinfection. Ozone Sci. & Engrg., 5:79 (1983). 7. RECKHOW, D.A. & SINGER, P.C. The Removal of Organic Halide Precursors by Preozonation and Alum Coagulation. Jour. AWWA, 76:4:151 (Apr. 1984). 8. LAWRENCE, J. The Oxidation of Some Haloform Precursors With Ozone. 3rd Intl. Sym. on Ozone Technology, Intl. Ozone Inst., Paris, France, May 1977. 9. UMPHRES, M.D. ET AL. The Effects of Preozonation on the Formation of Tri- halomethanes. Ozonews, 6:3, Part 2 (1979). 10. JACANGELO, J.G. ET AL. Ozonation: Assess- ing Its Role in the Formation and Control of Disinfection By -Products. Jour. A WWA, 81:8:74 (Aug. 1989). 11. DORE, M. ET AL. Interactions Between Ozone, Halogens, and Organic Com- pounds. Ozone Sci. & Engrg., 10:53 (1988). 12. WALLACE, J.L. ET AL. The Combination of Ozone -Hydrogen Peroxide and Ozone -UV Radiation for Reduction of Trihalometh- ane Formation Potential in Surface Wa- ter. Ozone Sci. & Engrg., 10:103 (1988). 13. LYKINS, B.W. ET AL. Granular Activated Carbon for Controlling THMs. Jour. A WWA, 80:5:85 (May 1988). 14. Chlorination By -Products: Production and Control. KIWA Communication 74. AWWARF, Denver, Colo. (Jan. 1986). 15. SYMONS, J. M. ET AL. Removal of Organic Contaminants From Drinking Water Using Techniques Other Than Granular Activated Carbon Alone: Progress Report. USEPA, Cincinnati, Ohio (1979). 16. SINGLEY, J.E. ET AL. Trace Organics Removal by Air Stripping. AWWARF, Denver, Colo. (1980). 17. WAGNER, R. & ELEFRITZ, R.A. Ozonation for Effective THM Control. Public Works Magazine (Apr. 1983). About the author: Anthony G. Myers is an environmental en- gineer in the Water Supply and Treat- ment Division of CH2M HILL Inc., 310 W. Wisconsin r► Ave., P.O. Box 2090, Milwaukee, WI 53201. He is a graduate of Michigan Technological University, Houghton (BS), and the University of Illinois, Urbana (MS), and is member of A WWA and WPCF. 84 RESEARCH AND TECHNOLOGY JOURNAL AWWA Evaluation of Treatment for Removing Giardia Cysts Jerry E. Ongerth The removal of Giardia cysts by three small (0.01- to 0.57-mgd) treatment plants—one with conventional filtration, one with in-line filtration, and one with diatomaceous earth filtration—was studied. Turbidity and cyst removals for each were compared with those of parallel pilot filters (1 gpm) seeded with Giardia cysts and with and without optimization of chemical treatment. Cyst removal efficiencies ranged from about 40 to 99 percent. Design and operating deficiencies likely contributed to the observed treatment performances. Major deficiencies included lack of or poor optimization of chemical treatment, on-off cycles of filters without backwashing, absence of operator control of backwashing, and no filtering to waste. Giardiasis is the most prevalent water- borne disease in the United States. Outbreaks generally occur in small water systems that use surface water without filtration. Several outbreaks have oc- curred, however, in water systems that employ complete treatment or at least simple filtration. In some instances in which complete treatment or filtration was employed, incorrect design, installa- tion, or operation of the processes may have contributed to the loss of system integrity. In other instances, no apparent problem could be detected nor could cysts be recovered from the water. Inspection of water treatment plants associated with outbreaks of giardiasis has shown that physical defects in plant construction or lack of effective operation or maintenance frequently results in increased potential for cyst passage. At Camas, Wash., an inspection of the multimedia pressure filters revealed significant loss of media and mounding of support gravel.' This, combined with less -than -optimum chemical condition- ing, could have resulted in cyst passage. In one plant at Berlin, N.H.,2 pressure filters had uneven media distribution, mud masses, and media clogging; sepa- ration of media from chamber walls; and malfunctioning of the air -scour system. Also at Berlin, at a newer, conventional gravity filtration plant, a construction defect was found in a common wall between settled- and filtered -water areas. This could have permitted cross -con- tamination of finished water by Giardia - containing raw water. At Reno, Nev., comparisons made between samples of finished water from JUNE 1990 one plant with direct filtration (without sedimentation) and from a second plant with only coagulation and sedimentation (without filtration) showed that under some conditions more cyst -size particles were found in the filtered water than in the water receiving only chemical treat- ment and sedimentation.' Effective cyst removal can be achieved by properly operated granular media filtration of various designs and opera- tional configurations. Tests show, how- ever, that some filter passage (2 to 8 percent of cyst -size particles) occurs in normal operation under conditions of heavy cyst loading.4 This may occur in all types of filters. Tests indicate that slow sand filtration is particularly effec- *A full report on this project is available from the AW WA Research Foundation, 6666 West Quincy Ave., Denver, CO sons. JERRY E. ONGERTH 85 tive for removal of cysts and cyst -size particles from low -turbidity water.5-7 Tests with diatomaceous earth (DE) filters suggest that they also are effective for cyst removal.8.9 Raw water that is cold and has low turbidity and alkalinity is hard to coag- ulate and filter to achieve efficient turbidity removal.4.6.10 Filtered water characteristically has a higher -than - average turbidity when granular media filtration is at one of the following three operating stages: (1) immediately follow- ing backwash, (2) at the end of the filter cycle just prior to backwash, and (3) depending on circumstances, when sig- nificant changes in flow rate, turbidity, or chemical treatment occur during a filter run. 11-13 Turbidity removal is typi- cally low following backwash as the filter bed becomes reconditioned by the chemically conditioned water passing through it. The period may last from a few minutes to an hour or more. Poorer - quality water produced in this period has led to the practice of wasting the filtered water during the reconditioning period. Pilot -scale tests indicate that cyst -size particles may pass through filters in relatively large numbers during this period.4.6,8 Increased effluent turbidity at the end of a filter run may result in passage of cysts and cyst -size particles.4.6.8 Some work indicates that cysts migrate through a filter bed as a function of filter run length and chemical conditioning. 4.6 Short-term increases in filter effluent turbidity associated with discontinuities in flow rate and major changes in tur- bidity or chemical conditioning have also been shown, in pilot studies, to result in increased passage of cysts or cyst -size particles.8 The primary objective of this project was to investigate and evaluate full- scale water treatment plants, at normal water quality and operating conditions, for the removal of Giardia lamblia cysts. The project focused on the design char- acteristics of water treatment processes considered critical in cyst removal and on operational practices that affect cyst removal. Disinfection was not included in this investigation. Methods Approach. Three types of treatment were studied: (1) a 0.51-mgd conventional treatment plant (operated without addi- tion of coagulant chemicals) at commu- nity A in Washington, (2) a 4-mgd pressure -gravity in-line filtration plant at community B in California, and (3) a 0.085-mgd DE filtration plant at com- munity C in Washington. A slow sand filtration plant was examined in a companion project sponsored by the AWWA Research Foundation.14 Two-week field studies at each plant included (1) monitoring the full-scale TABLE 1 Design speeifieiations for pilot plant components Component Type Size Detention Time Loading Rapid mix In-line, static 10 -in. x 1 -in. diameter 2s 1 gpm Flocculator Paddle, tapered 3 eachat 13 gal, 3 x 13 min 1 gpm variable -speed G = 0-1500 Sedimentation Rectangular, upflow 3.25 cu ft, 12 x 24 x 23 min 3,165 gpd/sq ft with tube settlers 20 in. at 30° 5 min Filters Multimedia, depth 0.2 sq ft 5 gpm/sq ft varies Backwash 2.5 gpm with air Controls Automatic -manual for 0-10 gpm backwash, manual needle valve for flow rate Alum Backwash water disposal pond Crystal Creek pipeline xculstloe Sedimentation Fitter section "cum Spent backwash Flocculation Sedimentation section section Flow control orifice Flash mix Fitter I I I I I I Backwash storage resenoir — Distribution system Figure 2. Schematic diagram of community A's full-scale treatment plant 100 eo Bo 40 20 Et Backwash 4/9 4/10 4/11 4/12 4/13 4/14 4/18 4/18 4/17 4/18 Dab -1985 Figure 4. Turbidity removal perfor- mance for full-scale plant in commu- nity A Fubtw N waste bac�'icwt ash to wash — I — — — Dlstrlbtttbn Fitter 4 storage reservoir community B Fitter 3 polymer supply tank Fln.r 2 I I I 1 I esckwash Filter 1 water L — — — — disposal pond `Screen chamber From t___N Gray Eagle Creek Figure 6. Schematic diagram of community B's full-scale treatment plant 86 RESEARCH AND TECHNOLOGY JOURNAL AWWA TABLE 2 Chemical conditioning used in operation of pilot plant I in community A Date Run Number* Time Chemical and Dose pH 4/12 1 11:25 a.m. Alum -10.5 mg/L 6.6 2 1:10 P.M. Alum -10.5 mg/L 6.5 3 2:10 p.m. Alum -10.5 mg/L 6.9 4 2:50 p.m. Alum -14.8 mg/L 6.4 5 4:45 p.m. Alum -14.8 mg/L 6.2 4/16 6 5:30 p.m. Calgon 233 as filter aid -0.33 mg/L, plus 6.5 alum -10.4 mg/L 7 4:10 p.m. Calgon 233, as filter aid, only- 6.5 0.0245 mg/L 8 6:00 p.m. Calgon 233 as filter aid -0.024 mg/L, 6.5 plus alum -10.4 mg/L 8A 10:00 P.M. Calgon 233, as filter aid, only -0.033 mg/L 6.5 4/17 9 9:00 a.m. Alum -16.5 mg/L 6.5 4/18 10 4:00 p.m. Alum -14.5 mg/L 6.5 *See Figure 5 0.6 Raw water -- Filtered water 0.5 0.4 ANN, a t 1, 0.3 c 0.2 My 0.1 ttactcwasn Backwash Backwash 0.0 , 4/5 4/6 4/7 4/8 4/9 4/10 4/11 4/12 4/13 4/14 4/15 4/16 4/17 4/18 Date -1985 Figure 3. Raw- and filtered -water turbidities for full-scale plant in community A 0.6 oil) a I IN �� I Backwash & 0.3 II Influent i ! J� •»«» pilot atter 1 (chemical treatment optimization) 0.1 -- Pitot filter 2 (no chemical treatment) ? ?♦ ♦ ♦♦ �v Run 1 2 3 4 5 6 7 8 8 10 oo 4/5 4/6 4/1 4/8 4/9 4/10 4/11 4/12 4/13 4/14 4/15 4/18 4/17 4/18 Dale -1985 Figure S. Raw- and filtered -water turbidities for pilot filters in community A, showing effect of chemical conditioning plant for cyst removal and treatment performance; (2) operation and monitor- ing of the pilot plant to supplement the cyst removal and performance data from the full-scale plant and to determine optimum chemical treatment conditions; and (3) a sanitary survey of the water- shed and the treatment plant. Sampling for Giardia focused on periods associated with turbidity passage through the fil- ters: (1) during filter conditioning im- mediately after backwash, (2) imme- diately before backwash as turbidity breakthrough or time -limit criteria were reached, and (3) during mid-cycle, under normally peak efficiency operation, when performance was affected by flow rate changes or off -on cycles. Optimization of chemical treatment was carried out with one of two parallel pilot -plant treatment trains. Turbidity removal was examined as a function of pH and of alum and polyelectrolyte concentrations. Giardia cyst removal and turbidity were monitored after the optimal conditioning dosage had been established. Measuring performance. Treatment plant performance was measured by the continuous monitoring of raw- and fil- tered -water turbidity, periodic particle counts of raw and filtered water, and periodic enumeration of Giardia cysts in raw and filtered water. The effectiveness of chemical conditioning was evaluated by comparing the full-scale plant per- formance with pilot -plant performance under optimum conditions. Sampling. Turbidity of the raw and filtered water of both the full-scale and pilot plants was monitored continuously. Samples for cyst enumeration were collected from sampling taps at the turbidimeters in 20-L plastic. buckets. Samples for particle counts were col- lected in 20 -mL counting vials* at the sampling taps. Samples for Giardia enumeration were collected (1) imme- diately following backwash at the start of the filter -conditioning period, (2) just prior to backwash when the filters were most heavily loaded with particulates, and (3) during mid-cycle under normal filter operating conditions. Analyses. Giardia cyst concentrations were determined by membrane filtration with selective cyst concentration15 and immunofluorescence assay cyst iden- tification.16 Raw- or filtered -water sam- ples of from 10 gal (40 L) tn 120 gal (480 L) were filtered under a 20 -in. -Hg (10 -psi) vacuum through a 293 -mm -diameter, 5- µm -pore -size polycarbonate membrane.t Particles, including Giardia cysts, were recovered from the filter membrane by the method of DeWalle.8 Recovered particles were concentrated by centrif- uging and decanting and were then •Coulter Electronics. Hialeah. Fla. Muclepore, Pleasanton, Calif. JUNE 1990 JERRY E. ONGERTH 87 TABLE 3 Community A -Giardia cyst removals in Pilot treatment trains with and without chemical conditioning Pilot Sampling Sample Cyst Concentration-number/gal Overall Removal Cyst Removal percent Plant Sedimentation Filter Train Date Number Influent Effluent Effluent percent Comments I (chemical optimization) 4/17 16,18 527 13 98.0 0-15 after backwashing 1.1 95 19 1.5 99.7 15-30 after backwashing 20 28 0 1.1 30-45 after backwashing 24,25,26 527 0.6 99.9 4/18 28 0.03 0.9 99.8 0-15 after backwashing 6/10 31 29 7.2 71 0 5-30 after backwashing 30 2nd 10 gal 0 30-45 after backwashing 12.5 31 3rd 10 gal 0 33 45-60 after backwashing 22.3 7 40 528 6/11 0 15 45 528 42 0 25.2 37 528/750 347 38 660/750 21.1 94 2 (without chemical 4/09 1,2,3 426 362 41 90.3 0-15 after backwashing conditioning) 4 Negative control = 0 130 69.0 15-30 after backwashing 18.0 5 194 54.0 30-45 after backwashing 6 2 (without chemical 6/5 292 31.0 45-60 after backwashing 4/12 14,13,12 202/650 185 54 73.0 13 1st 10 gal 15 519/650 18.2 9 2nd 10 gal 4/17 16,17 527 91 16 3rd 8.5 gal 7 27 650 143 78.0 6/6 10 4/18 39 33 11 34 41 528 27 97.0 0-15 after backwashing 8.1 75 42 153 71.0 15-30 after backwashing 80 1st 10 gal 44 20 155 70.0 30-45 after backwashing 2nd 10 gal 44 21 73 86.0 45-60 after backwashing 46,47 1 20 1 12 79 4th 10 gal TABLE 4 Giardia cyst removals in Pilot treatment trains in community B with and without chemical conditioning Pilot Train Date Sample Number Influent Cyst Concentration number/gal Effluent Cyst Concentration number/gal Cyst Removal percent Comments 1(with chemical 6/8 24 3.8 84 optimization) 25 0 >99 26 1.1 95 27 24 28 1.1 95 29 0.03 >99 30 0.03 >99 1st 10 gal 6/10 31 7.2 71 2nd 10 gal 32 12.5 48 3rd 10 gal 33 22.3 7 4th 10 gal 6/11 37 15 42 25.2 6/12 47 347 49A 21.1 94 49s 50 Negative control = 0 6/13 53 18.0 84 54 27 2 (without chemical 6/5 3 20 conditioning) 4 17.5 13 1st 10 gal 5 18.2 9 2nd 10 gal 6 16.8 16 3rd 8.5 gal 7 6/6 10 22.8 33 11 34 6/7 14 8.1 75 19 6.8 80 1st 10 gal 20 16.0 53 2nd 10 gal 21 11.1 67 3rd 10 gal 22 7.3 79 4th 10 gal 6/11 35 13.4 61 36 17 44 6.5 62 1st 20 gal 45 18.8 2nd 20 gal 6/12 46 11.0 35 3rd 24 gal 6/13 51 Negative control = 0 55 34 3.5 1 89 88 RESEARCH AND TECHNOLOGY JOURNAL AWWA formalin -fixed, refrigerated, and shipped by overnight express to the California Department of Health Services Labora- tory for processing as described by Riggs.16In summary, the Riggs procedure consists of centrifuging the concentrated particles on 40 percent potassium citrate, rinsing, and filtering the final cyst -con- taining particles onto a 25 -mm -diameter, 5 -µm -pore -size polycarbonate mem- brane.* The particles were stained by a fluorescein -labeled, polyclonal, guinea pig anti -Giardia antibody. The stained filters were mounted on glass slides and examined by ultraviolet illumination microscopy at 100 or 250X to enumerate the Giardia cysts. Turbidities were monitored using a turbidimetent Particles in the Giardia cyst size range (5-20 µm) were counted using an electronic particle counter.$ Temperature was measured using a certified glass capillary mercury ther- mometer. The pH was measured using a meter with a combination electrode.§ Sanitary survey. Sanitary surveys were conducted to evaluate factors related to the presence of Giardia cysts in the raw water and to their removal (or nonre- moval) in treatment. The surveys were divided into four categories: (1) the watershed, (2) water quality history, (3) plant physical description, and (4) plant operation. Pilot plant To ensure that useful data would be obtained on Giardia cyst removal, a mobile pilot plant was taken to each plant site and was modified to provide physical and operating conditions as much as possible like those of the full- scale plant. Heat -inactivated G. lamblia cysts were added to the pilot -plant in- fluent to allow the monitoring of cyst removal. The pilot plant (Figure 1) con- sisted of two parallel, hydraulically independent, complete treatment trains, each of 1-gpm capacity, housed in a 10 ft X 32 -ft mobile trailer. Design specifica- tions of the pilot -plant components are summarized in Table 1. The clear acrylic filters were 6 in. in diameter and 8.3 ft long. Piping provided for downflow filtration and backwash with air scour and hydraulic surface wash. The filters had flanged ends permitting the media to be changed to meet individual treatment plant specif- ications. The 6 -in. -diameter filters (0.2 sq ft) were designed for a normal operat- ing flow rate of 1 gpm or 5 gpm/sq ft. One of the two pilot treatment trains was used to simulate operation of the full-scale plant. The other pilot treat- ment train was operated under identical physical conditions but with altered chemical conditioning to assess its effects on cyst removal. Inactivated G. lamblia cysts were fed continuously to the in- fluent of the pilot treatment trains. A 0.93 -sq ft pilot DE filter** was used during field investigations at community C. The pilot filter package included a single filter leaf with housing, tanks, and pumps for precoat and body feed and for control of the filtration rate. Field investigations and results Community A. Physical characteristics of plant. This plant, located in the foot- hills of the Cascade Mountains, served a population of 1,100 people. Source water was an impoundment of about 30 acre -ft (10 mil gal) on a small stream draining a watershed of about 100 acres. The plant was a 400-gpm conventional facility, though during this study and for many months preceding it, no condi- tioning chemicals were applied. Principal components included an influent flow- meter, chemical feed facilities and injec- tion pumps, a static in-line flash mixer, and two parallel 200-gpm treatment modules, each with a baffled flocculation section, a sedimentation section with tube settlers, and dual -media gravity filters (Figure 2). The media design included 12 in. of graded gravel (3 in. each of #1, #2, #3, and #4);16 in. of sand (3 in. of F5, 2 -mm effective size [es]; 3 in. of F9,1 -mm es; 10 in. of F16,0.45 -mm es with 1.5 uniformity coefficient [uc]); and 18 in. of coal (C10, 0.9 -1.0 -mm es with 1.5 uc). Backwash was controlled auto- matically on a timed cycle. The plant was designed for constant - head, declining -rate operation regulated by a float -controlled throttling valve on the filter effluent line. Observation of plant operation indicated a constant cycling of the water surface of approx- imately 3 -in. rise and fall. Because of limited storage capacity, the plant runs continuously for long periods of time during warm summer days to meet demand. During the night and when the weather is cool, only intermittent opera- tion is required. Thus, much of the year, typical operation includes numerous on-off cycles between backwashes. Plant performance. Investigations were conducted from Apr. 1 to Apr. 18, 1985. Plant monitoring records from mid -1981 through 1984 showed turbid- ities were typically low from November through March. Turbidities increased almost continuously from spring through fall. Although raw -water turbidity varied between 0.25 ntu in winter and 0.7 ntu in the fall, turbidity of the filtered water was consistently about 0.1-0.2 ntu less than that of the raw water. Filtered - water turbidities ranged from 0.12 to 0.58 ntu on a monthly average basis. The lack of a more consistent filtered - water turbidity was due partly to the practice of not applying coagulating chemicals. The operator rationalized this on the grounds that the filtered -water turbidity generally met nominal drinking water standards. From April 1 to 18, plant production averaged about 160 gpm. Because the plant can operate only at the fixed rate of 400 gpm, it was filtering for about 40 percent of the time. Filtration occurred during five to eight periods per day, ranging in length from 1 to 2 h. The plant backwashed five times during the period of observation, on April 1, 5, 9, 12, and 18. Raw -water turbidity ranged from 0.25 to 0.5 ntu (Figure 3). Finished water was consistently about 0.2 ntu below the raw -water turbidity. As in- dicated by the turbidity removal ef- ficiency, overall treatment performance ranged from 45 to 70 percent except for periods immediately following back- wash (Figure 4). Eleven samples from the full-scale plant were analyzed for Giardia cysts, two of raw water and nine of finished water. Cysts were found in one of the two raw water samples at approximately 0.25 cyst/L. Cysts were found in seven of the nine filtered water samples, ranging from 0.012/L to approximately 0.25/L. Cyst concentrations were generally highest in filtered -water samples taken immediately following backwash. The concentration of cyst -size particles in raw and finished water ranged from 100 to 500/mL and had a positive corre- lation to turbidity. Particle concentra- tions in the filtered -water samples were 15 to 70 percent lower than in the raw water. The removal of cyst -size particles, however, did not appear to be closely related to turbidity removal. Pilot plant. Fresh filter media were washed and installed on April 1. The pilot -plant influent line was connected to the full-scale plant influent line on that date, and the pilot plant was oper- ated without monitoring, other than flow rate adjustment, until April 5. Monitoring of influent and effluent turbidity began on April 5, and addition of Giardia cysts to the pilot plant began on April 9. Detailed pilot -plant monitor- ing was conducted alongwith monitoring of the full-scale plant from April 9 through 18. During operation the pilot - plant flows were maintained at 1 gpm, which resulted in a clarifier overflow rate of 3,165 gpd/sq ft, a detention time of 23 min, and a filtration rate of 5 gpm/sq ft, approximately equal to those of the full-scale plant. Pilot -plant opera- tion was tied directly to that of the full- scale plant by connecting the pilot -plant raw -water supply line to the main plant supply line downstream of the influent shut-off valve. Therefore, the pilot plant operated for the same five to eight periods per day as the full-scale plant, each period ranging from about 1 to 2 h. *Nuclepore, Pleasanton, Calif. tModel 1720A or 2100A, Hach Co., Loveland, Colo. $Model Z131, Coulter Electronics. Hialeah, Fla. §Model 100, Orion Research, Cambridge, Mass. **JM Corp., Denver, Colo. HINE 1990 JERRY E. ONGERTH 89 Pilot filter 2 was operated exactly like the filters of the full-scale plant, with no chemical conditioning, and was back - washed manually when the community plant was backwashed. Pilot filter 2, operating with new media (identical to those of the full-scale plant), was inca- pable of reducing the turbidity of the feedwater (Figure 5). In fact, the fresh filter media, although prewashed, con- tinued to release fines, which contributed to an effluent turbidity consistently higher than that of the raw water. Pilot filter 1 was used to establish optimal chemical conditioning combina- tions and to determine the treatment performance capabilities of the plant, including cyst removal, when operated with chemical conditioning. Chemical treatment evaluations were made for 10 separate runs on April 12,16,17, and 18 (Table 2). Performance of this pilot filter was significantly better than that of the full-scale plant for several of the condi- tioning combinations (Figure 5). Giardia cysts were fed continuously to pilot filters 1 and 2 at concentrations ranging from 500 to 750 cysts/L. A total of 37 samples and controls were collected from the two pilot filters for Giardia cyst enumerations. Cyst removals observed for pilot filter 2 ranged from 31 to 90.4 percent (Table 3). Cyst removal efficiency during mid -run operating conditions averaged about 75 percent (samples 12 and 27). Continuous samples collected in consecutive 15 -min operating periods beginning immediately following back- wash indicated lower removal efficiency during the initial 60 min. The somewhat higher removal observed in the initial 15 min following backwash was likely due to the volume of cyst -free backwash water remaining in the filter column, which became part of the sample. Chemical conditioning. Several combinations of alum, with and without pH control, and polyelectrolyte used as a filter aid were tested to determine their effects on turbidity and cyst removal (Table 2). Use of alum alone at about 10 mg/L or in combination with the nonionic polyelectrolyte used as a filter aid at 0.03 mg/L significantly improved turbidity removal and filter stability. At optimal pretreatment, filtered -water turbidity ranged from 0.03 to 0.07 ntu (Figure 5). Giardia cyst removal by pilot filter 1 during optimal chemical conditioning ranged from 98 to >99 percent. Eleven samples of filtered water were collected during chemical addition; cysts were found in four, at a relatively low concen- tration, during the initial 30 min follow- ing backwash. No cysts were found in the samples collected after the 30 -min period (Table 3). Sanitary survey. A survey of the watershed and treatment plant identified several factors that indicated the likely presence of Giardia cysts in the raw water and the potential for cyst carryover into the finished water. The watershed is a small, heavily wooded basin drained by an unnamed creek flowing into the raw -water im- poundment. The watershed provides an unmanaged and relatively secluded animal habitat. Evidence of beaver and muskrat was found at various locations. Dense undergrowth and the absence of litter suggest that sources of Giardia are likely animal rather than human. Several physical features provided an opportunity for contamination. The clearwell is directly beneath the floor slab of the treatment units. Spent back- wash water is collected in a sump formed by the floor slab and concrete wall sec- tions between the two parallel treatment modules. A crack in the floor slab would permit passage of the backwash water directly into the clearwell. Drainage from the road adjacent to the plant can percolate into the ground beside the subgrade clearwell. A wooden privy serving the plant stands at the corner of the plant building. Leakage from it would present a serious source of contamination should the concrete plant walls develop cracks into the interior. Several factors of plant operation contribute to the potential for passage of Giardia cysts. The plant is operated by one person who has multiple mainte- nance responsibilities in the community. He spends an average of one hour per day at the plant. Most significant, however, is the complete absence of chemical conditioning, although chemical feed equipment and coagulants are on hand. The hydraulic control scheme causes the plant to cycle on and off from five to eight times in a 24-h period. This imposes hydraulic shocks on the clarifers and filters, whose performance is already compromised by the lack of chemical pretreatment. The backwash cycle is initiated automatically by a timer. Therefore, backwashing normally occurs when the operator is not present to observe its effectiveness. Backwash conditions are thus never adjusted. Community B. Physical characteris- tics of plant. The plant served a recrea- tional community in the mountains northeast of Sacramento, Calif. The sys- tem had 600 service connections. Winter use was about 0.29 mgd (200 gpm). Peak demand in summer months was 2.9 mgd (2,000 gpm). The climate was typical of the western mountains, little different from communities A and C. The raw -water source was a creek, which drained a watershed of about 7 sq mi. The intake was at a small diversion dam near the treatment plant. Annual water temperatures ranged from <4°C to between 15 and 20°C. Water in the creek at the plant intake was soft, with low alkalinity and total dissolved solids and neutral pH. The treatment plant employed dual - media filtration with addition of polymer as a filter aid. It was considered an in- line filtration plant, having addition of chemicals but no mixing for coagulation other than pipeline turbulence and no settling prior to filtration. The plant had four filters, each with two independent cells (Figure 6). The filters were in closed, horizontal, cylindrical steel tanks, 6 ft in diameter by 24 ft long. Filter media included 9 in. anthracite, 0.6-0.8 mm es, 1.85 uc; 12 in. sand, 0.45-0.55 mm es, 1.45-1.55 uc; 6 in. fine gravel; 9 in. medium gravel; and 6 in. coarse gravel. The plant could operate under gravity feed or under pressure from parallel 1,900-gpm centrifugal pumps. Plant capacity under gravity flow was from 2.0 to 2.6 mgd (1,400 to 1,800 gpm). Design capacity with pumping was 4.0 mgd (2,800 gpm), corresponding to a filtration rate of 5.3 gpm/sq ft. The listed nominal filtration rate was 3.75 gpm/sq ft, corresponding to a total plant capacity of about 2.8 mgd (2,000 gpm). Plant operation. Raw water flowed in an open channel from the diversion dam to a screen chamber, then by pipe to the plant, for gravity or pressure filtra- tion. The pumps were used primarily during peak demand in the summer. A cationic liquid polymer* was pumped into the filter supply line about 20 ft upstream of the filters. The polymer concentration used was 0.25-0.5 mg/L. The only mixing was from turbulence in the filter face piping. During the winter, when the total water demand averaged 200 gpm, only two filters were used with gravity feed. Two filters operating in this mode produced about 800 gpm. Thus, the plant operated in off and on cycles con- trolled by switches that sensed the water level in the finished -water reservoir. With a 200-gpm demand, the plant operating cycle was about 1 h on and 4 h off , or about five complete cycles per day. When the demand exceeded 800 gpm, all filters were operated by gravity until pumped flow was required. Regardless of the number of filters in operation or of the operating mode, the production was matched to demand by the level controls in the distribution reservoir. The filters under gravity flow were essentially op- erated with constant head and declining rate. When operating under pump pres- sure, the mode was increasing head, declining rate. As head loss across the filters increased, increased pumping head decreased output as dictated by the pump's head -discharge curve. The fil- tered water flowed by gravity to the finished -water distribution reservoir. Filters were backwashed once a week during low -use periods and about once every four days during high -use periods. 08102, Naloo Chemical Co., Oak Brook, III. 90 RESEARCH AND TECHNOLOGY JOURNAL AWWA 0.5 Raw water 0.4 0.3 Filtered water 0.2 1 0.1 �►-1.�1ww....`-- `Z� r-�•-.�-� Flail backwsSh Backwash Backwash 0.0 6/3 6/4 615 6/6 6/7 618 819 6/10 6/11 6/12 6/13 W14 Date -1865 Figure 7. Raw- and filtered -water turbidities for full-scale plant in community B, June 3-14, 1985 The backwash cycle was actuated manually, and backwashing was carried out by a preset, timed sequence of automatic valve manipulation. The two cells of an individual filter were back - washed separately, in sequence. Back - washing of filters was usually staggered, i.e., on different days, to maintain production continuity. Backwashing of an individual filter cell took about 6 min. When a newly backwashed filter was returned to service, filtered water was wasted for 5 min. This was controlled automatically. Plant performance. Field work was conducted from June 1 through June 14, 1985. Water temperature during the study ranged from 12°C at night to 17°C in the afternoons. According to plant records, raw -water turbidity was typically low, averaging <0.5 ntu annually. Recorded peak tur- bidity was as high as 10 ntu. Filtered - water turbidity varied with the raw - water turbidity, averaging between 0.1 and 0.2 ntu. Turbidity removal efficiency ranged between 50 and 70 percent. This study was conducted during a period of low runoff and low raw -water turbidity. Total water production from June 6 to 14 was 1.1 x 106 gpd. Thus, the average production over this period was about 800 gpm on a continuous basis. However, with the plant operating in pumped - pressure mode using all four filters, the production rate was measured at about 1,900 gpm. This means that the plant averaged just over 13 h of production per day from June 6 to 14. In fact, the production period increased gradually from a minimum of about 4 h June 6 to a maximum of 14.5 h June 13-14. The plant production cycle during the study period was typically alternating on and off periods of approximately equal length ranging from about 0.5 to 2 h, resulting in 8 to 12 cycles per day. The cycling frequency was strictly a function of the plant production rate, storage tank volume between the high and low float - switch elevations, and water demand. Plant performance during the study was indicated by the record of raw- and filtered -water turbidities (Figure 7). Raw -water turbidity averaged about 0.3 ntu with brief periods as low as 0.2 ntu and as high as 0.4 ntu. The turbidity of filtered water averaged about 0.09 ntu with little variation other than that associated with backwash. Turbidity removal efficiency was consistently about 70 percent except for lower re- movals observed immediately following backwash (Figure 8). Analysis for Giardia cysts was con- ducted on 18 samples from the plant, 5 of raw water and 13 of filtered water. Cysts were found in 3 of the 5 raw -water samples and in 7 of the 13 filtered -water samples. Based on a recovery efficiency of 20 percent, raw -water cyst concen- trations were uniformly about 0.3 cyst/L, and filtered -water cyst concentrations were between 0.03 and 0.07 cyst/L. In a set of four effluent samples taken at 15 - min intervals immediately following backwash, the cyst concentration de- creased from 0.9/1, to 0.5/1, to 0.1/L to 0/L. No cysts were found in a second set of samples from the preceding backwash period. The concentration of cyst -size particles found in the raw water ranged from about 250 to 500/mL (x = 362, s =126, n = 5). Filtered -water particle concentrations ranged from about 40 to 240/mL (x =110, s = 52, n =14). The removal efficiency of cyst -size particles calculated from paired values of raw and filtered water ranged from about 50 to 90 percent (x = 64 percent, s = 34 percent, n = 6). Based on the limited number of observations made, 100 60 40 20 1= 0 6/4 6/5 6/6 6/7 8/5 6/9 6/10 8/116112 6/13 8/14 Date -1985 Figure S. Turbidity removal perfor- mance for full-scale plant in commu- nity B the concentrations of cyst -size particles in raw and filtered water appeared closely related to corresponding turbidity mea- surements (Figure 9). The removal effi- ciency for cyst -size particles, which averaged 64 percent, was comparable to the observed turbidity removal effi- ciency, which averaged about 70 percent. Pilot plant. Pilot -plant piping and control connections were madeJune 1-3. Following polymer addition, pressurized raw water was piped directly to pilot filter 2. Raw water for pilot filter 1 was taken upstream of polymer addition. New filter media were installed to match those of the full-scale filters. The eleva- tion of the filter media surface in the pilot filters was within a few inches of the full-scale plant media surface. Power for operating the cyst feed and supple- mental chemical feed pumps was taken from the raw -water supply pump control circuit of the full-scale plant. Accord- ingly, pilot -plant flow cycling was closely matched to that of the full-scale filters. Flow through the pilot filters began June 3, accompanied by continuous turbidity monitoring. Pilot filter 2 was operated as a model of the full-scale plant throughout the study. Pilot filter 1 was used for testing chemical -conditioning schemes. The flow rate used for each pilot filter was 0.7 gpm, corresponding to the full- scale plant's filtration rate of 3.5 gpm/sq ft. Pilot filter 2 was backwashed manu- ally for 6 min at the same time that the full-scale plant filters were backwashed. Pilot filter 1 was backwashed as needed. Raw -water turbidities for the full-scale plant and the two pilot filters were identical. The filtered -water turbidity of pilot filter 2 decreased continuously over the period of observation from about 0.2 ntu to 0.13 ntu (Figure 10). Performance was not as good as that of the full-scale plant (Figure 8). This was likely a result JUNE 1990 JERRY E. ONGERTH 91 0 Fall -scale plant 0 6 O Pilot plant 2 • Pilot plant 1 0.4 0.4 0 0 Influent c 0.3 • 000 ? 0.3 ILI Raw water Pilot titter 1 i 9 --------------------------- 0.2 Filtered water 0.2 r "'• 0.1 0% O 0.1 • pilot titter 2 Nalco titaQatfloc 6330 O O 6102 2cow 33 0.0 0.0 0 100 200 300 400 600 6/3 6/4 6/6 8/6 6n hie Gro 6/10 8/11 6/12 6/13 W14 Cyst-St:e Particles—number/mL Figure 9. Relationship between cyst- °at' -'sem size particles and turbidity in com- Figure 10. Raw- and filtered -water turbidities for pilot filters in community B, munity B June 3-14, 1985 of the use of new filter media in the pilot plant. Giardia cysts were fed continuously to pilot filter 1 and pilot filter 2 at concentrations ranging from 200 to 500 cysts/L. A total of 38 samples, including 5 positive and negative controls, was collected from both of the pilot filters and then analyzed to determine Giardia cyst concentrations. Some problems were experienced with cyst breakup (i.e., fragmentation). Based on hemacytometer counts, the cyst feed concentration should have been between 1,000 and 2,500 cysts/mL. Upon analysis of positive controls, only approximately 10 percent of the expected concentration of whole, intact cysts was found. Estimates of the total number of cysts recovered based on cyst fragments found were comparable to the expected number of cysts. The fragmentation of cysts, however, should have affected filter influent and effluent concentrations similarly. In fact, frag- mentation might be expected to bias the effluent samples to the low side more than the influent samples, tending to overestimate cyst removal. With this in mind, the influent and effluent cyst concentrations for the seeded pilot plant are summarized in Table 4. Cyst removal efficiency observed for pilot filter 2 was relatively low, averaging 64 percent (s = 24, n = 4) for four samples not taken immediately following back- wash. Cyst removal observed for pilot filter 1 was significantly higher than for pilot filter 2. The average removal, including three samples taken imme- diately following backwash, was 79.5 percent (s = 28.6, n =11). Excluding the three low postbackwash removal values, the average cyst removal in the filter subjected to additional chemical condi- tioning was 93.6 percent (s = 6.3, n = 8). As was observed in samples collected in the full-scale treatment plant, cyst pas- sage to the finished water appeared greater in the initial 30-60 min following filter backwash. Chemical conditioning. Several chemical conditioning combinations were used, with various dosages of alum and cationic polymers. Flash mixing of the chemical coagulants was provided in pilot plant 1, but no flocculation. Using between 10 and 16 mg alum/L in com- bination with the polymer used in the full-scale plant produced filtered -water turbidity of 0.06-0.20 ntu (Figure 10). Use of polymers (at 0.3 mg/L) did not produce finished water significantly different from the current plant treat- ment. When alum was used as the coagulant for pilot filter 1, cyst removal was higher than with the polymer treatment scheme used at the full-scale plant and duplicated in pilot filter 2. Sanitary survey. The sanitary survey of the watershed and treatment facility identified features that are pertinent to the potential presence of Giardia cysts in the raw water and to the potential for cyst carryover into the filtered water. The watershed is attractive and heav- ily used for outdoor recreation. The habitat throughout the watershed is suitable for a wide range of wildlife. Near the center of the watershed, the creek passes through a 100+ -acre marsh where there is evidence of beaver. The plant does not have a full-time operator but receives careful attention on a daily basis. Three features of the plant's design and operation were iden- tified as contributing to potential cyst passage: (1) The on and off cycling of filter operation imposes hydraulic shocks from 8 to 12 times per day or 50 to 75 times between backwashings; (2) the filters are completely enclosed, prevent- ing routine observation of media condi- tion, particularly during backwashing; and (3) the physical design provides neither flash mixing nor coagulation, thus drastically limiting flexibility for chemical conditioning. Upon inspection of the medium in one filter through an access hatch, it appeared to be clean and in good condition, both on and below the surface. Community C. Physical characteristics of plant. Community C was a ski resort about 50 mi east of Tacoma, Wash. Peak water demand, on winter weekends, was about 85,000 gpd (60 gpm). Winter weekday and summer usage was <1,000 gpd. The raw -water sources were two small creeks that drain adjacent steep mountainsides. Community C had a 0.085-mgd (60- gpm) DE filtration plant. Principal com- ponents of the plant include a flowmeter, precoat-body feed tank, pressure filter, turbidimeter, automatic control system, and associated pumps, valves, and piping (Figure 11). The filter had a surface area of 60 sq ft and operated at a nominal rate of 1 gpm/sq ft. Filtered water flowed into two 50,000 -gal steel tanks. Plant operation. Water production was controlled by level sensors in the finished -water tanks. When the tanks were full, a 10-gpm recirculation pump was started and water was recirculated through the filter to prevent loss of filter cake from the filter leaves. The end of a filter cycle was determined by the head loss across the diatomite cake. Cycles were as short as a day during peak - demand or high -turbidity periods but could be as long as two or more weeks when demand was low. Cleaning re- quired plant shutdown and removal of the filter cake by automatic and manual operations. Visual inspection of the septum to ascertain completeness of cake release was not practical because of the configuration of the filter (Figure 11) and the confined stacked septum ar- rangement. A new cycle was initiated by pumping a DE slurry through the filter to deposit a precoat cake of about 0.2 lb/sq ft of filter area. 92 RESEARCH AND TECHNOLOGY JOURNAL AWWA After the precoat was applied, normal )peration began. The precoat-body feed tank was refilled, and DE for body feed was added. Flow to the filter was started along with the body feed pump to provide 20 mg/L of diatomite in the raw water. The filter operated continuously until the filter cycle was terminated for clean- ing. A brief (about 1-2 s) pressure transi- tion occurred when the operating mode changed from normal filtration to recir- culation and again when normal fil- tration was reinitiated. Performance of full-scale and pilot plants. Field investigations were con- ducted from May 20 to 24 and from July 29 to Aug. 2,1985. The field investigation was planned for a single week. The second week was scheduled after an interruption in the supply line made it impossible to operate the main plant at design capacity. Limited observations of the full-scale plant and of the pilot filter, operated in parallel with the full-scale plant, were made on May 23 and 24. Raw -water turbidity varied from 1.3 to 2.8 ntu, and the filtered -water turbidity of the full- scale plant varied from 0.33 to 0.5 ntu. Both decreased between successive mea- surements made on May 23 and 24. No Giardia cysts were found in a single 11 - gal raw -water sample or in two 50 -gal filtered -water samples taken in this period from the full-scale plant. Using 0.2 lb/sq ft of precoat* and 20 mg/L of body feed, the pilot filter was operated at a flow rate of 1 gpm/sq ft for 24 h. Cyst addition was then initiated at a rate of 6,800/gal. During the first 6 h, the turbidity removal increased from 60 to 75 percent (Figure 12). In three suc- cessive samples of the pilot filter effluent, concentrations of 14,18, and 23 cysts/gal were found. These concentrations cor- respond to about 99.7 percent cyst re- moval. Observations were terminated at this point because the raw -water supply was interrupted. The full-scale plant and the pilot filter were monitored again between July 29 and August 2. During this period water use was at a very low level, so the full- scale plant was operating in recirculating mode for all but brief periods. The raw - water turbidity, which was recorded continuously between July 31 and August 2, varied between 0.1 and 0.2 ntu except for peaks of 0.6-0.7 ntu corresponding to periods of rainfall. During two brief periods of normal filtration by the full- scale plant, raw -water and filtered -water turbidity were, respectively, 0.22 and 0.06 ntu or 73 percent removal, and 0.12 and 0.05 ntu or 58 percent removal. During the latter period, a 90 -gal sample of raw water and 120 gal of the full-scale plant filtered water were processed for Giardia cysts. Cysts were present in both samples, corresponding to a con- centration of 0.13 cyst/gal in the raw water and 0.02 cyst/gal in the filtered water. Filtered -water turbidity observed dur- ing operation of the pilot filter between July 31 and August 2 was typically 0.03-0.06 ntu. Higher turbidity was observed in the filtered water during the initial period of filtration following completion of the precoat. The concentra- tion of cysts in filtered water from the pilot unit appeared higher during the initial filtration period. The concentration of cyst -size particles in the raw -water samples ranged from about 260 to 1,600/mL. Corresponding filtered -water particle concentrations ranged from about 180 to 550/mL. Based on the limited observations made, the concentration of cyst -size particles ap- pear closely related to turbidity. The removal efficiencies of turbidity and cyst -size particles were more closely related than was observed during the study at community A. Sanitarysurvey. The sanitary survey of the watersheds and treatment plant indicated conditions conducive to the presence of Giardia cysts in the raw water and the potential for passage of cysts or cyst -size particles into the fin- ished water. The watersheds, two small basins of about 100 and 300 acres, are alpine in character, partially forested with fir and hemlock, with a partially grassy mead- ow. The habitat is suited principally for small mammals and rodents. Deer and elk may be present in midsummer months. The watersheds are readily accessible to humans in both summer and winter. Two features of the plant may have contributed to cyst passage. First, the plant did not use a filter -to -waste cycle. This presented a limited opportunity for cyst passage in the initial volume of water in the treatment system after a new filtration cycle was initiated. The second potential for cyst passage was associated with the transitions between normal filtration mode and low -demand recirculation mode. Discussion Giardia cyst removal. Giardia cyst re- moval efficiencies for the two granular media filtration plants were calculated to be between 70 and 90 percent. A reduction in cyst concentration of ap- proximately one log is significantly less than reported by others. Previous pilot studies using waters of differing char- acteristics with a range of coagulant dosages, mixing intensities, and floccu- lation conditions have achieved cyst removal efficiencies between 99 percent (two logs) and 99.9 percent (three logs).4.7.17These studies have shown that with in-line filtration, direct filtration, and conventional treatment, coagulant selection and dosage can affect cyst removal by one log or more. The cor- respondence of cyst removal efficiencies observed in this study between full-scale and pilot filters, and the significantly higher cyst removal efficiencies observed in pilot filters operated with optimized chemical conditioning were similar to the observations of others. Thus, it is confirmed that effective chemical condi- tioning increases cyst removal efficiency by at least one order of magnitude. It is also clear from observations at the community A and B plants that filtered - water turbidity is a useful parameter for estimating cyst removal only if water is properly chemically conditioned. Al- though the filtered -water turbidity at both plants was consistently 0.1 ntu or less, Giardia cysts were found in more than half of the filtered -water samples. This is consistent with the observations of Al -Ani et al,"' with low -turbidity water like that in this study, indicating that cyst removal efficiency depends on tur- bidity reduction. The irrelevance of filtered -water tur- bidity by itself as a parameter useful in controlling filtration for removal of Giardia cysts is not understood by some water treatment operators and regula- tory agencies. The relatively low turbid- ity of filtered waterin community A was the initial basis for not using chemical coagulation. It was also the basis for not exploring other potentially more effec- tive chemical combinations and seasonal adjustments at community B. Clearly, optimization of chemical conditioning, with seasonal adjustments correspond- ing to changes in water quality and temperature, is an essential feature of making effective use of the capital invested in water treatment hardware. Filter operating conditions. The variation in and control of filtration rate are major aspects of filter operation, with the potential for a significant impact on Giardia cyst removal. The adverse effects of variations in filtration rate on filter performance are well documented.12 Increased passage of Giardia cysts through filters in response to changes in filtration rate has been documented in previous pilot studies.4.10 In this study, data on cyst concentration and turbidity in the filter effluent were examined to assess the potential effects of changes in filtration rate, which were significant features of the operation at each of the treatment plants. Two major factors contributed to flow rate changes at the community A plant: (1) the on-off cycles (5 to 8 per day or 30 to 50 between backwashing) used to match plant production to reservoir capacity and system demand, and (2) filtration rate control by a throttling valve on the filter effluent line regulated by a float on the water surface above the "Hyflo Super -Cel. JM Corp.. Denver. Colts. JUNE 19% JERRY E. ONGERTH 93 filter. During filtration, the head on the filter fluctuated (cycled continuously) over a 2- to 3 -in. range within a period of about 10 min. At community B, three factors con- tributed to flow rate changes: (1) the on-off cycles (8 to 12 per day or 50 to 75 between backwashing) used to match plant production to system demand and storage capacity, (2) declining filtration rate with no automatic filtration rate control, and (3) variations caused by the conditions of discharge from the filters to the finished -water reservoir. Dis- charge from the filters is restricted, depending on the water surface elevation in the reservoir. The effect is significant, reducing filter output into a nearly full reservoir by more than 20 percent. These changes are slow, however, occurring over a period of hours and are much less detrimental than rapid changes. At community C, three factors con- tributed to changes in filtration rate: (1) alternation between filter production and recycle modes, required to match pro- duction to storage and demand and to retain the filter cake on the septum; (2) declining -rate, constant -head filtration in production mode, and (3) discharge against the variable water surface eleva- tion of the finished -water reservoir. An examination of the filtered -water turbidity record for each plant showed that during the initial period of a filter production cycle between backwashes (i.e., the initial period of filtration follow- ing restarting without backwashing), only slight changes resulted. Giardia cyst analyses of samples corresponding to filter restart periods were positive in most cases, but cyst concentrations were not distinguishable from those in other filtration periods. The lack of clear indi- cations of adverse effects on performance of the unquestionably troublesome op- erating conditions may reflect the gen- erally poor to mediocre performance. Previous work clearly shows the adverse effect of flow rate changes on the release of accumulated particles from a filter. Cleasby et al lz have demonstrated that the balance between forces tending to retain particles in a filter and those tending to dislodge particles is delicate and easily tipped in favor of the dislodg- ing, contributing to particle migration through the filter. Further, they de- scribed the magnitude of the effects resulting from changes in filtration rate as directly proportional to the magnitude of the change and its abruptness. In this study, the limited evidence of adverse performance resulting from severe changes in flow rate can most likely be attributed to the low turbidity of the raw waters and to the relatively small amount of solids accumulated in the filters. The consistent appearance of Giardia cysts in the filtered water can be interpreted, at least in part, as evidence teQAMV rslre 4—Body food pump S—ptmal tNd pump 6—Reotrculatlon pump 7—Alr rent 6 -Filter vasal 9—Finished-water tim t0—Dtstrlbutlon reservoir Figure ii. Schematic diagram and hydraulic profile of community C's full-scale treatment plant of a flow rate change effect, and can possibly also be attributed to poor or no coagulation. In this study, the removal efficiency of cyst -size particles was relatively low compared with that re- ported by others.4•10 This may also be due, at least in part, to flow change conditions imposed on each of the plants in this study, as well as to the water - conditioning practices. Filter backwashing. Effects of filter backwash practices on treatment per- formance are apparent in data on treat- ment performance (turbidity, particle counts, and cyst concentrations) and in visual observations. (This discussion does not apply to the DE filter at com- munity C.) Conditions for effective cleaning of the filter media through backwashing were least favorable at community A, where backwash was carried out on a preset time cycle, oc- curring strictly as a function of head loss and without the operator present for observation. The author observed that the backwash water was still visibly turbid at the end of the pretimed back- wash cycle. This is illustrated by the record of turbidity in filtered water immediately following backwash (Figure 3). Not only was the turbidity elevated in this period, but the break-in period was unusually long, undoubtedly due in part to the lack of chemical conditioning. The absence of more severe effects from poor backwashing on media condition at community A is likely due to the lack of chemical conditioning and to the low turbidity of the raw water. Without conditioning chemicals, particle accumu- lation in the filter depends entirely on natural interparticle forces and biological activity. The filter bed can, in a way, be seen as an extension of the stream bed. Inspection of the filter media at commu- nity A, in fact, revealed attached aquatic organisms more characteristic of slow sand filter conditions. Observations at community B indi- cated more normal operating conditions. The media had a clean, uncoated ap- pearance. Backwashing was conducted with an operator present to observe the clarity of the spent backwash water as it was discharged into the disposal pond. The condition of the media, however, cannot be monitored routinely because of the enclosed pressure -vessel filters. Filter -to -waste practice. In this study, only community B practiced filter to waste and only for 5 min. As data from the pilot filter show, this is far too short. Because Giardia cysts are resistant to disinfection, filtering to waste is par- ticularly important to maximize cyst removal by filtration. Turbidity records for filtered water at the community A and B plants and from the pilot filter at community C show that the period of elevated turbidity extends over the first hour of filtration following backwash. Concentrations of Giardia cysts found in filtered water at each plant were highest during the initial 15 min of filtration following backwash. Elevated cyst concentrations in samples from the unconditioned or less than optimally conditioned full-scale or pilot filter runs extended for as much as an hour after filter run initiation. In pilot - filter runs with optimal conditioning, the filter break-in period was reduced 94 RESEARCH AND TECHNOLOGY JOURNAL AWWA PM tun stat at 1:40 p.m.. lst reading at 1:45 pm. 1.5 Raw want-� .10,000 I � 1.0 � r I I 1.000 I I a i I Start Terminate is I � 0.5 ° '0 +meter turefdity I 100 I °\o 0 I 10 1200 1400 1500 1500 2000 2M Time of Day Figure 12. Turbidity and Giardia cyst removal by pilot diatomaceous earth filter in community C, May 24, 1985 significantly. In pilot -filter runs at community C, higher cyst concentrations were found in samples taken in the initial 15 min of filter operation than in later samples (Figure 12). Too few sam- ples were taken from the full-scale plant in community C to determine whether its characteristics might be similar. The observations presented here clearly indicate the potential for passage of Giardia cysts through filters of various types and operating characteristics dur- ing the initial period of filtration follow- ing backwash. The period of higher potential cyst passage can be as long as an hour in filters operating without chemical conditioning or with poorly optimized chemical conditioning. This emphasizes the need for a filter -to -waste period in which Giardia are likely to be present in the raw water. It also high- lights the importance of effective chem- ical conditioning if the filter -to -waste period is to be of a practical duration, e.g., 15 min as opposed to 1 h. Other design considerations. The filter media design, the raw -water turbidity, and the chemical conditioning scheme are important factors in determining the particle storage capacity of a filter, the turbidity reduction capability, the rate of head loss accumulation, character- istics of the filter relative to turbidity breakthrough, and terminal head loss. The community A plant had 18 in. of 0.9 -mm coal over 16 in. of sand, including 10 in. of 0.5 -mm, 3 in. of 1 -mm, and 3 in. of 2 -mm sand. The community B plant had 9 in. of 0.7 -mm coal over 12 in. of 0.5 -mm sand. In previous work, Cleasby et all!' observed that in comparison with filters used by Tate"" containing 20 in. of 1.1 -mm coal and 20 in. of 0.5 -mm sand, their pilot filters containing 14 in. of 1.4 - mm coal over 20 in. of 0.5 -mm sand exhibited one log removal fewer cyst - size particles than those of Tate 0.5 log versus 2.5 log removal). Studies of Giardia cyst removal in pilot -scale in-line filters, reported by Al - Anil" and by Mosher,t% were conducted using a media design consisting of 22 in. of 1.0 -mm coal over 10 in. of 0.5 -mm sand. This relatively conservative de- sign, in conjunction with effective chem- ical conditioning, contributed to the observed 2+ log cyst reduction efficiency corresponding to periods of effective (>90 percent) turbidity reduction. In compar- ison, the community B filter medium is relatively shallow although the coal size is smaller in partial compensation. Media design at communities A and B may have contributed to the observed per- formance characteristics. The number of major factors affecting performance at those plants, however, make it vir- tually impossible to determine the rela- tive importance of each. Other design factors at one or more of the plants that may have contributed to cyst passage into the filtered water include (1) a common wall separating raw (or spent backwash) and filtered water, and (2) the potential for contam- ination of the finished water in the clearwell by sanitary wastes or surface runoff. Furthermore, at the community C plant, the lack of redundancy in design and the possibility of power failure dur- ing periods of demand present the po- tential for direct bypassing of the plant. Other operating factors. At community A, the operations staff was one person— responsible not only for the plant but for all other aspects of water system opera- tions and maintenance, including the dam, reservoir, chlorination, pumps, wells, distribution system, and refuse collection. Time available for plant maintenance and operation was limited to about 30 min twice a day. The obvious conclusion is simply that it is not reasonable to expect a community of this size and character to provide the level of operator skill, training, and attention required to assure consistently effective performance of a conventional treat- ment plant. Accordingly, an alternative means of providing water of suitable quality should have been selected. This is the logical responsibility of the de- signer and of the regulatory agency (and the community). Giardia cysts in source water. Giardia cysts were found in the raw water at each treatment plant, with concentra- tions ranging from 0.13 to 0.3 cyst/L. These concentrations are comparable to cyst concentrations found in other simi- lar water sources elsewhere in the Cascade and Sierra Nevada mountains."' The three watersheds had distinctly different physical features but each was typical in its own way of broad areas of western mountain watersheds. The cyst concentrations found do not appear to be unique. The presence of cysts in Cali- fornia mountain watersheds has been reported by Riggs"i and Suk."" Cyst con- centrations are likely to be significantly lower than in streams below domestic wastewater discharges, which typically have relatively high concentrations of Giardia cysts.2324 In fact, cyst concen- trations reported for the Youghiogheny River in northern West Virginia and southwestern Pennsylvania, influenced by upstream secondary treatment plant effluents, ranged from 0.03 to 12/L, assuming 10 percent recovery. It appears prudent to expect Giardia cysts to be present in even relatively remote water sources. Accordingly, source -related factors can be identified that may be of significance to water treatability and to minimizing cyst concentrations in filtered water. Cysts originate in fecal material at high con- centrations, about 106/g.2, From the assumption that cysts may enter a stream in fresh feces, it follows that a small steep watershed with an unim- pounded in -stream diversion, as at community C, has the potential for con- veying water with the highest (though more likely short-term) cyst concentra- tions to the treatment plant. Impound- ments, increasing tributary area, and increasing stream flow rate would all be expected to favor lower and more uniform cyst concentrations. However, in some locations impoundments might provide JUNE 1990 JERRY E. ONGERTH 95 more habitats for beaver and muskrat and lead to more cysts. Conclusions • Giardia cyst contamination is likely to be appreciable even in relatively re- mote, high-quality raw -water sources. Accordingly, water treatment processes should be operated to provide effective cyst removal at all times. Special atten- tion must be given to chemical pretreat- ment for granular rapid sand filters. • The effectiveness of casually oper- ated filtration processes for removing Giardia cysts is relatively poor and is approximately equal to the turbidity reduction efficiency. Thus, when treat- ing raw water with naturally low tur- bidity, <1 ntu for example, simply meeting the often -used turbidity goal of 0.5 ntu will not ensure effective cyst removal. Operators, design engineers, and state regulatory agencies must understand and deal with this. Man- agement philosophy should be to produce drinking water with the lowest attain- able turbidity. • Giardia cyst removal by filtration of well -conditioned water resulting in 90 percent or better turbidity reduction will produce effective cyst removal of 99 percent or more. Accordingly, monitoring water -quality conditions and adjusting chemical conditioning to maintain op- timal coagulation are essential to con- sistent cyst control. 0 Where turbidity reduction efficiency is <90 percent, Giardia cyst concentra- tions are likely to be relatively high in the filtered water produced immediately following backwash. This condition may persist for up to an hour. In well -condi- tioned waters, the period will be shorter. Accordingly, an appropriate period of filtering to waste or other procedures such as preconditioning backwash water or gradually increasing filtration rates from zero to the normal operating value should be considered. • Changes in flow rate, particularly the practice of using on-off cycling between backwashings, should be avoided to reduce the potential for the passage of cysts into filtered water. • It is unrealistic to expect poorly trained part-time plant operators to be able to maintain the optimum chemical conditioning required to assure con- sistently effective Giardia cyst removal in small, relatively inflexible water treatment plants. • If small plants are unable to achieve performance levels described in this study to ensure cyst removal, super - chlorination with prolonged contact times or another equally effective disin- fection procedure may be necessary as a short-term measure. Ultimately, such plants need improved operation. • Diatomaceous earth plants must recirculate (during periods of nonpro- duction) to keep the media on the septum. Given the control problems facing small rapid sand plants (to match production to use) and the problem of dislodging accumulated particles if the filters are turned repeatedly off and on between backwashes, it might be advantageous to recirculate in these plants to maintain flow conditions. Acknowledgment This study was funded in part by the AWWARF with participation by the University of Washington, the California Department of Health Services, and the Seattle Water Department. The assis- tance of Henry J. Ongerth in preparation of the final manuscript was indis- pensable. Successful completion of the project would not have been possible without the skillful work of Joegir Engeset in operating the pilot plant and of Lynn Salter in preparation of the manuscript. The contributions of com- munities A, B, and C through their respective representatives, Jim Bogart, Harvey West Jr., and Greg Durban, made the project possible. References 1. KIRNER, J.C.; LITTLER, J.D.; & ANGELO, L.A. A Waterborne Outbreak of Giardiasis in Camas, Washington. Jour. AWWA, 70:1:35 Oan.1978). 2. LIPPY. E.C. Tracing Giardiasis Outbreak at Berlin, New Hampshire. Jour. A WWA, 70:9:512 (Sept. 1978). 3. NAVIN, T.R. ET AL. Case -Control Study of Waterborne Giardiasis in Reno, Nevada. AmcricanJorrrnal Epidemiolony, 122:2:269 (1985). 4. LOGSDON, G.S. ET Al.. Alternative Filtra- tion Methods for Removal of Giardia Cysts and Cyst Models. Jour. AWWA, 73:2:111 (Feb. 1981). 5. BFI.I.AMY, W.D. FT AI.. Removing Giardia Cysts With Slow Sand Filtration. Jour. AWWA, 77:2:52 (Feb. 1985). 6. CLEASBY, J.L.; HILMoI;, D.J.; & DIMITRA- coPoULos, CJ. Slow Sand and Direct In- line Filtration of a Surface Water. Jour. AWWA, 76:12:44 (Dec. 1984). 7. PYPF.R, G.R. Slow Sand Filter and Package Treatment Plant Evaluation Operating Costs and Removal of Bacteria, Giardia, and Trihalomethanes. Proj. Rept., Coop. Agreement CR809284010. Drinking Wa- ter Res. Lab., USEPA, Cincinnati, Ohio (1985). 8. DFWAu.F, F.B.; ENGESET.J-; & LAWRENCE. W. Removal of Giardia lamblia Cysts by Drinking Water Treatment Plants. Proj. Rept., Grant R806127. Drinking Water Res. Div., USEPA, Cincinnati, Ohio (1982). 9. LANGE, K.P. ET AL. Diatomaceous Earth Filtration of Giardia Cysts and Other Substances. Jour. AWWA, 78:1:76 (Jan. 1986) 10. ENGESET, J. & DEWALLE, F.B. Removal of Giardia lamblia Cysts by Flocculation and Filtration. Proc. 1979 AWWA Ann. Conf., AWWA, Denver, Colo. (1979). 11. Committee Report. The Status of Direct Filtration. Jour. AWWA, 72:7:405 (July 1980). 12. CLEASBY, J.L.; WILLIAMSON, M. BAUMANN, E.R. Effect of Rate ChangK---" Filtered -Water Quality. Jour. AWN, 55:7:880 (July 1963). 13. AMIRTHARAJAH, A. & WFTSTF.IN, D.P. Int. tial Degradation of Effluent Quality Dur-'. ing Filtration. Jour. AWWA, 72:9:518 ( Sept. 1980). 14. SFFLAUS. T.J.; HF.NDRICKs, D.W.; & JAN. ONIS, B.A. Design and Operation of a Slow Sand Filter. Jour. A WWA, 78:12:35 ( Dec. 1986). 15. LUcH•rEl., D.L.; LAWRENCE. W.P.; & DF- WAI.I.F, F.B. Electron Microscopy of Giardia lamblia Cysts. Appl. Envir. Microbio!., 40:4:821 (1980). 16. RIGGs, J.L. Use of Immunofluorescence for Detection of Giardia lamblia Cysts 00*centrated From Drinking Water. Final Proj. Rept. Contract 52-83. AWWARF, Denver. Colo. (1985). 17. MOSHER, R.R. & HENDRICKS, D.W. Rapid Rate Filtration of Low -Turbidity Water Using Field -Scale Pilot Filters. Jour. A WWA, 78:12:42 (Dec. 1986). 18. AL -ANI, M.Y. ET AL. Removing Giardia Cysts From Low -Turbidity Waters by Rapid Rate Filtration. Jour. AWWA, 78:5:66 (May 1986). 19. CLEASBY, J.L. ET AL. Effective Filtration Methods for Small Water Supplies. Coop. Agreement CR808837-01-0 Proj. Rept. Drinking Water Res. Div., USEPA, Cincinnati, Ohio (1984). 20. TATE, C.H.; LANG, J.S.; & HUTCHINSON, H.L. Pilot -Plant Tests of Direct Filtra- tion. Jour. AWWA, 69:7:379 (July 1972). 21. ONGERTH, J.E. Giardia Cyst Concentra- tions in River Water. Jour. AWWA, 81:9:81 (Sept. 1989). 22. SUK, TJ.; RIGGS, J.L; & NELSON, B.C. Water Contamination With Giardia in Back -Country Areas. Proc. Natl. Wil- derness Res. Conf.: Current Res. Forest Service, USDA, S. Tahoe, Calif. (1986). 23. Fox, C J. & FITZGERALD, P.R. Presence of Giardia lamblia Cysts in Sewage and Sewage Sludges From the Chicago Area. Waterborne Transmission of Giardiasis (W. Jakubowski and J.C. Hoff, editors). USEPA, Cincinnati, Ohio. (1979). 24. SYKORA, J.L. ET AL. Monitoring of Water and Wastewater for Giardia. Proc. 1986 WQTC, AWWA, Denver, Colo. 25. FAUBERT, G.M. ET AL. Comparative Studies on the Pattern of Infection With Giardia spp. in Mongolian Gerbils. Jour. Parasitol., 69:802 (1983). About the author: 4 Jerry E. Ongerth is an assistant professor in the Department of Environmental Health, University of Washington, Seattle, WA 98185. Ongerth is a graduate of the University of California, Berkeley (BS) and the University of Michigan, Ann Arbor (MS and PhD). His work has been published by Applied and Environmental Microbiology, Journal of Clinical Micro- biology, American Journal of Public Health, and Journal WPCF, as well as JOURNAL AWWA. 96 RESEARCH AND TECHNOLOGY JOURNAL AWWA W 00 o y 4 Q. �C 5 ,,� ,� 11 o „o aA� S O cr. � o C � (D E- o ►� o ?' � o It C 'i7 arc G� coo' o " aha ' CD CCL Fd ° o o a. o - 5 . on �- v,cq .4 c o a �• g �. o CD CD En o a"�e�nc R1,5.Ro 140041ob • CD 0 �' n y• 0 g r. �. X39 A A !D rr CD �. 0 CD 3TO�VO�0 r moo. 0, 0 CD cD oQ fDyo¢oara Cl. p� O O 0 o, 0 9- S O C v,uq o�� ° 7 ° Grp! rt fD Y��+ r.. a'•`C 0 Cr a O a. 4 C to r7,�� ti bd 'R' , � CL o w ,.,, O • M, O ff. o a. E3 9 o .0 cD� CD cfD o o` ►°,marc°°,��a ��c��o7.���� o C 8 O �, o x�' p fD CD 0ire 101, lot .0 Cv',O f� cD 0 A n C a.N CD n Ern :7•m "t W ' ° a CD ° N 'U C °� p. O p- m p m O � n U9 C. 'o,-- ¢, cn C of crop� o,°a•�•�yo �DQ�. �o 40 -.0 W' o✓ ° � Oa �D � o cr. � o x�0 (7no�oCa< n! CITY OF SEBASTIAN UTILITIES DEPARTMENT DATE: March 22, 1993 TIME: 4:45 P.M. CONFERENCE WITH: [ X ] TELEPHONE CONVERSATION WITH: Joe McNamara of the DER SUBJECT: Trihalomethanes (THM's) RESUME OF CONVERSATION: Robb McClary and I called Joe from Robb's Office. Robb explained that Mayor Powell had spoken to Mr. McNamara. He had informed the Mayor that the City of Sebastian could impose stricter requirements on its own water treatment system in order to implement THM control. Robb told Joe that GDU is still operating their system in Sebastian. Joe said that Florida State Law images the Federal Law unless the State wants to make the Florida law more stringent. Joe McNamara said that he thought the counties and the cities in Florida could enact an ordinance that was more stringent than the DER rules, especially if they owned the treatment plants. Robb told Joe that the existing franchise agreement the City has with General Development Utilities states that the water quality and quantity should be in accordance with DER Rules and Regulations. If the City now passed a new ordinance requiring GDU to install a THM control system at their plant, the City would violate its own franchise agreement. Consequently, the City would most likely end up in court being sued by GDU. Joe McNamara said that he hoped he didn't give any false hope to Mayor Powell by suggesting that the City could impose more rigorous standards in regard to the City-GDU situation. Robb asked Joe about the logic behind the fact that EPA does not require water systems having 10,000 or fewer customers to institute THM control. Joe said that there may not be any logic to it. He said that the EPA standard for MCL purposes is based on a person drinking 2 liters of water per day for 70 years. He said that we could write to the Secretary of the DER. He would ask for our help in requesting THM control for small systems. We could also call the American Cancer Society, AWWA, and National Health Institute. ORIGINATED BY: //ZX444//j,- 4z..p V i'; I.'1n4r—Dlrc,dor COPY TO ILLkN G'' ASSOCIATIES, INC. engineers, hydrogeologists, surveyors & management consultants April 19, 1993 Mr. Richard Votapka, P.E. Utilities Director City of Sebastian P.O. Box 780127 Sebastian, Florida 32978-0127 Subject: Public Notice for Ammoniation System Dear Mr. Rich: HAI #92-023.06 aC- � •;eg3 �tC5 `��, Attached is a copy of the General Development Utilities, Inc. (GDU) Public Notice, which was folded and put into the bills as a bill stuffer to all the customers of the GDU system. Also, this notice was printed in the newspaper for local distribution twice prior to the conversion to the ammoniation system. The first time it was put into the newspaper was one month prior to conversion, and the second was two weeks prior to conversion. The notice was also published in the Palm Bayer, which is the City's newspaper. The City cooperated with GDU at the time in disseminating this notice. In an effort to have as many things accomplished in a rapid manner, I am providing you a copy of the notice such that you can rework it to your satisfaction and coordinate the public notification activities which would be required in conversion from a free to a combined chlorine residual. Thank you for your time and consideration of the above matter. Very truly yours, Hartman & Associates, Inc. ieral C. an, P.E. President GCH/ch C20/Votapka.gch Attachment cc: Hal Schmidt, HAI 201 EAST PINE STREET - SUITE 1000.ORLMXDO. FL 32801 TELEPHONE (407) 839-3955 - FAX (407) 839-3790 PRINCIPALS: JAMES E. CHRISTOPHER • CHARLES W. DRAKE • GERALD C. HARTMAN - MARK 1. L[:KE -.MARK A. RYNNING • HAROLD E. SCH,MIDT. JR. a PUBLIC NOTICE TO. OWNERS OF TROPICAL FISH AND OPERATORS OF MEDICAL DIALYSIS FACILITIES: General Development Utilities is advising owners of tropical fish and operators of med- ,ical dialysis facilities and other chlorine -sensitive processes that current procedures to test 'for and remove chlorine from the Port Malabar water supply system may no longer be effective. In early March, 1983 General Development Utilities will modify its water treatment process in order to comply with new state and federal regulations. The modification involves changing disinfectants from chlorine to chloramine. As used in public water supplies, chloramine is, like chlorine, harmless to humans but can adversely affect some aquatic species and certain chlorine -sensitive processes such as dialysis. Problems arise because test procedures used to detect and measure chlorine may -not indicate the presence of chloramine. Also, the technique of "aging" water to dissipate chlorine before use in aquariums is not advisable, because chloramines are more persistent than chlorine. Water disinfected with chloramine can be used safely for all applications if certain Precautions are followed. General Development Utilities urges its customers who may be affected by this change to seek advice from professional aquarists or medical authorities competent to give such advice. On the positive side, chloramines will provide longer lasting protection against water- borne disease organisms without the strong taste and odor of chlorine. Further, chloramine will not produce chloroform and other by-products which are potential human carcinogens. Please watch your local daily newspaper for the date in which the change to chloramines will occur, or call our local office at 723.2877 if you have any questions. Feb., 1983 General Development ,y litilrtim Inc. Customer Relations Department Pbr<- `i"i m s/� CoKt7P_oL Palm Bay Utility Corporation Booster Pump Station Ammoniation System Addition Equipment List Provided by Blankenship and Associates, Inc. 1. Capital Controls Company, Inc. Flow Proportioning Ammonia Gas Pressure Feeder Model # 4611A — 50 PPD gas flowmeter capacity — 4 — 20 mAdc controller signal — check valves coupled to 1 1/2—inch corporation cock with diffuser with rubber sleeve — IPS threaded corporation cock — 2— cylinder manifold — 120 Vac, 60 Hz, Single Phase power — interconnecting piping, manual isolation valves, and appurtenances from cylinder manifold to cabinet — interconnecting piping with manual isolation (ball) valve from cabinet to union outside of enclosure 2. Blankenship and Associates, Inc. Equipment Enclosure — model BA— 505 with standard options and/or the following (all power items should be wired to breaker box ready to use) — 10—inch Dayton Model 2C819 exhaust fan, mounted and wired to breaker and switch — safety chains and mounting hardware for 4 ammonia cylinders (mounted) — automatic thermostat, 30-110 degrees, switched — 1000—Watt heater unit — dual electronic scale for ammonia cylinders, mounted and ready to use — ammonia detection and alarm system with audible alarm and extenal light, mounted and wired to breaker, ready to use Note : Switches and breakers shall be located outside enclosure and shall be rated NEMA 4X. Provided by Other Contractors or PBUC 3. 84" x 84" concrete pad for enclosure, (see plans) — holes drilled for enclosure—mounting with internal flange and expansion anchors 4. Concrete sidewalk to enclosure for transport of ammonia cylinders (see plans) 5. Interconnecting piping (1—inch PVC) and appurtances from union at outside of enclosure to corporation cock at feed point 6. Electrical conduit and wiring from power source to breaker box in enclosure Bulletin A1.14610A.2 SERIES 4610A G AMMONIA GAS PRESSURE FEEDER, CAPITAL CONTROLS FLOW PROPORTIONING TYPE COMPANY, INC Capital Controls Series 4610A represents state of the art In automatic proportioning control of anhydrous ammonia gas feeding for water/ wastewater treatment. The modular dispensing system provides the user with remarkable flexibility in configuring the components for maximum operator convenience and operating safety. Four feed capacities are available for feeding anhydrous ammonia up to 250 PPD (5 kg/h) Into open vessels or pipelines against pressure up to 10 psig (0.7 bar). The Series 4610A system has been human engineered for ease of operation, observation, control, and service. Basic Feed System The dispensing system is comprised of four gas regulating, metering, control, and application components. The anhydrous ammonia gas pressure reducing valve with gas flow rate in- dicator is located at the gas source. It may be direct cylinder mounted, wall mounted on the gas valve of a single or multiple cylinder manifold, or on the gas discharge valve of a bulk storage tank. Ammonia gas at source pressure enters through the inlet valve and filter assembly. Gas pressure is reduced and controlled to approximately 20 psig (1.4 bar). Low pressure gas is con- veyed from the pressure reducer to the floor cabinet mounted metering and rate controls. In the floor cabinet the ammonia gas flow is metered and its flow rate automatically controlled by a linear modulating valve. The automatic valve features local manual valve ad- juster. The floor cabinet instrument panel houses the ammonia gas pres- sure gauge, dosage adjustment, and gas flow indicator. The position of the automatic gas flow control valve In the floor cabinet is controlled locally with a manual rate ad- juster, or automatically from the controller which is mounted where most convenient for operator observation and system safety. The electronic controller receives the flow signal and adjusts valve posi- tion. Lights illuminate to show valve operation, power, and no flow conditions. From the floor cabinet, the low pressure gas is conveyed to the check valve and gas diffuser assembly located at the applica- tion point. The gas is diffused into the water or wastewater through an expandable rubber sleeve which automatically removes precipitate that forms on the diffuser In water containing hardness. A spring loaded diaphragm check valve prevents backflow of water. Flow Diagram General Specifications Model Information Code MODEL o a a ❑ ❑ Floor Cabinet Anhydrous Ammonia Gas Pressure Feeder L Capacity Flow ProportioningControl 1 - 50 PPD (1 kglh) 2 -100 PPD (2 kg/h) 3 - 250 PPD (4.8 kglh) FEED CAPACITIES *Rigid pipe to be Schedule 80 PVC threaded or solvent weld. CHECK_ VALVE DIFFUSER Model Open Channel or Maximum Capacity Flowmeter Capacity Model Number PPD (Metric) PPD (metric) 4611A 50 PPD (1 kg/h) 50 PPD (1 kg/h) diffuser with rubber sleeve. tubing or 114" thread. 20 PPD (0.4 kglh) 4612A 100 PPD (2 kg/h) 100 PPD 4613A Gasdrtlet connection for one (1) diffuser with rubber PPO (1 kg/h) sleeve. pipe. 20 PPD (0.4 kg/h) 4613A 250 PPD (5 kglh) 250 PPO (4.8 kg/h) a Gas Pressure Regulator with 100 PPD (2 kglh) 1. Yoke Assembly 6. Manual Exhaust Valve 50 PPD 0 kglh) Accuracy—within f 4% of maximum flow capacity TUBING CONNECTIONS 8.25" (8 m) 318" Plastic Pressure Model Number Maximum Capacity Pressure Vent' 4611A 50 PPD (1 kglh) 318" tubing WIN 4612A 100 PPD (2 kglh) 112" pipe WIN 4613A 250 PPD (4.8 kglh) 112" pipe 318" Went from regulator (1) and exhaust valve (1). PRESSURE LINE SIZE.REQUIREMENTS Type of Pressure Maximum Diameter of Pressure Line Line Feed Rate 100'(30 m) 2001(61m) 500'(152m) Material 50 PPD (1 kglh) 318" 318" 112" Tubing or Rigid Pipe 100 PPD (2 kglh) 112" 112" 314" Rigid Pipe 250 PPD (4.8 kg1h) 112" 314" 1 " Rigid Pipe *Rigid pipe to be Schedule 80 PVC threaded or solvent weld. CHECK_ VALVE DIFFUSER Model Open Channel or Number Tank Type Pipeline Type 4611A Check valve and diffuser Check valve coupled to with rubber sleeves. Gas 1112 ff corporation cock with Inlet connection for 318" diffuser with rubber sleeve. tubing or 114" thread. 4612A Check valve and diffuser Check valve coupled to and each with rubber sleeves. 1112 " corporation cock with 4613A Gasdrtlet connection for one (1) diffuser with rubber 112" thread or solvent weld sleeve. pipe. Shipping Weight -200 lbs (90 kg) Standard Equipment" (AIL Models) a Gas Pressure Regulator with 1. Yoke Assembly 6. Manual Exhaust Valve 2. Gas Supply Indicator 7.25" (8 m) 318" Vent Tubing 3. Pressure Relief Valve 8.25" (8 m) 318" Plastic Pressure 4. Gas Filter Tubing (Model 4611A Only) 5. Vent and Pressure 9. 12 Lead Gaskets Connections* 10. One (1) Set Spare Parts 'Models 4612A and 4613A provide, in addition to threaded gas pressure con- nections, spare adaptors for solvent weld gas pressure piping connection. EUROPEAN HEADQUARTERS Crown Quay Lane • Sittingbourne • Kent ME 10 3JG • U.K. Tel: 0795.76241 • Telex: 965636 CAPCO G IN BELGIUM Rue de Mont 31 • B-6330 Sombreffe • Belgium Tel: 071-889122 • Telex: 51601 CAPCO B b. Remote Controller c. Floor Cabinet with 1. Automatic Valve with Local Manual Valve Adjuster 2. Metering and Rate Controls 3. Gas Pressure Gauge d. Check Valve with Expandable Rubber Sleeve Diffuser Assembly{Patented) Electrical Requirements a. 120 Vac, 60 Hz, Single Phase c. 100 Vac, 50 Hz, Single Phase b. 240 Vac, 50 Hz, Single Phase d. Power Consumption: 50 watts Wiring—power and signal wires broght to remote mounted controller. Signals to Controller a. 4-20 mAdc c. 1050 mAdc e. 0-5 Vdc b. 1.5 Vdc d. 0.20 mAdc Options a Anhydrous ammonia cylinder manifold (recommended with Model 4611A, required with Models 4612A and 4613A unless feeder is mounted on an ammonia storage tank vapor discharge valve) Dimensions t sEXHAUST �p {---� VALVE EXHAUST VALVE ON TEE w Wo Le 1 wiwaw� u" aerx wwt AiiLlaT _ io�oaaows IORK0010UlT `� _ MALLOKAIIQ � 1q rouNsuo sw�o�r � e 1plMliLMJ1IRN0.A0I1ldir� 1�R1�Ot1�® ��— O AMMONIA OAK a/fYKe.OrKN tNANML Oe TANK T Kraoa,enw oeoiauarwras / —71 as i w � I A,Ko,nar w oow"ws.aw TwK as At»Ow�Mp/111»�•MtY/ »1T[11P. tleO1KAwevti ap"sww.saee TANK Ordering Please specify: (1) model number, (2) das flowmeter capacity, (3) controller signal, (4) check valve and diffuser assembly type, (5) desired options. Design Improvements may be made without notice. Ci CAPITAL CONTROLS COMPANY INC P.O. Box 211. Colmar, PA 18915 U.S.A. Tel: 800.523.2553, in PA 800-242-7590 Outside U.S.A. 215.822-2901 Telex 6851074 CAPCO UW FAX: 215-82248640 Pub. No. 1284-2 8188 4M Copyright 1988 Capital Controls Company, Inc. S Bulletin A1.24610A.0 SPECIFICATION G ADVANCE SERIES 4610A CAPITAL CONTROLS AMMONIA GAS PRESSURE FEEDER, COMPANY, INC FLOW PROPORTIONING CONTROL Po. Box 21 Colmar. PA 18915 1.0 SCOPE This specification describes the ADVANCE Series 4610A Ammonia Gas Feeder as manufactured by Capital Control Company, Inc., 3000 Advance Lane, Colmar, PA 18915. 2.0 DESCRIPTION The gas ammoniator shall be Model (4611A) (4612A) (4613A) gas pressure operated, flow proportioning control. The ammoniator shall have a maximum capacity of (50 PPD) (100 PPD) (250 PPD) of anhydrous ammonia gas feed equipped with an anhydrous ammonia gas flowmeter of PPD. 3.0 DESIGN The ammoniator design shall provide for conveying the gas under a reduced pressure, from the ammonia supply valve to the diffuser check valve assembly. The ammoniator shall be constructed of materials specially suited for wet and dry ammonia gas service. All springs used in the ammoniator shall be stainless steel. 4.0 COMPONENTS The ammonia dispensing system shall be comprised of five (5) gas regulating, metering, control and application components: an ammonia gas pressure regulator with integral ammonia valve yoke and gas flow indicator, manual exhaust valve, floor cabinet gas flow controls, electronic controller, and gas diffuser with check valve. 4.1 PRESSURE REGULATOR The pressure regulator reduces anhydrous ammonia gas from source pressure to approximately 20 psig (1.4 bar). This spring -opposed regulator shall be factory set and shall not require any field adjustment for operation. Excess pressure will be prevented from building up in the system by means of a spring loaded, integrally mounted, diaphragm actuated pressure relief valve. The excess pressure shall be vented to the outside. 4.2 MANUAL EXHAUST VALVE A manual exhaust valve shall be provided for mounting in the ammonia gas line between the regulator and the floor cabinet. The valve shall be manually opened during gas supply replenishment to vent gas pressure to a safe location. 4.3 FLOOR CABINET GAS FLOW CONTROLS The floor cabinet shall house the automatic gas flow control valve with integral manual adjustment knob for control of gas feed when not in automatic operation, differential pressure regulator, gas flow indicator to show gas feed rate, and an ammonia gas pressure gauge. These components shall be housed in a floor mounted fiberglass cabinet having dimensions of 64" (1625 mm) high, 27" (685 mm) wide, and 19-3/4" (500 mm) deep. The gas flow indicator and gas pressure gauge is mounted in the 24" (610 mm) x 17" (430 mm) cabinet window. Dual front panels (top bezel and bottom access cover) \ - shall be removable without the use of tools for ease of access to all cabinet internals for servicing. The cabinet shall include rigid side frames, dual front panels, total rear access, and a ventilated kick plate to facilitate air circulation. All utility inlets and outlets shall be bulkhead connections mounted in the rear. Bulletin A1.24610A.0 Page 2 4.4 ELECTRONIC CONTROLLER A solid-state electronic controller suitable for wall mounting at a location remote from the floor cabinet shall be provided to control valve position. The controller shall contain a calibrated dosage potentiometer located in the cabinet next to the flowmeter enabling a 10:1 control range. The adjustable potentiometer will provide 5:1 turn -down ratio and 2:1 turn -up ratio. An internal selector switch shall provide for a fixed 1:1 ratio to replace the adjustable potentiometer as required. Field modification of the input signal is achieved with plug-in components. The controller shall be remotely mounted from the gas feeder and shall be customer wired to the floor cabinet mounted gas flow control valve, and the controller shall provide: 1. Customer power connection terminals. 2. Test switches to simulate flow signal for monitoring the controller and valve action. 3. DPDT alarm contacts for: a. No flow condition (select either flow signal or valve closed). b. Contact ratings 10A-120 Vac or 28 Vdc; 2A-240 Vac. 4. Lights will illuminate to show valve operation, power, and no flow conditions. Signals to controller: 1. 4-20 mAdc 2. 1-5 Vdc 3. 10-50 mAdc 4. 0-20 mAdc 5. 0-5 Vdc 4.5 GAS DIFFUSER WITH CHECK VALVE The diffuser -check valve assembly shall consist of a spring loaded check valve to prevent water from backing up into the ammonia feed system. The diffuser disperses fine gas bubbles into the water being treated. The maximum back pressure at the point of application shall be 10 psig (0.7 bar). (For open channel) the diffuser shall be a slitted, expandable rubber sleeve type. (For pipeline addition) the diffuser shall be a slitted, expandable rubber sleeve type close coupled with a corporation cock assembly having (Mueller thread) (IPS thread). 5.0 STANDARD EQUIPMENT The following standard equipment shall be furnished with each ammoniator: 1. 25 feet (8 m) of plastic pressure (Model 4611A only) and 25 feet (8 m) of 3/8" vent tubing. 2. Twelve (12) lead gaskets. 3. One (1) set of spare parts. CITY OF SEBASTIAN UTILITIES DEPARTMENT DATE: March 16, 1993 TIME: 4:30 PM [ ] CONFERENCE WITH: [ X z] TELEPHONE CONVERSATION WITH: EPA Drinking Water Hotline SUBJECT: Trihalarethanes RESUME OF CONVERSATION: I called the EPA Drinking Water Hotline (1-800-426-4791). I asked if EPA had any practical consumption comparison as to equating how much water would an individual have to drink with a certain level of THH' s in it to cause cancer. I was told that only the numerical thresholds are the acceptable guidelines now. EPA does not have any comparison to the 100 UG/L (micrograms per liter) for total trihalomethanes. I was also told to call back in the beginning of June, 1993. By then the proposed rule changes for THH's will be established by EPA. ORIGINATED BY: CITY OF SEBASTIAN UTILITIES DEPARTMENT DATE: March 26, 1993 TIME: 2:30 PM [ X ] CONFERENCE WITH: Gerry Hartman and Hal Schmidt of Hartman & Associates [ ] TELEPHONE CONVERSATION WITH: SUBJECT: Ammoniation System for THM Control RESUME OF CONVERSATION: I discussed a typical "Ammoniation" system for the purpose of installing one at the GDU Water Plant once the City takes over. Gerry and Hal said that basically the system is similar to a chlorine system. It is a "dry" system; no liquids are involved. Ammonia gas is installed in a large horizontal tank which is normally rented from the supplier for a nominal fee per year. Black steel pipe is used from the tank to the water system piping. Rotameters (regulators) are used to control the flow of gas from the tank to the system. The gas can cause calcification in the orifice of the injection tee so proper selection of materials is essential. The ammonia will react with "humics" (precursors of THM's) found in the raw water supply and the chlorine used for disinfection. A "Monochloramine" free residual will remain. No THM's are associated with this residual. ORIGINATEDBYa ///24 COPY TO: City of Sebastian POST OFFICE BOX 780127 o SEBASTIAN, FLORIDA 32978 March 15, 1993 TELEPHONE (407) 589-5330 o FAX (407) 589-5570 Mr. Glenn Schuessler Assistant Director Environmental Health Department HRS -Indian River County Public Health Unit 1900 27th Street Vero Beach, Florida 32960 RE: Trihalomethanes at General Development Utilities Water Treatment System, F1. DER No. 33111369 Sebastian Highlands Subdivision, Sebastian, Florida Dear Mr. Schuessler: As you are aware, the City of Sebastian is currently trying to negotiate with General Development Utilities to purchase its potable water system within the City. One of the concerns that the City has is the problem with the total trihalomethanes (TTHM) exceeding the maximum contaminant level as established by the U. S. Environmental Protection Agency. The last samples of water from the GDU system that were collected by your Department were on April 28, 1992, (see attached letter). Since the analysis was done almost a year ago, I would like to request another trihalomethane anaylsis on the GDU System to determine the present level of TTHM. I would like to be present when you or a member of your staff take the test samples. Please call me at your convenience so we can schedule a visit to the existing General Development Utilities water treatment plant for the purpose of collecting the samples. Also, I would like to take samples at various other locations throughout the distribution system to provide a comparison with the results that will be determined at the plant site. Sincerely, Richard B. Votapka, P.E. Utilities Director RBV/ar CC: Robb McClary, Sebastian City Manager ' HARTMAN c7 ASSOCIATES INC. engineers, hydrogeologists, surveyors & management consultants April 13, 1993 HAI #92-023.06 Mr. Richard B. Votapka, P.E. Utilities Director City of Sebastian P.O. Box 780127 Sebastian, Florida 32978-0127 Subject: Ammoniation System Design - Trihalomethane Control System for Water Treatment Plant Dear Mr. Votapka: This letter constitutes our engineering proposal to design, permit, inspect, certify and prepare record drawings for the installation of an ammoniation system at the General Development Utilities, Inc. (GDU) Filbert Street Water Treatment Plant (WT?) for the purpose of trihalomethane control. The ammomation feed system will quench the trihalomethane formation reaction and limit the formation potential of the compound in the drinking water. The cost for survey of the site for which the improvements are to be placed is $650. The final design drawings and specifications will be prepared on behalf of the City at a cost of $7,500. The cost for the permit fee to the Florida Department of Environmental Regulation (FDER) is $4,000, and the preparation of the permit application is $500, for a total of $4,500. The cost of the day of inspection, preparation of record drawings and certification to FDER is $850. The cost of the preparation and coordination with the ammonia supplier and rental agreement with the ammonia supplier for facilities located at the WTP site to be provided by the ammonia supplier is $1,000. The total survey, engineering, permitting and technical services cost for this assignment through completion and record drawings is $14,500. 201 EAST PINE STREET - SUITE 1000. ORLANDO. FL 32801 TELEPHONE (407) 839-3955 - FAX (407) 839-3790 PRINCIPALS: JAMES E. CHRISTOPHER -CHARLES W. DRAKE -GERALD C. HARTMAN • MARK I. LUKE -MARK A. RYNNING -HAROLD E. SCHMIDT. JR. L Mr. Richard B. Votapka, P.E. April 13, 1993 Page 2 The engineer's cost estimate for this project is as follows: Site preparation and clearing - $1,000. Yazdpiping - $61,900. Slabs - $3,800. Other concrete work - $1,000. Structure - $99200. Metering equipment, valves, gauges, and related instrumentation - $17,200. Electrical - $2,700. Special supports and related equipment accommodation facilities - $3,100. Total $449900 Contingency at 10% $4,490 TOTAL $499390 The above is a conceptual cost estimate and subject to revision once specific client desires are obtained and field investigations are completed. The project schedule would be to have the design complete within one (1) month of notice to proceed; permitting complete within three (3) months thereafter; and construction complete within seven (7) months thereafter. The hourly costs and other directs costs shall be as contained in our other contracts with the City of Sebastian. Mr. Richard B. Votapka, P.E. April 13, 1993 Page 3 We look forward to providing the technical expertise which you desire. If the above is acceptable to the City of Sebastian, please execute one (1) copy of this proposal and return it to our offices. Witness Witness Witness Witness GCH/ch P10/Votapka.gch Very truly yours, Hartman & Associates, Inc. -Gerald C. Hartman, P.E. President City of Sebastian, Florida Authorized Signature Date City of Sebastian POST OFFICE BOX 780127 ❑ SEBASTIAN, FLORIDA 32978 TELEPHONE (407) 589-5330 ❑ FAX (407) 589-5570 M E M O R A N D U M DATE March 19, 1993 FROM Richard B. Votapka, Utilities Director TO Trihalomethane File SUBJECT Meeting with Glenn Schuessler, Asst. Director, Indian River County Environmental Health Dept. Glenn told me that in his experience with trihalomethane testing (THM), he found that the longer the water remains in the distribution system, the higher the THM's seem to be. He plotted this on a map of the GDU system where he had taken tests. . A � b a . t. {" & ASSOCIATES, INC. engineers, hydrogeologists, surveyors & management consultants Mr. Richard B. Votapka, P.E. Utilities Director City of Sebastian P.O. Box 780127 Sebastian, Florida 32978-0127 April 13, 1993 HAI #92-023.06 ,"1 ap,�p/A Subject: Ammoniation System Design - Trihalomethane Control System for Water Treatment Plant Dear Mr. Votapka: This letter constitutes our engineering proposal to design, permit, inspect, certify and prepare record drawings for the installation of an ammoniation system at the General Development Utilities, Inc. (GDU) Filbert Street Water Treatment Plant (WT?) for the purpose of trihalomethane control. The ammoniation feed system will quench the trihalomethane formation reaction and limit the formation potential of the compound in the drinking water. The cost for survey of the site for which the improvements are to be placed is $650. The final design drawings and specifications will be prepared on behalf of the City at a cost of $7,500. The cost for the permit fee to the Florida Department of Environmental Regulation (FDER) is $4,000, and the preparation of the permit application is $500, for a total of $4,500. The cost of the day of inspection, preparation of record drawings and certification to FDER is $850. The cost of the preparation and coordination with the ammonia supplier and rental agreement with the ammonia supplier for facilities located at the WTP site to be provided by the ammonia supplier is $1,000. The total survey, engineering, permitting and technical services cost for this assignment through completion and record drawings is $14,500. 201 EAST PINE STREET - SUITE 1000.ORLANDO, FL 32801 TELEPHONE (407) 839-3955 - FAX (407) 839-3790 PRINCIPALS: JAMES E. CHRISTOPHER - CHARLES W. DRAKE - GERALD C. HARTMAN - MARK I. LUKE • MARK A. RYNNING • HAROLD E. SCHMIDT, JR. Mr. Richard B. Votapka, P.E. April 13, 1993 Page 2 The engineer's cost estimate for this project is as follows: Site preparation and clearing - $1,000. Yard piping - $6,900. Slabs - $3,800. Other concrete work - $19000. Structure - $9,200. Metering equipment, valves, gauges, and related instrumentation - $17,200. Electrical - $22700. Special supports and related equipment accommodation facilities - $3,100. Total $44,900 Contingency at 10% $4,490 TOTAL $499390 The above is a conceptual cost estimate and subject to revision once specific client desires are obtained and field investigations are completed. The project schedule would be to have the design complete within one (1) month of notice to proceed; permitting complete within three (3) months thereafter; and construction complete within seven (7) months thereafter. The hourly costs and other directs costs shall be as contained in our other contracts with the City of Sebastian. v 1 0 Richard B. Votapka, P.E. April 13, 1993 Page 3 We look forward to providing the technical expertise which you desire. If the above is acceptable to the City of Sebastian, please execute one (1) copy of this proposal and return it to our offices. Witness Witness Witness GCH/ch P10/Votapka.gch Very truly yours, Hartman & Associates, Inc. Gerald C. Hartman, P.E. President City of Sebastian, Florida Authorized Signature Date City of Sebastian POST OFFICE BOX 780127 o SEBASTIAN, FLORIDA 32978 TELEPHONE (407) 589-5330 0 FAX (407) 589-5570 H E H O R A N D U H DATE : Harch 12, 1993 TO : File FROM : Rich Votapka, Utilities Director /M/ SUBJECT : Trihalomethames (THH's) In my meeting with Glenn Scheussler, Assistant Director of the DHRS Environmental Health Department on Harch 10, 1993, Glenn gave me the following information re: Trihelomethanes: TTHHs or Total Trihalomethanes consist of four (4) compounds: 1) Chloroform ( CP C't3 ) 2 ) Bromoform CCW Br3 J 3 ) BDCH - Bromo Di Chloro Methane 4 ) CDBH - Chloro Di Bromo Methane (CN 3r? C 1 Tests for the above are expressed in UG/L which means micrograms per liter or ppb -parts per billion. The aggregate total for the four compounds above must be less than 100 ug/1 to not constitute a health threat. The Environmental Protection Agency (EPA) maintains the curent standard for THRs is 0.1 mg/L for systems serving more than 10,000 people. The EPA's D -DBP rule (Disinfectant & Disinfection By -Products Rule) which will become effective after June, 1995, will impose that level on all systems and may reduce the allowable level. Trihalomethanes (THHs) are compounds suspected to be cancer -forming when ingested in sufficient quantities. RBV/ar I STATE OF FLORIDA DEPARTMENT OF HEALTH AND REHABILITATIVE SERVICES HRS - INDIAN RIVER COUNTY PUBLIC HEALTH UNIT ENVIRONMENTAL HEALTf TE6EPHONS (407) 778.8321 1800 27TH STREET St1N+COM 240-6321 VERO BEACH. FL 329W FAX 778-6303 March 30, 1992 Ms. Janine Morris U. S. Environmental Protection Agency 345 Courtland Street, N. E. Atlanta, GA 30365 RE: General Development Utilities FL DER ID # 3311136 Sebastian Highlands Sebastian, FL Dear Ms. Morris: In reference to your request, I have enclosed copies of laboratory results regarding total trihalomethane analysis for the above referenced facility. Our initial investi- gation was conducted on October 26, 1990 which consisted of collecting two (2) samples, one inside a residence (950 Beach Lane) and one outside the residence. Subsequent collection and sampling of the water system at other locations was conducted on March 29, 1991 and October 22, 1991. If you have any questions or we can be of further assistance, please feel free to contact me. ' S irl4erfily, ' ,en R. Schuessler Asst. Env. Health Director cc: Paul Morrison, FL Dept. of Env. Reg., Central District LAWTON CHILES. GOVERNOR STATE OF FLORIDA DEPARTMENT OF HEAUP I AND REHABILJTATIVE SERVICES HRS - INDIAN RIVER COUNTY PUBLIC HEALTH UNIT ENVIRONMENTAL HEALTH TELEPHONE (407) 778-6321 1$00 27TH STREET SLIN-COM 240.6321 V ERO BEACH. FL 32560 FAX 7 78 6303 May 27, 1992 Ms.. Janine Morris U.S. Environmental Protection Agency 345 Courtland Street, N.E. Atlanta, Ga 30365 RE: General Development Utilities FL. DER #3311136 Sebastian Highlands Subdivision Sebastian, Florida Dear Ms. Morris: Enclosed are the most current copies of laboratory results regarding total Trihalomethane analysis for the above referenced facility. The samples were collected on Apr.tl 28, 1992. If you have any questions or if we can be of further assistance, please feel free to contact me. Sincerer,-. Glenn R. Schuessler Asst. Environmental Director LAWTON CHILES, GOVERNOR