Optimization For Disinfection Byproducts_11!15!2008

Optimization of Treatment for Disinfection Byproducts Control

Optimization of Treatment for Disinfection Byproducts Control

Optimization of Treatment for Disinfection Byproducts Control Jeffery L. Droll [email protected] Warren National University FPPE490: Final Paper Dr. Magdy Girgis PhD November 12, 2008

1

Optimization of Treatment for Disinfection Byproducts Control

2

Abstract In January, 2002 modifications to the Clean Water Act went into effect that lowered the maximum level of specific contaminants associated with the disinfection of drinking water with chlorine referred to as disinfection byproducts (DBP’s). This rule had the greatest impact on purveyors that treat water from surface sources such as rivers, lakes and reservoirs The reason for this impact is the presence of naturally occurring organic matter (NOM) produced by the organisms living in the water or within the water shed from which the water is collected. The NOM then reacts with the chlorine being applied for disinfection producing undesirable byproducts. This essay will present a case study of one facility where conventional treatment methods and each unit process was successfully optimized to yield a finished water that met both the regulatory requirements and the aesthetic demands of the customers.

Optimization of Treatment for Disinfection Byproducts Control

3

Table of Contents I) Introduction A) The Ottawa Ohio Water Treatment Plant Case Study B) What is Optimization C) Disinfection Byproducts 1) What Are Disinfection Byproducts? 2) What is Naturally Occurring Organic Matter?

7 7 11 12 12 14

II) Optimization A) Examination of Historical Records 1) Water Quality for the Previous Two Years a) Raw Water Source and Raw Water Characteristics b) Water Quality after Treatment in the Clarifiers c) Water Quality after Recarbonation d) Water Quality after Filtration e) Water Quality at the Plant Tap 2) Evaluation of Individual Unit Processes a) Existing Pretreatment Capabilities b) Raw Water Supply and Low Service Pumping c) Rapid Mixing d) Clarifiers (Coagulation, Flocculation,

16 16 16 18 17 19 21 22 24 24 25 26

Sedimentation) i) Coagulation ii) Flocculation iii) Sedimentation e) Recarbonation Basins f) Filtration g) Transfer Pumping (Intermediate Pumping) h) Clear Well Storage (Finished Water Storage) and Disinfection i) High Service Pumping 3) Chemical Feed Systems a) Potassium Permanganate b) Powdered Activated Carbon c) Coagulant Feed System d) Lime Feed System e) Soda Ash Feed System f) Carbon Dioxide Feed System g) Phosphate Feed System h) Chlorine Feed System

28 28 29 31 34 36 44 44 46 47 47 48 49 50 51 52 53 54

Optimization of Treatment for Disinfection Byproducts Control

III) Jar Testing A) Initial Phase (Establishing a Baseline) 1) Verification of Present Conditions and Treatment a) Establish Rapid Mixer Energy Transfer Values b) Establish Flocculation Detention Time c) Establish Sedimentation Time 2) Pretreatment Chemical Dosages 3) Coagulant Dosages 4) Softening Chemical Dosages 5) Simulated Recarbonation (pH Adjustment) 6) Phosphate Dosages 7) Disinfection and Simulated Distribution System Testing B) Experimental Phase (Chemical) 1) Modification to Chemical Dosages a) Pretreatment Chemical Dosages i) Sequencing of Chemical Additions b) Coagulant Dosages i) Single Basin Treatment ii) Split Treatment c) Softening Chemical Dosages d) Disinfection, Simulated Distribution System Testing 2) Compilation and Comparison 3) Modifications to Unit Processes a) Addition of Reaction (Contact) Basins b) Modifications to Rapid Mix Basins Required c) Separation of Treatment between Clarifiers d) Modifications to Sludge Recirculation Piping on Clarifiers e) Control of Solids in the Clarifier Slurry Pool and Reaction Zone

4

56 56 56 57 58 59 60 61 61 63 63 64 64 64 65 66 67 68 68 70 71 72 72 72 73 74 75 76

IV) The Human Factor

77

V) Summary A) Results of Optimization 1) Summary of Changes a) Unit processes b) Chemical Treatments VI) Looking Forward References

79 79 79 79 81 83 84

Optimization of Treatment for Disinfection Byproducts Control

Appendices Abbreviations and Acronyms Tables Figures

5

89 90 92 107

Optimization of Treatment for Disinfection Byproducts Control

6

I) Introduction A) The Ottawa Water Treatment Plant Case Study I chose the Ottawa Water Treatment Plant for this paper for several reasons, but primarily because the success achieved demonstrates the effectiveness of process optimization on obtaining compliance with the recently promulgated limits for specific contaminants in drinking water. The Ottawa Water Treatment Plant and its water source are typical of most surface water treatment plants constructed during the 1960’s and 1970’s in Ohio. However, extra thought was given to specific details of this treatment plant’s design by the engineers that allowed it to be adapted and the treatment optimized with minimal physical modifications during this process. The water source for the Village of Ottawa is the Blanchard River. This source was chosen when the demands for water within the village exceeded the capacity of the aquifer in the region. According to the United States Geologic Survey (USGS)* the river has an estimated drainage area (water shed) of approximately 350 square miles and an average daily discharge of approximately 271 cubic feet per second (2,027 gallons per second) (gps). This average daily flow combined with the 121 million gallon reservoir along the north bank of the river which was constructed at the same time as the treatment plant was, allows for an abundant supply of water at all times. Additionally, the reservoir capacity allows the treatment plant operators to monitor the quality of the water in the river and to be selective when impounding water from the river.

Optimization of Treatment for Disinfection Byproducts Control

7

The Ottawa Water Treatment Plant is what is generally considered a “conventional” treatment plant using chemical precipitation of suspended contaminants and excess hardness. However, this particular treatment plant was designed with solids contact clarification units installed at different elevations allowing for both series and parallel operation of the units. The importance of this feature will be expanded upon later. A general overview of the unit processes used and their purpose is as follows. Processes added or found to need substantial modification as a result of this optimization are denoted with an asterisk. Low service pumping: Low service pumping is used to supply the energy required to lift the raw water into the treatment plant when the water elevation in the reservoir is not at a sufficient height to accomplish this. Pretreatment Basins:* The pretreatment basins are used to provide contact time for the addition of potassium permanganate (KMnO4) and powdered activated carbon (PAC) that are used to condition and adsorb organic contaminants, respectively. Each basin provides 30 minutes of residence time with mixing to assure that the chemicals have been brought into contact with all of the water present and to allow adequate time for the chemicals to fully react. Rapid mixing:* A rapid mixer is a small chamber in the influent channel equipped with a high intensity mechanical mixer that is used to thoroughly disperse chemical coagulants throughout the raw water before the chemicals can fully react with the alkalinity in the water. Clarification:

Optimization of Treatment for Disinfection Byproducts Control

8

These units are large circular basins that are partitioned into three zones. They combine three processes in on physical structure. These processes are: Reaction/Mixing zone: Here additional chemicals whose reaction times are less critical than the coagulation chemicals can be added and the reactions from all of the chemicals added to that point are allowed to go to completion. Additionally, here a controlled quantity of previously precipitated solids are brought back up from the bottom of the clarifier and mixed with the water to act as a substrate and ballast for the newly forming precipitates. Flocculation zone: This is a zone where the flow is downward. Here a gentle turbulence is induced into the water by the manner in which the water is discharged from the mixing zone. These turbulence cause collisions between the newly formed precipitates and recirculated solids that because of their shape and structure allow them to link together to form even larger faster settling particles referred to as “floc”. Settling zone: This zone is the largest portion of the clarifier and the flow is upward. As the floc particles formed in the flocculation zone pass under the partition the greater majority of the particles are too large and heavy to be carried upward and continue downward to the bottom (sludge sump) of the clarifier. Recarbonation: Recarbonation is the process of adding carbon dioxide back to the water to realign the alkalinity species so the water is neither corrosive nor depositing. Carbon dioxide forms carbonic acid that reacts with any residual hydroxide ions present and then a portion of the carbonate anions to produce bicarbonate alkalinity. An approximately even balance between carbonate and bicarbonate anions produces the most stable water. Filtration:

Optimization of Treatment for Disinfection Byproducts Control

9

Filtration is in essence the final polishing of the water before it is disinfected and delivered to the distribution system. Here beds of sand particles of nearly uniform size (approx. 0.5 mm) trap the small particles that would not settle in the clarifier are removed along with a substantial number of bacteria and bacterial cysts. Disinfection: Disinfection is usually performed by injecting a solution carrying dissolved chlorine gas or sodium hypochlorite (bleach) into the filtered water. The normal dosage is approximately 4 milligram per liter (mg/L) Clearwell Storage: The clearwell is the storage tank used to hold the filtered and disinfected water. Here sufficient time is provided to allow the chlorine to subdue the disease causing microorganisms. Additionally, the volume of the clearwell provides a reserve for high demand periods and storage during low demand periods to reduce variations in plant production. It may be likened to a warehouse. However, the clearwell and the storage time it provides are both a blessing and a curse for surface water treatment plants. This point will be expanded on in later sections. High Service Pumping: These pumps provide the energy required to move the treated water out into the distribution system at the volume and the pressure required. Distribution system storage: Distribution system storage is a means of storing water at different locations out in the community to provide the ability to maintain both pressure and volume in locations where such could not be reliably accomplished by the high service pumps at the treatment plant. This storage can be in the form of elevated tanks commonly seen around communities or in the form of underground storage tanks where the water is transferred back to the distribution system by pumps when needed.

Optimization of Treatment for Disinfection Byproducts Control

10

Each of these unit processes were looked at in detail during this study and a method was developed to optimize the performance of each to reduce the production of disinfection byproducts in the water without adding substantially to the operational complexity of the facility. B) What is Optimization? One dictionary definition of optimization is to make something function at its best. Another definition of optimization is to accomplish a task with the smallest amount of effort. The later definition is often confused with the concept of minimization. By either definition, something must be given up. In the past, disinfection byproducts (DBP’s) were not considered when treating water. Chemical dosages and process unit sizes were determined base on what would produce an acceptable product for the smallest investment. When DBP’s were found and determined to pose a significant health risk to those consuming the water, (Chloral Hydrate in Drinking Water (2007)) many previous treatment strategies and treatment plants proved to be, for the most part, inadequate. (Hill, Fred (2007)), (World Health Organization (2004)). In optimizing a water treatment plant for DBP control it was found that there were optimum chemical dosages and optimum methods of operating individual treatment units. I some cases a study of the chemical dosages produced a Bell Curve while others reached a point of diminishing return (Perišio, M. (2006)). When studying the physical treatment unit sizes, most often the principle of diminishing return was the determining factor. However, the strategy for operating the treatment units often produced a Bell Curve (Heinonen, Pekka and Pisto, Sannimaria -Lopez (2007)).

Optimization of Treatment for Disinfection Byproducts Control

11

In this study we will examine what was done to optimize chemical dosages and the operation of the treatment units for DBP control. C) Disinfection Byproducts 1) What Are Disinfection Byproducts? Disinfection byproducts are produced when the chlorine that is added to the water for disinfection reacts with the small amounts of organic matter that is left in the water after treatment. The terms applied to these contaminants are THM’s and HAA5’s (Weeks, Daniel P., PhD (2003)). These terms are acronyms for Trihalomethane and the five Haloacetic acid compounds presently being tested for. On the periodic table of elements, the elements in the column in which chlorine resides are referred to as halogens. Therefore the term “halo” is applied to these compounds. The chlorine compound most commonly used for disinfection is Hypochlorite. It is produced when chlorine gas reacts with water. The hypochlorite molecule acts as an anion with a charge of negative one (1-). The molecule consists of one chlorine atom, one oxygen atom and depending upon the pH, a hydrogen atom or an alkali metal (as found in bleach, sodium hypochlorite). The effectiveness of this molecule comes from the fact that it contains two of the most highly active, negatively charged elements, oxygen and chlorine (Weeks, Daniel P., PhD (2003), pg. 36). This combination disinfects by disrupting the structure of the molecules that make up the outer membranes of microorganisms and viruses, and then by upsetting metabolic activities within the cells. Organic molecules can be generally classified as saturated and unsaturated. Saturated organic molecules have hydrogen atoms at all locations on the carbon atom or carbon

Optimization of Treatment for Disinfection Byproducts Control

12

atom chain where substitution can take place. Unsaturated organic molecules have something other than hydrogen at one or more of the available sites. (Basuray, Sagnik and Chia Chang, Hsueh, (2007)) In the study of biology you will find that the molecules making up the membranes of microorganisms are unsaturated organic molecules, therefore a reaction with hypochlorite is possible. In the concentrations used for disinfection, hypochlorite will seldom react with “saturated” organic molecules. The unsaturated organic molecules already have one of the hydrogen atoms substituted by an oxygen atom or carbon atoms that are double bonded to other carbon atoms. These substitutions upset the balance of charges (electrons) around the carbon atom or carbon atom chain allowing the highly negatively charged components of the hypochlorite molecule to remove the remaining hydrogen atoms and substitute for them. These molecules that are not fully saturated with hydrogen come from the metabolic activities of other living things or from the remains of those living things that have died. Since these substituted molecules tend to be slightly polar in nature because of the presence of the oxygen or the double bond, they tend to be highly soluble in water. An example of these molecules would be methanol alcohol, acetic acid (vinegar) and ethanol alcohol. All of which are common precursors for disinfection byproducts. Chlorine is not the only halogen element found in these DBP compounds. Small amounts of bromine will be commonly found in natural waters with trace amounts of iodine being occasionally found. These two elements are not normally found in a state that will allow them to engage in substitution reactions. They are oxidized to hypobromite and hypoiodite by the hypochlorite allowing them to engage in these reactions in the same manner as the hypochlorite (Pascal Roche, Benanou, David. (2006)).

Optimization of Treatment for Disinfection Byproducts Control

13

Although it is said that several hundred halogenated organic compounds have been identified, the compounds tested for in these two groups are produced in the largest quantities and have been identified as posing a threat to the health of those who consume the water. The THM compounds tested for are Chloroform, Bromo-dichloro-methane, Dibromochloro-methane and Bromoform. The HAA5 compounds tested for are Monochloroacetic acid, Monobromoacetic acid, Dichloroacetic acid, Trichloroacetic acid, Bromochloroacetic acid and Dibromoacetic acid. 2) What is Naturally Occurring Organic Matter? Naturally occurring organic matter (NOM) consists of soluble and semi-soluble organic (carbon based) molecules. The molecules are the remnants of organisms that have died, were eaten, or were produced from the metabolic activities of the organisms that inhabit the water and the surrounding water shed that provides the source of the water to be treated. These molecules vary in size from single carbon (methyl) alcohols and aldehydes, to massive humic acid molecules that may contain more that 10,000 carbon atoms. These larger molecules are the ones that give healthy river and pond water its characteristic earthy or musty smell (Prashant, Kumar (2003)). In most natural systems the tendency is for the organisms to utilize the organic molecules as food and successively break them down into ever smaller parts. Often these remaining small molecules are utilized by bacteria as food and converted to a final mineral (inorganic) state. However, small quantities of them often remain. These remaining molecules are the molecules that provide the substrates (precursors) for THM and HAA5 compounds.

Optimization of Treatment for Disinfection Byproducts Control

14

Often during treatment strong acidic, alkaline or oxidizing chemicals are applied to the water. Some of these chemicals have been found to increase the number of precursors in the water by breaking pieces off of the very large molecules during the treatment process. These are some of the concerns we will be addressing during the optimization process. Those are; how to remove the larger molecules without creating more precursors and to utilize the chemical conditioning and adsorptive characteristics of certain treatment chemicals to remove as many of the precursors already present as possible.

Optimization of Treatment for Disinfection Byproducts Control

15

II) Optimization A) Examination of Historical Records The examination of historical operating records is a critical part of any optimization work. This should include any information the plant operators have compiled over the years. Most of the records will be the results of standardized laboratory tests. This type of information is the most useful when comparing the performance of the processes to industry benchmark standards. It allows for trending of critical parameters and the building of relationships between operation of the plant and finished water quality. However, the most often overlooked information is notes made by the operators expressing subjective opinions about the performance of a piece of equipment, some attribute of the water that is not tested for regularly such as the color or odor of the raw water, weather conditions or an opinion pertaining to the quality of a treatment chemical. Any of these personal insights have the potential to help direct energy in the most useful direction. Many times this information is only available through lengthy conversations with the staff or from operator’s logs that do not contain test data. Often these notes are about events within the plant or the water shed. Although these notes may not directly affect the data, they may explain anomalies and allow parsing of the data that would otherwise tend to skew the analyses of the data. 1) Water Quality for the Previous Two Years Because the amount of data collected on a daily basis at most water treatment plants can be immense, it is necessary to limit it to a range that will provide the most useful information for the work at hand. Experience has shown that the overall source water

Optimization of Treatment for Disinfection Byproducts Control

16

quality will change very little over time if there have been no major changes within the watershed. Therefore two years of data will allow the observation and trending of both seasonal and weather related changes along with changes in the process units Table 1 is a listing of the average water quality data for the two years prior to this study. a) Raw Water Source and Raw Water Characteristics Initially the raw water source is looked at to determine if it is capable of supplying enough water to meet the needs of the community. This is an important consideration because the quality of the water in the stream will vary over the course of the year and if the amount of water available in the stream at any given time is sufficient to allow impounding; the plant operators can be selective and impound only when the water quality is determined to be acceptable. The basic parameters used to determine water quality are: • • • •

Turbidity Organic content (determined by UV254 scan) Alkalinity Hardness

The first parameter directly affects the longevity of the reservoir since the turbidity in a river is primarily composed of small suspended particles that will become sediment in the reservoir. The other three parameters directly impact the economics of treating the water by dictating the amounts of treatment chemicals required to achieve an acceptable finished product. The raw water source is the Blanchard River. According to the USGS the Blanchard River serves a water shed of approximately 350 square miles (USGS 04189000

Optimization of Treatment for Disinfection Byproducts Control

17

Blanchard River) and the river has an average monthly flow of 271 cubic feet per second (2,027 gallons per second), which equates to 175 million gallons per day. From the data indicating an average flow past the treatment plant of this magnitude and the community’s average need being less than 3 million gallons per day, the treatment plant operators can be selective. Conversations with the operators indicated they were using this to their advantage. b) Water Quality after Treatment in the Clarifiers The quality of treatment obtained through the clarifiers is probably the most critical stage of the treatment process. The clarifiers are in essence large high rate chemical reaction vessels. As with any chemical process there are several factors which will determine the overall efficiency of the process. The basic characteristics of the clarifiers and their average effluent water quality are detailed in Table 2. The operators perform laboratory tests on the clarified water to determine the effectiveness of treatment. Although these tests are not mandated by any regulatory agency, they are considered necessary by the operators to control the process. These tests are: • • • • •

Turbidity Alkalinity Hardness pH UV254

Turbidity on clarified water is an indicator of the effectiveness of treatment as it relates to the removal of suspended and precipitated matter from the water. The largest portion of

Optimization of Treatment for Disinfection Byproducts Control

18

these vessels is designed to provide quiescent conditions with a decreasing flow velocity as the waster rises through the outer portion. The alkalinity of the water is a measure of the amounts of calcium and magnesium present that is associated with a carbonate or hydroxide anion. Often times this is referred to as the buffer capacity of the water. The hardness of the water allows the operators to determine the total amount of calcium and magnesium in the water and differentiate between the amount of calcium and magnesium that is alkalinity (carbonate) and the amount of calcium and magnesium that is associated with non-carbonate anions. The pH of the water allows the operators to determine the species of the carbonate anions present since the balance of carbonate and bicarbonate anions is critical to the stability (corrosiveness) of the water; this allows them to determine the amount of adjustment that is required in the recarbonation process. The results of the UV254 analysis performed on the clarified water is compared to the results of the UV254 analysis performed on the raw water to determine the process’s efficiency at removing organic material from the water. As with the raw water, laboratory analyses records from the previous two year was examined. The information shown in Table 2 reflects the averages of the clarified water quality for the two years prior to this study. c) Water Quality after Recarbonation

Optimization of Treatment for Disinfection Byproducts Control

19

During precipitative softening in the clarifier the pH of the water is raised by the addition of calcium hydroxide (lime) to a level where all of the alkalinity is converted to either carbonate or hydroxide species. These alkalinity species associated with calcium and magnesium, respectively, have a low solubility at the elevated pH allowing them to come out of solution (precipitate). Under these conditions the remaining alkalinity is barely soluble and is just at equilibrium. This remaining alkalinity is considered unstable because even slight changes in either temperature or pressure will cause it to plate out on the inside of pipes and plumbing fixtures slowly plugging them. More importantly, if this unstable water is applied to the filters, which immediately follow recarbonation, this alkalinity will plate out on the filter sand increasing the effective size of the sand grains. This increase in grain size will reduce the ability of the filters to remove small particles. Recarbonation is the process used to alter the alkalinity species and adjusted them to a more soluble form. The process is the addition of carbon dioxide gas directly to the water. Upon contact with the water, the carbon dioxide gas reacts with the water to form the weak acid, carbonic acid, and dissolves. The carbonic acid first reacts with the hydroxide anions to form carbonate and water. Once all of the hydroxide anions have been converted, the remaining carbonic acid reacts with the carbonates to produce bicarbonate anions, which are slightly acidic. (Edwards, M., Scardina, P. (2006)) A proper balance of the carbonate to bicarbonate alkalinity will produce finished water that will neither corrode metal components nor deposit on the interior surfaces of pipes and fixtures. At this point the water is considered chemically stable. Here the operators performed laboratory analyses to monitor the process. These analyses are:

Optimization of Treatment for Disinfection Byproducts Control

• •

20

Alkalinity pH

The alkalinity test indicates the species and concentrations of the alkalinity present. Since at this point the pH is a function of the balance of the alkalinity species, (OTCO, 2007, Precipitative Softening) it serves as an indicator of the correctness of the alkalinity tests. d) Water Quality after Filtration Following recarbonation, the treated water is applied to three high rate (rapid) sand filters. Filtration is normally looked upon as a polishing process. However, the filtration process, when operated correctly provides a highly efficient barrier against microorganisms and the cysts of several pathogens. Because of this process’ importance to the health of the system’s users, the performance of the filtration process is the most tightly regulated of all of the treatment processes. There are three analyses performed of the filtered water. These are: • • •

Turbidity Alkalinity pH

The overall performance of the filter and the preceding processes are evaluated by the turbidity of the filtered water. By federal law, the filtered water turbidity must be less than 0.3 Nephelometric Turbidity Units (NTU) in 95 percent of the samples taken every 15 minutes and the turbidity may not exceed 0.5 NTU at any time. The reason for this is that it has been found that microorganisms can live on and within the small particles passing through the filter and possibly avoid deactivation by the disinfection process in this way.

Optimization of Treatment for Disinfection Byproducts Control

21

Therefore, when the filter operation records were examined, great attention was given to any exceedances and the quality of the water being applied to the filters. There are several other operating parameters that can be extracted from the basic filter operating data pertaining to the economic efficiency of the filters. These parameters will be elaborated on in Section II, 2, Evaluation of Individual Unit Processes. The alkalinity of the filtered water is monitored closely. This parameter directly indicates if carbonate deposits are being formed on the sand grains. If there is a decrease in the alkalinity of the filtered water it would indicate that the water had not been fully stabilized by the recarbonation process. If deposition is determined to be occurring, the amount that is being deposited can be estimated by converting the difference in alkalinity between the filtered and unfiltered water to pounds per day. Deposition on the sand grains will increase the effective size of the sand grains and reduce the filter’s effectiveness at removing smaller particles. Again here pH is looked at to verify the accuracy of the alkalinity test. e) Water Quality at the Plant Tap The finished water at the plant tap is subjected to a battery of tests that are mandated by the Ohio and US Environmental Protection Agencies (inclusive, EPA). A list of these tests is shown in Table 1 and the two year average of these tests is shown in the Finished Water column. The parameters most frequently tested for are:

Optimization of Treatment for Disinfection Byproducts Control

• • • • •

22

Turbidity Alkalinity pH Hardness Chlorine residual

Here again the turbidity of the finished water is the primary indicator of the success of the preceding processes. Here also the same limits for turbidity are enforced that apply to the filter effluent and for the same reasons. The alkalinity, pH and hardness are used to determine the stability of the water. These parameters combined with the water temperature, allow stability indexes to be calculated. On Table 1 these indexes are the Langelier Index (LI) and the Calcium Carbonate Precipitation Potential (CCPP). The Langelier Index is a non-dimensional number. Values greater than zero (0) indicate a tendency for carbonates to precipitate and values less than zero (0) indicate corrosive water. The CCPP calculation is much more complex than the Langelier Index, and will give a result that indicates in milligrams per liter how much alkalinity may be lost from the water (precipitated) or gained by it (corrosion) to reach stability. The chlorine residual present in the finished water is a measure of the chlorine left following disinfection with chlorine gas or sodium hypochlorite. This parameter is also regulated by the EPA (Formation of Chlorinated Organics in Drinking Water (2005)). To remain in compliance with the regulations the amount of free chlorine (hypochlorite ion) remaining in the water must be >0.2 mg/L at any point in the distribution system but <4.0 mg/L at the water’s point of entry into the distribution system.

Optimization of Treatment for Disinfection Byproducts Control

23

One parameter closely studied in an optimization study is the chlorine demand. This is the difference between the chlorine residual present in the finished water and the amount of chlorine gas or hypochlorite ion applied to the water. Since the primary focus of this optimization study is to reduce disinfection by product and since the level of disinfection by product produced can be closely correlated to the chlorine demand of the water; this parameter is of great interest and value. 2) Evaluation of Individual Unit Processes During this phase each of the individual unit processes are evaluated from three perspectives. These are: 1) Their average performance over the preceding two years 2) Physical characteristics in relationship to accepted minimum industry standards 3) Their present mode of operation. a) Existing Pretreatment Capabilities Pretreatment is generally used as a conditioning process. The most common chemicals used for pretreatment are chemical oxidizers and adsorbents. The purpose of these chemicals is to oxidize both organic and inorganic contaminants and to adsorb contaminants and some of the contaminants conditioned by oxidation prior to them entering the following treatment processes. Currently there were no facilities dedicated to providing pretreatment. There are two chemicals available for pretreatment. These are potassium permanganate and powdered activated carbon (PAC) (Gaulinger, Siegfried. (2007)).

Optimization of Treatment for Disinfection Byproducts Control

24

b) Raw Water Supply and Low Service Pumping

The Blanchard River serves as the raw water source. An above ground reservoir, constructed in 1971 is used to store raw water for the treatment plant. A river pumping station is located adjacent to the Blanchard River. The pumping station contains screening equipment and three pumps. Pump capacities are shown in Table 4.

The reservoir is a 28-acre impoundment basin that has a storage volume of 121 million gallons at an average depth of 21 feet. The usable volume based on ODNR water overflow restrictions is approximately 106 million gallons. The reservoir does not provide sufficient storage volume based on current guidelines. According to current guidelines, the Ohio EPA recommends that total raw water storage should provide a minimum of 270 days based on the average plant production rate. The current average plant production rate is approximately 1.6 MGD. Based on this the reservoir currently only provides 66 days of storage. At the current plant design flow, the reservoir will provide only approximately 35 days of storage. Additional raw water reservoir storage should be considered. Any additional reservoir capacity needs to be evaluated based on existing river flow data and Ohio EPA recommendations. The elevation of the water in the reservoir is normally maintained at a height sufficient to supply the head (pressure) necessary to move the water through the initial treatment processes at a rate of 2,400 gpm. When the water elevation in the reservoir is below this minimum, two raw water pumps provide the necessary head. Normally these pumps are not used. Each of these pumps has a capacity of 2,100 gpm at a head of 25 feet. Each is driven by a 20 horsepower motor. Firm raw water pumping capacity (capacity with one pump out of

Optimization of Treatment for Disinfection Byproducts Control

25

service) is 3.0 MGD. Any future plant expansion plans should address additional raw water pumping capacity. c) Rapid Mixing One in-line mixer is located in the raw water inlet pipe, although it is currently not used. The mixer is driven by a one horsepower motor. The estimated detention time in this mixer is less than one second. The estimated velocity gradient (G value) for the mixer, when used, is between 2,270 sec-1 and 3,320 sec-1. High G values are necessary for in-line mixers particularly when charge neutralization coagulation is employed (Zeta-Meter Inc. Fourth Edition). Here however, sweep coagulation treatment is used rather than charge neutralization for both turbidity removal and organics control. Mixing energies of this magnitude have been found to be detrimental to sweep coagulation where the intent is to develop a large fragile floc particle. When the mixer is operating, the G value produced is greater than that required for sweep coagulation and the detention time provided is too short. For these reasons, it appears that the in-line mixer serves little purpose in the existing treatment process and should be removed. There are two rapid mix basins complete with mechanical mixers. The basins can be configured for either series or parallel flow depending on treatment needs. One basin is dedicated to each clarifier. The primary (west) rapid mixer currently was operated at approximately 30 percent of its maximum speed. The east rapid mixer is currently not used. Each rapid mix basin has a volume of approximately 1,500 gallons. This provides a detention time of 43 seconds at the design flow rate under series operations. The

Optimization of Treatment for Disinfection Byproducts Control

26

detention time in each basin under parallel operations is 87 seconds. The current recommended standard for rapid mixing establish a detention time of not more than 30 seconds (Recommended Standards for Water Works, (2003)). The existing rapid mixers are five horsepower units that can produce an estimated velocity gradient (G value) between 505 sec-1 and 740 sec-1 depending upon the raw water temperature. The primary (west) rapid mixer currently is operated at a speed that produces a G value between 290 sec-1 and 424 sec-1 depending on the water temperature. The current recommended standard is a minimum G value for rapid mixers of 900 sec-1 (Recommended Standards for Water Works, (2003)) The existing rapid mix basins and mixers do not meet current recommended standards. The excessively long detention time is most likely interfering with the coagulation process because the newly formed floc particles are not taken away from the vicinity of the mixer quickly enough and are being sheared prior to leaving the basin. Additionally, the large cross sectional area of the basin in relationship to the diameter of the mixer impeller is likely allowing substantial amounts of water to bypass the mixing process. The existing basin flow configurations exhibit desirable rapid mix design characteristics. Water flow enters the top of each rapid mix basin and is directed downward toward the clarifier influent piping. Properly sized, each basin would be very beneficial to the optimization efforts for coagulation treatment. Optimum rapid mixing conditions employ the proper G value for mixing, pinpoint application of the coagulant chemical, and the use of counter current mixing. Coagulant dispersion is best accomplished when the coagulant is applied near the eye of the mixer impeller. To take advantage of counter current mixing technology, the rapid mixer is configured to push water against the flow through the basin. This is

Optimization of Treatment for Disinfection Byproducts Control

27

accomplished by setting the mixer drive unit for the correct rotation based on the impeller design and water flow characteristics. d) Clarifiers (Coagulation, Flocculation, Sedimentation) i) Coagulation At this time coagulation is provided by adding ferric chloride to the raw water through a connection to the in-line mixer, even though the mixer is not operated. Coagulation is used to foster the gathering of suspended particles and colloids into a settleable floc material. Coagulation also can be used to remove dissolved and suspended organics from the water (Enhanced Coagulation for Surface Water (2006)). The average dosage of ferric chloride presently applied for treatment (32 mg/L) indicates that sweep coagulation is the mechanism of coagulation. Sweep coagulation is a term used to indicate how particles and solids are removed from the water. In sweep coagulation the coagulant reacts with the alkalinity in the water to form sticky hydroxide floc particles. These hydroxide floc particles act like a net to sweep other particles and solids from the water. Particles and solids absorbed into the floc particles materials are later removed during sedimentation. However, the G value produced in the rapid mixing appears to be insufficient for sweep coagulation. As previously mentioned, new rapid mix units should be installed to match the proposed basin modifications. Enhanced coagulation, either by pH adjustment or by increased coagulant dosage, is most likely needed to effectively reduce DBP precursor concentrations prior to chlorination of the water (Enhanced Coagulation for Surface Water (2006)). DBP precursors are organic

Optimization of Treatment for Disinfection Byproducts Control

28

contaminants that create disinfection byproducts (DBPs) following chlorination of the water. The DBPR established requirements for enhanced coagulation, TOC removal, and maximum contaminant levels for DBPs. Bench-scale evaluations will be performed to predict the needs for coagulation and to identify the needs for effective DBP precursor and TOC removal (García, Indiana (2005)). Currently, lime and soda ash are added to the raw water flume upstream of the primary (west) rapid mix basin for softening. The lime may be acting as a coagulant aid for improved turbidity removal. The current application of lime and soda ash at this location are most likely interfering with coagulation treatment. These two chemicals quickly increase the water pH well above the optimum pH for coagulation. Since sweep coagulation is employed in treatment, lime and soda ash should be applied at least 10 seconds downstream of the coagulant feed point. ii) Flocculation The flocculation zone of each clarifier was designed to aid in the development of floc material from the chemically treated water to increase the removal of certain contaminants. The existing flocculation mixing equipment has been out of service since 1973. It appears that lime scale accumulations on the mixers rendered them inoperative. No mechanical mixing of the chemically treated water occurs at this time. For effective flocculation, it is recommended that the mixing intensity (G value) be maintained between 20 sec-1 and 70 sec-1. Adjustment of the flocculation mixer speeds is crucial to the optimization of flocculation treatment. Mixer speeds must be adjusted based on floc density, floc settling characteristics, and water temperature.

Optimization of Treatment for Disinfection Byproducts Control

29

The original flocculation mixers had variable speed capabilities. New mixing equipment is needed for flocculation and should have variable speed capabilities. Floc settling rates and filterability index tests were performed as part of the evaluation of existing treatment. The floc settling rate test helps determine the ability of floc material to settle and allows the estimation of the clarifier up-flow rate necessary to allow effective settling of the floc. The test results showed that the floc size (0.5 mm diameter) is sufficient, but the floc density is too low for effective sedimentation. Floc density is a function of coagulant dosage, mixing intensity, and detention time. Variable speed mixing equipment is needed to enhance flocculation and to increase floc density. The maximum clarifier up-flow rate determine from the test procedure was estimated to be 0.63 gpm/ft2. On the day of the test, the actual up-flow rate was calculated to be 0.66 gpm/ft2. More agitation (higher G value) is needed to increase the density of floc material produced during the flocculation process. Filterability index is a rough measure of the efficiency of particle removal from treatment. Index values typically range between 1.05 and 1.3 in well optimized treatment processes. A filterability index of 1.6 was determined for the settled water produced by the primary (west) clarifier. A filterability index of 1.5 was determined for the settled water produced by the secondary (east) clarifier. Index values greater than 1.3 indicate a need to improve treatment. Necessary improvements may include proper dosage control, changes in rapid mix and flocculation mixing intensity, and the use of a polymer as a coagulant aid. Deficiencies in rapid mixing and flocculation mixing were previously discussed. Polymer addition and other coagulants will be evaluated to determine if they could improve floc density and settling capabilities.

Optimization of Treatment for Disinfection Byproducts Control

30

Recommended Standards establishes a detention time for flocculation at no less than 30 minutes at the design flow rate. The current rated design detention time for flocculation is approximately 45 minutes using both clarifiers in parallel operation. At the normal treatment flow rate (approx. 2.25 MGD), each clarifier provides a flocculation detention time of approximately 30 minutes. Actual flocculation detention time is approximately 30 minutes since only the primary (west) clarifier is presently used for flocculation treatment at 2.25 MGD. The secondary (east) clarifier is presently only used for additional settling.

Although the clarifiers appear to meet the detention time requirements for flocculation, the mixers are inoperative. Flocculation, therefore, fails to create the proper floc density for effective particle removal. New mixers are necessary to optimize flocculation and to improve floc density.

An estimate of the solids concentration in the reaction (mixing) zone of the primary (west) clarifier was determined to be 19 percent by volume. An estimate of the solids concentration in the reaction zone of the secondary (east) clarifier was not determined since its present function is only as a secondary sedimentation basin. Optimal flocculation treatment occurs when the reaction zone solids concentration is between 5 percent and 25 percent by volume. It appears that an optimal solids concentration is maintained in this clarifier for treatment. Based on the evaluation of flocculation treatment in the existing clarifiers, improvements are needed to optimize flocculation.

iii) Sedimentation

Optimization of Treatment for Disinfection Byproducts Control

31

The outer portion of the two up-flow clarifiers provides sedimentation for the treated water. The clarifier design characteristics are shown in Table 5.

The existing clarifier design characteristics appear to meet current recommended standards. Clarification and sedimentation detention time must be at least four hours based on current recommended standards. The combined detention time for sedimentation also exceeds the detention time recommended under current recommended standards (Recommended Standards for Water Works, (2003)). Each clarifier has an up-flow rate lower than current recommended standards. The maximum up-flow rate for clarifiers using softening treatment is 1.75 gpm/ft2 based on recommended standards. Up-flow rates for clarifiers using coagulation treatment only is 1.0 gpm/ft2 based on recommended standards using the present calculated up-flow rate of 0.66 gpm/ft2. The existing clarifiers currently meet the standards for coagulation treatment established in the recommended standards (Recommended Standards for Water Works, (2003)). The weir overflow rate (WOR) for clarifiers is established in Recommended Standards at not more than 20 gpm per linear foot of weir. Both the design and operational WOR are less than the recommended standards. The WOR for each clarifier was calculated to be 1.8 gpm/ft and 3.6 gpm/ft at design flow for parallel and series operations, respectively. These low flow velocities reduce the tendency for solids billowing and vortexing leading to excessive solids carryover from the basins. The settled solids concentration in the sludge withdrawal pipe from the primary (west) clarifier was found to be 100 percent by volume. Optimal treatment has been found to occur when this solids concentration is maintained between 70 percent and 95 percent by volume. It appears that settled solids are not removed from the clarifiers at an adequate

Optimization of Treatment for Disinfection Byproducts Control

32

rate. Improvements to the solids management practices and adjustment of the solids removal schedule are needed for this clarifier. Adjustments to the solids removal schedule for the secondary (east) clarifier were also found to be needed. The solids removal rate must be adjusted to match the solids production rate during treatment. Estimating the solids production rate could be added to the routine process control schedule. Results of the calculations can be used to establish solids removal rates for each clarifier. The sludge collector operating speed was measured as part of the evaluation. Based on measurements, the tip speed of the sludge collector was determined to be 2.8 feet/minute. Recommended Standards establish the operating speed of sludge collector mechanisms at no more than 10 feet/minute. The collector mechanism is operating at a speed significantly lower than permitted by current recommended standards. This desirable operating condition helps prevent solids billowing within the clarifiers. The existing clarifiers can be operated either in a parallel or series flow mode. Ohio EPA requires that both parallel and series operating conditions be available for solids contact clarifiers. Currently, the clarifiers are operated in series flow with treatment only being provided in the primary (west) clarifier. The secondary (east) clarifier is used only for secondary sedimentation. The water quality produced by this operation meets current drinking water standards. Compliance with anticipated drinking water standards should be achieved utilizing some improvements or modifications. Proposed clarifier improvements are discussed later. Settled water turbidity averages 1.0 NTU with maximum turbidity levels at approximately 2.3 NTU. Sedimentation should produce water having a turbidity level less than 5 NTU and preferably less than 3 NTU, placing the present resulting turbidity within acceptable parameters.

Optimization of Treatment for Disinfection Byproducts Control

33

Lime is applied to the water for softening. The average alkalinity concentrations typically produced in the treated water produces a total alkalinity and hydroxide alkalinity concentration of 53 mg/L and 20 mg/L, respectively. An optimum total alkalinity concentration for effective corrosion control is typically between 60 mg/L and 80 mg/L.

The hydroxide alkalinity concentration necessary for optimum softening treatment is a function of the raw water magnesium concentration which determines the lime dosage necessary for magnesium precipitation. Further discussion on optimum hydroxide concentrations is provided under softening treatment.

Enhanced coagulation requirements under the DBPR were previously described. Current TOC removal was determined to be approximately 37 percent. Although the minimum TOC removal requirements appear to be met using the current treatment, additional TOC removal is needed to achieve compliance with future THM requirements. However, compliance with enhanced coagulation and TOC removal requirements does not assure compliance with the maximum disinfection byproducts limits.

e) Recarbonation Basins

A recarbonation basin and carbon dioxide generating equipment were constructed as part of original plant. In 1993, the carbon dioxide generating system was replaced with a liquid carbon dioxide storage tank and a diffuser system.

The existing basin configuration has a mixing zone and a stabilization zone. The mixing zone provides a detention time of three minutes, while the stabilization zone provides a detention time of 31 minutes. Recommended Standards establishes a total recarbonation

Optimization of Treatment for Disinfection Byproducts Control

34

detention time of 20 minutes including a mixing detention time of at least three minutes at design flow. The existing recarbonation basin provides a detention time that exceeds current recommended standards.

The recarbonated water typically has average carbonate alkalinity and bicarbonate alkalinity concentrations of 24 mg/L and 25 mg/L, respectively. Optimal recarbonation for stability control should produce a water quality that is nearly equal in bicarbonate alkalinity and carbonate alkalinity concentrations. Based on this guidance, it appears that recarbonation of the settled water will produce adequate stability control. Water having carbonate alkalinity concentrations greater than approximately 35 mg/L will tend to plate out on surfaces including the filter sand grains causing growth of the grains. Filter sand growth can increase turbidity levels by allowing the smaller particles to pass through the filter.

Recarbonation treatment is discussed further under chemical feed applications in this section.

The existing recarbonation basin has no roof. Consequently, sunlight and warm water conditions foster the growth of algae in the basin. Algae growth creates a distinct green coloration in the treated water and may be producing additional organic contaminant concentrations in the water applied to the filters. These organics, if left untreated before chlorination, could contribute significantly to the disinfection byproduct (THM and HAA5) concentrations created by disinfection.

A roof should be constructed over the recarbonation basin to reduce algae growth during warm weather months. This roof would provide additional protection against birds and

Optimization of Treatment for Disinfection Byproducts Control

35

other small animals that could enter the basin and cause bacterial contamination of the treated water.

f) Filtration There are three rapid sand dual media filters to remove fine particles for final polishing treatment of the water. Each filter is provided with backwash and surface wash capabilities. Currently no equipment or piping is supplied for filter-to-waste operations. Filtration typically is the last barrier to remove contaminants that may be present in the treated water. Treatment ahead of the filters should remove a significant portion of the contaminants so that the filters can be operated effectively for particulate and microbial removal. Each of the filters has a surface area of 351.5 ft2, providing an approved treatment capacity of 1.0 MGD each at the current approved filtration rate of 2 gpm/ft2. Each filter bed consists of approximately 12-inches of gravel, 15-inches of sand, and 7.5-inches of anthracite. The existing filter media and gravel was installed in 1989. The filter bed was examined by probing and excavation to determine the media depth. This examination indicated that approximately 4.5-inches of anthracite had been lost since the media was installed. The approved filtration rate of the filters is 2 gpm/ft2. This is in accordance with current recommended standards. Current operating practices produce an average rate of approximately 1.5 gpm/ft2.

Optimization of Treatment for Disinfection Byproducts Control

36

Average filter run times are 86 hours, although the established maximum filter run time is 100 hours. Filter runs longer than 72 hours are desirable and usually indicate effective filtration. Excessively long filter runs (greater than 150 hours) can deposit too many solids deep into the media making it difficult to clean the filter media during backwash cycles. It may be possible to increase filter run times without adversely affecting filtered water quality. Any efforts to increase filter run times should be accompanied by careful monitoring of the primary filtered water quality parameters. Gross water production (GWP) is commonly used to help determine filtration efficiency and filterability of the treated water. GWP of approximately 10,000 gal/ft2 per filter run is accepted as indicating good filterability of the treated water. A typical GWP was determined to be approximately 7,840 gal/ft2 per run. Based on the GWP determined, the treated water appears to be filterable and the filters appear to be operating properly. Increasing the filter run times would increase the GWP. Filter performance can also be evaluated using filter efficiency calculations. These calculations determine the percentage of water filtered versus wash water used to clean the media. Filter efficiencies greater than 95 percent are accepted as indicating effective treatment. Using this method, the filtration efficiency was determined to be 98 percent. One backwash pump is used to supply wash water to the filters for cleaning. The backwash pump has a capacity of 5,200 gpm. The typical wash water usage is approximately 867,000 gallons per month or 1.8 percent of the monthly raw water treated. Normal wash water usage should average approximately two percent to four percent of the monthly raw water pumpage.

Optimization of Treatment for Disinfection Byproducts Control

37

The filters are currently backwashed at a rate of approximately 12.0 gpm/ft2. This rate is lower than the existing design backwash rate and current recommended standards for backwash. Records indicate that plant personnel do not change the backwash rate based on water temperatures. Some media loss has been noted based on data from the original filter media records. The cause of this media loss may be due to excessively high backwash rates during the winter months when the water is colder and therefore denser. As a general guideline, for each 1oC change in water temperature, the wash water rate should be adjusted by two percent. This would provide a higher flow as the water warms and lower flow as the water cools. Adjustment to the backwash rate should improve backwash efficiency. The backwash pump is capable of providing a maximum rate of 5,200 gpm which is approximately 15 gpm/ft2. The actual filter media expansion was measured using a bed expansion device during a typical backwash cycle. Results of the bed expansion test are given later in this section. The backwash rate should be established by direct measurement of the bed expansion at different water temperatures. The expansion of a filter bed should be at least 30 percent. The average backwash cycle uses approximately 54,400 gallons of water. As a guideline, a typical backwash cycle would range between 100 gallons and 150 gallons per square foot of filter area. Using this guideline, each filter backwash should use between 35,200 gallons and 52,700 gallons of wash water. Based on this, it appears that the filters are backwashed too long during a normal wash cycle. This excessive backwash duration may in part be caused by the backwash rate being too low (12.0 gpm/ft2 vs. 15.0 gpm/ft2) for proper cleaning of the media at certain water temperatures.

Optimization of Treatment for Disinfection Byproducts Control

38

The condition of the existing filter beds, current filter operation, and current backwash procedures were reviewed. The bed of Filter 2 was assumed to be representative of all of the filters. It was probed to identify the position of the existing gravel layers. Probing revealed that the gravel layers were intact with no significant mounding of the gravel. Some areas of the filter bed, however, did show some disruption of the gravel. Measurements of the gravel profile analysis revealed some areas of the filter bed had in excess of 2-inches difference in media depth. A difference in media depth of more than one-inch indicates potential problems with the gravel layer. Filter 2 was cored using a coring tube to examine the media for particle retention and backwash efficiency. Acid solubility testing and sieve analyses were performed to evaluate the filter media to determine if the media is suitable for continued use or will require replacement. Core samples revealed that particle retention was close to normal. Identical quantities of these core samples were washed with identical quantities of water. The turbidity of this water was measured to give an indication of the quantity of particles being retained in the media samples. Based on these turbidity measurements, particle retention was found to be greater in the first 2-inches of the media than would be expected. Particle retention turbidity measurements from below the top 15-inches of media should be less than 150 NTU/100 grams. The existing filters showed a particle retention measurement of less than this limit. A graph of the particle retention profile is shown in Figure 1. Core samples taken after backwash confirmed that the backwash efficiency was less than required. Turbidity measurements taken from the media following a typical backwash cycle show that the filter media retained a significant amount of particles in the top

Optimization of Treatment for Disinfection Byproducts Control

39

portion of the filter following the backwash cycle. As noted previously, the bottom media layers were found to retain few particles. The top layers of the media had turbidity measurements greater than the recommended 30 NTU/100 grams to 60 NTU/100 grams. Turbidity measurements in this range indicate the filter media to be sufficiently clean. Based on the turbidity profiles found, the filter examined is considered to be dirty at the top and too clean at the bottom. This indicates improvements are needed in backwash practices. Typical backwash profile criteria are shown in Table 5. Figure 2 is a graph of the backwash profile found. Current operating practice is to start the surface wash operation and then begin the backwash flow at 3,200 gpm. Each filter is washed for approximately two minutes at 3,200 gpm along with the surface wash. The backwash flow is then increased to approximately 4,200 gpm and continues for eight minutes. The surface wash is allowed to operate approximately 1.5 minutes into the high rate wash period and is stopped. After the eight-minute high rate wash, the backwash rate is reduced over three minutes from 4,200 gpm to zero gpm. Analysis of the wash water during a typical backwash period revealed that the backwash appears to be too long and is at a rate too low for proper cleaning of the media. It appears from the data that a shortened backwash cycle at a higher rate is needed to improve filter media cleaning. Turbidity analyses obtained during a typical backwash cycle are shown in Table 6.

Established operating criteria indicate that a backwash cycle is complete when the wash water turbidity falls below 10 NTU. It is generally accepted that once the wash water turbidity falls to 10 NTU, the backwash cycle should be terminated. Removing too many solids from the filter media increases the filter ripening period and the initial filtered water turbidity. The current backwash turbidities are reduced well below 10 NTU after a typical backwash period. Based on the backwash turbidities collected, it appears that the

Optimization of Treatment for Disinfection Byproducts Control

40

filter media is clean after approximately 4.5 minutes. Periodic evaluations can easily be made to adjust the backwash length as needed based on actual operating conditions.

A bed expansion test was performed during the filter inspection. During the test, the bed expansion was measured to be 20.8 percent. Based on this low bed expansion, a higher backwash rate is needed. A second test was performed at the maximum backwash rate (5,277 gpm or 15 gpm/ft2). Bed expansion measured at the higher flow rate was found to be 29 percent. The second test confirms that proper bed expansion can be achieved. An acid solubility test was performed on the filter media to identify the calcium carbonate deposition rate on the media. Calcium carbonate deposition increases the diameter of the media grains and can increase filter effluent turbidities and particle counts. Typical acid solubility should be no more than 2 percent per year. The acid solubility for the anthracite and sand media were found to be 0.8 percent and 0.3 percent, respectively. These values equate to 0.08 percent per year for the anthracite and 0.03 percent per year for the filter sand. Based on this analysis, it is apparent that recarbonation is being used properly and is providing stable water to the filters preventing excessive calcium carbonate deposition. Representative samples of the filter media were analyzed using sieve analysis to identify the current effective size (ES) and uniformity coefficient (UC) of the filter media. Sieve analysis results showed the anthracite media has an ES of 0.39 mm and a UC of 3.85. The 1989 specifications for the anthracite during media replacement indicated an ES of 0.93 and a UC of 1.60. The anthracite media no longer meets its original specifications. Sieve analysis results for the sand showed its ES to be 0.42 mm and its UC to be 2.06. The 1989 specifications for filter sand were an ES of 0.492 and a UC of 1.44. The sand media, like the anthracite, no longer meets its original specifications.

Optimization of Treatment for Disinfection Byproducts Control

41

Current Recommended Standards for filter media establish an ES for anthracite at 0.8 mm to 1.2 mm and a UC less than 1.65. Recommended Standards establishes an ES for filter sand at 0.45 mm to 0.55 mm and the UC to be less than 1.65. The existing filter media does not meet the recommended design criteria. The high uniformity coefficient indicates that a higher backwash rate is necessary to properly clean the media. Based on computer modeling, the sand and anthracite exhibit dissimilar backwash needs for proper bed expansion. The calculated backwash rate needed for existing anthracite is 45 gpm/ft2. The calculated backwash rate needed for the existing sand is 31 gpm/ft2. These backwash rates cannot be achieved using the backwash equipment and piping available. Replacement of the filter media is considered necessary at this time to meet turbidity regulations. Sieve analysis and acid solubility analysis should be preformed at least every two years to determine the condition of the filter media. If the ES of the sand changes more than 10 percent from design specifications, the media should be replaced. Media having a low uniformity coefficient should be considered for the next filter media replacement to reduce the backwash rate needed. This would more closely match the filter backwash requirements to the capabilities of the backwash pump. The Surface Water Treatment Rule (SWTR) establishes criteria for filter ripening to reduce turbidity and particle count spikes during the initial portion of a filter run. Filter ripening is a method used to properly condition the media in a freshly washed filter to achieve the lowest possible turbidity and solids in the water when a filter is initially placed in service. One of the ripening techniques is the use of filter-to-waste operations. Filter-to-waste is a procedure of wasting the first portion of water during a filter run until

Optimization of Treatment for Disinfection Byproducts Control

42

the effluent turbidity is within acceptable limits or goals. Current filter-to-waste goals establish that the turbidity in the filtered (wasted) water should be at or below 0.1 NTU after a 15-minute filter-to-waste period and should not result in turbidity spikes above 0.3 NTU.

A filter-to-waste simulation was performed by plant personnel following the initiation of a filter run. An evaluation of this showed that a filter-to-waste period of 15 minutes could not meet the turbidity goals for filter ripening.

The simulated filter-to-waste period showed that the turbidity did not spike above 0.2 NTU, but approximately 90 minutes was needed to reduce the filtered water turbidity to 0.1 NTU. Turbidity measurements taken during the simulated filter-to-waste period are shown in Table 7. It is apparent that the solids necessary for proper filter ripening are being removed during backwash.

Individual filter effluent turbidity meters are needed to optimize filtration. The Long Term 1 ESWTR requires continuous monitoring of individual filter effluents for turbidity.

Filter wash water currently is drained to a waste wash water wet well below the recarbonation basin. The wet well was installed as part of the original plant and has a capacity of approximately 46,000 gallons, which currently is less than one filter wash. Wash water is pumped back to the raw water reservoir for reuse. It is not recommended to recycle wash water for reuse unless proper treatment of the wash water is provided. Wash water typically contains microorganisms, suspended particles, and organic contaminants that can be 20 times greater than the raw water concentrations. Recycling these

Optimization of Treatment for Disinfection Byproducts Control

43

contaminants can lead to difficulties in treatment especially while trying to meet disinfection byproduct and enhanced coagulation requirements.

A significant amount of biological activity was observed in the filter media during the evaluation. It is unknown at this time what effect this biological activity has on filtered water quality (i.e. organics), however, any such contributions are believed to be undesirable. Periodic chlorination of the filter influent at a dosage of approximately 0.5 mg/L should control the biological activity and could improve overall filtered water quality. g) Transfer Pumping (Intermediate Pumping)

Three transfer pumps currently are used to transfer filtered water from the plant to the above ground storage tanks (clearwells). Each pump has a capacity of 1,050 gpm at a head of 42 feet. The firm capacity for transfer pumping is 2,100 gpm (3 MGD, with one pump out of service.) It appears that the existing transfer pumps are adequate to meet the current needs of the treatment plant.

h) Clear Well Storage (Finished Water Storage) and Disinfection

Disinfection of the filtered water is accomplished by adding chlorine to the water and allowing a period of contact for the chlorine to react with the contaminants.

Two above ground clearwells provide finished water storage and provide the necessary detention time (contact time). One 0.5 million gallon (MG) clearwell and one 1.0 MG clearwell currently are operated in series following chlorine addition. The clearwells provide a hydraulic detention time of approximately 12.2 hours at the plant design flow

Optimization of Treatment for Disinfection Byproducts Control

44

rate. The Ohio EPA has assigned an effective volume factor (EVF) of 0.25 to each clearwell. Typical EVF’s in Ohio range from 0.1 to 0.6. The EVF is based on the design of the clearwell and physical testing by tracer to determine how soon water delivered to the clearwells appears at the outlet. This factor determines the effective contact time in each clearwell for CT calculations. At the minimum clearwell operating level and maximum production rate, the clearwells have an effective detention time of 112 minutes.

Based on plant records of chlorine residuals and calculated CT values, both the minimum chlorine residuals are maintained and the required CT values are met. Historical data showed that there were no maximum chlorine residuals above the maximum level of 4 mg/L established in the DBPR.

Future changes in the regulations may increase the required CT values significantly for inactivation of Cryptosporidium (and/or other microbials). Additional clearwell storage is not needed at this time to meet CT requirements.

Design standards require that CT values for disinfection be maintained during clearwell cleaning and maintenance operations. This requirement is met when the clearwell still in service is operated near its maximum water level.

Baffling within the clearwells would reduce short circuiting to provide a larger EVF and aid in meeting future CT requirements that are expected to be more stringent. This would allow the future regulations to be met using the existing clearwells.

Water industry practices are to maintain a free chlorine residual in the finished water of at least 80 percent of the total chlorine residual. This practice prevents chlorinous tastes and

Optimization of Treatment for Disinfection Byproducts Control

45

odors in the drinking water caused by the partial oxidation of inorganic contaminants. The ratio of the free to the total chlorine residual in the water produced here currently averages 86 percent. The Ohio EPA requires that provisions for chlorination be available for treating the raw water, settled water, filtered water, and water entering the distribution system. There are currently chlorination points available for raw water and filtered water. New chlorine feed points are needed at the high service pump suction header and for the filter influent. These additional feed points will allow compliance with current recommended standards. i) High Service Pumping

Currently, there are three high service pumps and an auxiliary connection to provide wash water from the high service pump header. Existing pump capacities are shown in Table 8.

Based on Recommended Standards, the high service pumping is adequate for the rated plant capacity. Recommended Standards states that the high service pumps should be capable of pumping the rated plant capacity with the largest pump out of service. There also appears to be sufficient variation in pump capacities to allow a variety of pumping rates. It may be beneficial at some future date to install variable frequency drives to allow more flexibility in pumping rates. An additional high service pump may be needed to match pumping capacity to a future plant capacity increase. Space was provided in the original plant design to install a fourth high service pump. 3) Chemical Feed Systems

Optimization of Treatment for Disinfection Byproducts Control

46

The chemical feed systems are at the heart of any conventional water treatment plant. These systems have to be capable of delivering the chemicals not only in sufficient quantity but deliver them in very accurate dosages to the right location.

Recommended Standards specifies that at least a 30-day supply of chemicals be available in storage based on the average plant production rates. An examination of chemical usage records and the chemical storage facilities show that there is sufficient storage for all existing chemicals.

Recommended Standards also recommends that redundant or backup feed equipment be provided for each chemical feed system. Currently, only the lime feed system and chlorination systems have this redundancy.

A list of the chemical feed equipment, their capacities and application points is given in Table 9.

a) Potassium Permanganate At the time of this study no potassium permanganate feed system was installed. It may be beneficial to install a potassium permanganate feed system. Potassium permanganate is primarily used as a taste and odor control chemical as well as an oxidizer. Potassium permanganate is commonly used for the removal of taste and odor compounds that cannot be adsorbed by activated carbon. Potassium permanganate can also be used to oxidize various organic compounds and provide an oxygen bond between suspended particles for improved coagulation treatment.

Optimization of Treatment for Disinfection Byproducts Control

47

Previous to this study, the Village was granted approval by the Ohio EPA to install a potassium permanganate feed system. The system has not been installed and should be included in any plant improvements. Potassium permanganate is usually added ahead of carbon pretreatment and coagulation. It has been found that even small doses of permanganate can improve overall treatment and water quality. Jar testing covered in Section III will bear out the viability of potassium permanganate treatment. Caution must be exercised when feeding potassium permanganate so that the chemical is fully reacted prior to the application of activated carbon to avoided interference. Activated carbon and potassium permanganate feed points must not be close to each other. b) Powdered Activated Carbon

Carbon dosages applied to the raw water appear to be in the normal range for a surface water treatment plant. Carbon is typically applied for taste and odor control. Plant personnel stated that taste and odor complaints are uncommon indicating that the chemical dosages may be at the proper concentration for effective treatment of tastes and odors. Use of a more reactive carbon product (higher iodine number) was recently implemented for improved removal of taste and odor causing compounds.

It is believed that continuous carbon feed helps to control TOC concentrations which in turn control the formation of disinfection byproducts after chlorination.

Optimization of Treatment for Disinfection Byproducts Control

48

The current carbon feed point allows for minimal carbon contact time before other treatment chemicals are added. Shortly after carbon treatment, lime and soda ash are applied to the water, which quickly increases the pH. Carbon is most effective at lower water pH values (below pH 8.5).

A pretreatment basin should be considered in the future to provide sufficient carbon contact time for optimal treatment. Normally, a 15-minute to 30-minute contact time is recommended for carbon pretreatment.

c) Coagulant Feed System The current ferric chloride dosage appears to be sufficient for turbidity control based on the filtered water turbidity levels and filter run hours. The treated water results in a low turbidity filter effluent averaging about 0.09 NTU. Jar testing was performed to determine the coagulant type and dosage needed for optimum organics removal, particularly for DBP precursor control. It is expected that a much higher coagulant dosage may be needed for organics control to meet future drinking water standards. The most effective application point for coagulant chemicals is to the water at or near the rapid mix impeller. This provides the best dispersion into the water in the shortest amount of time. Ferric chloride is currently applied to the raw water at the in-line mixer upstream of the primary (west) rapid mix basin. The feed point should be relocated to the rapid mix basin and should discharge immediately above the mixer impeller.

Optimization of Treatment for Disinfection Byproducts Control

49

d) Lime Feed System Lime is applied to the water for softening. It raises the pH through the addition of hydroxide anions that convert the more soluble calcium bicarbonate to the less soluble calcium carbonate form allowing a portion of the calcium to precipitate. There are two identical lime feeders currently in operation. Both feeders are capable of supplying substantially more than the required dosage individually. Since the feeders are supplied by separate silos they are operated in rotation on approximately a 30 day cycle. As one silo goes empty, the other is placed in service until a new load of lime arrives. Lime is applied to the raw water inlet flume ahead of the primary (west) rapid mix basin for softening. Typically, lime and coagulant chemical applications should be separated to allow the coagulation reactions to proceed to completion before the higher pH chemical is added. The most effective treatment method is to add a coagulant to the water first to take advantage of the lower raw water pH for more effective treatment. Additionally, lime treatment can interfere with activated carbon treatment. The lime feed point should be relocated to the mixing zone of the clarifiers since lime does not need high intensity mixing for effective treatment like a coagulant. Relocation of the lime feed point also should improve coagulation by providing a more effective pH for coagulation reactions in the rapid mix basins. Minor lime feed system modifications and the construction of an eductor box will allow lime to be fed to the existing clarifiers where needed. e) Soda Ash (Sodium Carbonate) Feed System

Optimization of Treatment for Disinfection Byproducts Control

50

Soda ash aids in softening because the sodium has a lower electronegativity than either calcium or magnesium and will exchange with them on sulfate and chloride anions. This exchange results in the formation of calcium or magnesium carbonate, depending upon pH, which will then precipitate. The soda ash is presently fed at the same point as the lime, and as with the lime, this point should be relocated to the clarifier mixing zone. As with lime, soda ash does not need high intensity mixing for effective treatment making the clarifier mixing (reaction) zone an ideal location for its application. Minor soda ash feed equipment modifications and the construction of an eductor box would allow it to be fed to the existing clarifiers as needed. The eductor box would use water pressure to carry soda ash solution to each clarifier. A lime dosage of 139 mg/L is currently used for softening and appears to be too high based on chemical stoichiometric calculations. The current soda ash dosage of 7.8 mg/L appears to be too low based on these same stoichiometric calculations. Softening scenarios were modeled during jar testing (Section III, A, 4 Softening Chemical Dosages) and using a spread sheet program to identify the required lime and soda ash dosages needed to produce a desirable softened water quality. Typically the raw water primarily contains calcium bicarbonate hardness and magnesium non-carbonate hardness. Carbonate hardness is removed using lime. Non-carbonate, sulfate and chloride associated hardness, is removed using soda ash. Over treatment with lime first reduces alkalinity then the excess lime increases the alkalinity back to a desired concentration.

Optimization of Treatment for Disinfection Byproducts Control

51

A more desirable and economical treatment would be to use only enough lime to reduce the alkalinity to the desired level. However, in light of the necessity to utilize enhanced softening to induce magnesium precipitation, it is unlikely that this more economical method of treatment is feasible. Non-carbonate hardness removal is considered minimal based on the soda ash dosage. An increase in soda ash dosage would reduce the non-carbonate hardness and the calcium concentration in the settled water. However, care must be taken because calcium concentration could be reduced to undesirable levels if too much soda ash is applied. A balance is needed to optimize softening while maintaining the necessary calcium and magnesium concentrations for corrosion control. Two softening scenarios were looked at to help identify optimum lime and soda ash dosages. These scenarios are summarized in Table 10. f) Carbon Dioxide Feed System Carbon dioxide is applied to the settled water for recarbonation. Recarbonation reduces the water pH and adjusts (or aligns) the water alkalinity species for stabilization and corrosion control. Based on the water quality analyses performed, carbon dioxide is properly applied and mixed into the water for treatment. Plant operators hold the bicarbonate alkalinity to carbonate alkalinity ratio to about 1 to 1. This is approximately midway in the range for proper stability control which is from 0.7 to 1 to 1.4 to 1. Carbonate alkalinity is maintained below 35 mg/L in the water applied to the filters which provides for minimal lime particle carryover.

Optimization of Treatment for Disinfection Byproducts Control

52

Carbonate alkalinity concentrations greater than 35 mg/L can cause excessive calcium carbonate deposition (scaling) of filter media. Current stability tests (marble tests) performed by the operators show the treated water produced an average stability of -1 mg/L as alkalinity. This value indicates the water is minimally corrosive and is likely providing the low acid solubility analysis results observed for the filter media. Stability values between +5 mg/L as alkalinity demonstrate adequate stability control. The operators holding the average stability at -1 mg/L shows their attentiveness to the process that they understand the process. Based on stoichiometric calculations, the existing carbon dioxide dosage appears to be the proper dosage for optimal treatment. g) Phosphate Feed System Sodium hexametaphosphate solution is applied to the settled water flume to prevent excess scaling of the filter media and for corrosion control. The existing phosphate dosage seems to be at the proper concentration for treatment. Calcium carbonate deposition on the filter media is minimal and excessive scaling in the distribution system does not appear to be a problem. The typical phosphate residual (mg/L as phosphorus) in the plant tap is approximately 0.17 mg/L. No change to the phosphate feed is recommended at this time. Current corrosion control practices appear to be adequate based on the water quality in the distribution system and other indicators like the iron, lead, and copper concentrations in the system. The typical stability range for optimum treatment was previously discussed.

Optimization of Treatment for Disinfection Byproducts Control

53

A Langelier index (LI) and/or calcium carbonate precipitation potential (CCPP) are commonly used as corrosion control indicators and normally are much more accurate than the marble test. Recommended LI values for corrosion control range from approximately -0.5 to + 0.5. Recommended CCPP values range from 4 mg/L to 10 mg/L as CaCO3. LI and CCPP values determined were 0.38 and 2.4 mg/L, respectively. Based on the recommended LI values, it seems that the water also meets corrosion control guidelines for this parameter. The historic lead and copper results and the lack of rusty water complaints support this conclusion. CCPP values are currently at a level lower than that recommended, reflecting the slightly corrosive condition found by the marble test. Proper alignment of the water pH, calcium concentration, and alkalinity concentration would increase CCPP values to the recommended range. LI values and CCPP determinations should be included to adjust treatment and optimize corrosion control practices. h) Chlorine Feed System Chlorine currently is added to the transfer piping ahead of the 1.0 million gallon clearwell for primary disinfection. Based on CT calculations, it appears that effective disinfection and contact time is being provided.

A secondary chlorine feed is available to the raw water flume. For compliance with Recommended Standards, feed points should be installed for the settled water and for the high service pump suction header. Application of chlorine ahead of the filters could be useful for periodic control of biological activity in the filter media. Chlorine application

Optimization of Treatment for Disinfection Byproducts Control

at the high service pump suction header is needed to maintain minimum chlorine residuals in the plant effluent in the event of a problem with treatment.

54

Optimization of Treatment for Disinfection Byproducts Control

55

III) Jar Testing Jar testing was conducted to determine the most appropriate strategy to enhance treatment and as a result reduce the concentrations of organic compounds in the finished water. Jar testing is not capable of exactly duplicating the physical processes but does permit reasonably accurate modeling of those processes at a scale suitable for trending variations in treatment methods and chemical applications. A) Initial Phase (Establishing a Baseline) Initially the goal is to produce a baseline modeled after the present method of treatment in both chemistry and sequence of the chemical additions. Also great care is taken to duplicate detention times in the different treatment units. It is important to note here that to prevent anomalies; the chemicals used in the testing are taken from the supply present at the plant and assayed. This removes the possibility of errors from assumptions based on bills of lading and manufacturers’ assay reports. Additionally, it has been found that certain treatment chemicals that are purchased in bulk and held in storage for extended periods of time may vary substantially in strength from when they were delivered. In optimization work isolating deviations in actual chemical strengths from that which is assumed can be a substantial discovery. 1) Verification of Present Conditions and Treatment The data reviewed provides an average of the raw water conditions and overall quality of the water at each stage of the treatment process. It must be remembered that the average is simply a line drawn down the middle of the conditions found on all of those days

Optimization of Treatment for Disinfection Byproducts Control

56

looked at. In a surface water treatment plant, each day the raw water quality will be to some degree unique. Therefore, the dosages used must be fine tuned to suit the water quality found on the days the testing was conducted. The water from each jar is tested at each stage in the same manner as the water being treated in the plant. Once the results of the jar testing have been aligned with the water quality found at the end of each stage of the processes to be tested, variations can then be tested and the results of those variations relied upon with a high degree of confidence. The time elements for each treatment stage throughout the jar testing are based on those shown in Tables 5 where a dedicated treatment unit for that process is present. The initial chemical dosages in the first jar were based upon the averages show in Table 9 where that chemical was used. The initial dosages applied to the second jar were the chemical dosage being use in the plant that day. This was done for comparison purposes. Also of importance was the timing of the chemical additions. Care was taken to estimate the detention time in the flumes and piping where chemicals are added so the baseline testing closely reflects not only the dosage and sequence, but the timing between additions. a) Establish Rapid Mixer Energy Transfer Values The amount of mixing energy required for different chemicals and at different stages varies substantially. When a mixer is designed and installed specifically for a job, the manufacturer will provide a data sheet detailing the energy value the mixer is designed to impart.

Optimization of Treatment for Disinfection Byproducts Control

57

In the absence of such data the horsepower of the mixer can be used to approximate this value. However, the actual horsepower imparted by a mixer must be verified by measuring the amperage being drawn by the motor. Since one horse power is 746 watts of electricity and the amperage times the voltage gives wattage; the mixing energy can now be estimated. In situations where static mixing (use of turbulence in the water) is the only mixing available, the energy can be estimated by the hydraulic head loss across the restriction or weir. Here the horse power is (gpm x head) / 3960. The mixing energy is calculated from the velocity gradient between the active element (mixer) and the water. The unit of measure for this is the G-value. A search of the literature on this subject revealed that there is no one value that can be applied to all situations. However, there appeared to be a general consensus that the optimum value for coagulant rapid mixing resides around 1000 Gsec-1. The equation used to calculate this is G= (P/µV)1/2 Where: P = Power µ = Absolute viscosity V = Volume of the basin Algebraic manipulation of this equation allows any of the primary factors to be determined if the others are known or required values can be determine through an iterative process. Once the G value has been determined, it and experimental values can be closely approximated in the jar test unit by using the graph shown in Figure 3. b) Establish Flocculation Detention Time

Optimization of Treatment for Disinfection Byproducts Control

58

In a clarifier, during flocculation there is a gentle mixing from the turbulence created by the water flowing through the ports on the mixing zone drum. On a clarifier the flocculation zone is the area under what is referred to as the bell or skirt. This area is cone shaped and its diameter increases as it extends downward. This increase in area causes the downward velocity of the water is reduced proportional to the amount the diameter of the cone increases, reducing turbulence and aiding in the separation of floc particles from the water. Although cone shaped, it is not a true cone, but rather the portion of the cone referred to as the frustum of a cone. The detention time of the flocculation zone is given at 22 minutes in Table 2 when the clarifiers are operated in series. c) Establish Sedimentation Time The detention time for the sedimentation zone of one clarifier is given at 2.4 hours in Table 2 when the clarifiers are operated in series. However, the use of the second clarifier for only additional settling rather than in the manner it was intended cannot be factored in because the focus must be solely with the performance of a single unit when operated in the manner in which it was intended. The design of a clarifier provides for an upward flow in the sedimentation zone to take advantage of gravity. This upward flow of the water is counter to the movement of the particles being removed. Additionally, the shape of the cone surrounding the flocculation zone aids in sedimentation. This is because its diameter decreases as it goes higher up in the clarifier. This narrowing of the cone at the top increases the area of the sedimentation

Optimization of Treatment for Disinfection Byproducts Control

59

zone nearer the top and reduces the up-flow velocity allowing lighter floc particles to be removed. In jar testing the maximum up-flow velocity that can be tolerated and still maintain treatment can be estimated my timing the settling rate of the particles formed in the jar. This ability provides the ability to correlate jar settling rates to the clarifier up-flow rates. This is invaluable to evaluating modifications in chemical treatments. For proper particle settling the up-flow rate must be less than the particle rate of descent. Up-flow rates close to the particle settling rates tend to produce a blanket or plume at a midpoint in the sedimentation zone. This can cause the immediate discharge of particles from the clarifier with only minor increases in flow rates. 2) Pretreatment Chemical Dosages At this time the only pretreatment chemical being fed is powdered activated carbon (PAC). The PAC is injected into the raw water piping less than one minute ahead of the addition of the coagulant and lime. Additionally, there is very little mixing of the PAC into the process stream. Because adsorption is a physical process, mixing is required to bring the PAC particles into intimate contact with the contaminants allowing them to be adsorbed onto the active sites of the PAC particle. When fed, the average PAC dosage was found to be 2.9 mg/L (See Table 9). Because the target compounds of PAC treatment are organic compounds the effectiveness of PAC treatment is evaluated using UV254 analysis. During this initial phase of jar testing the PAC was added at the average dosage with minimal mixing for 45-seconds. The carbon was filtered out of the water and the filtrate analyzed with a similarly filtered sample of the raw water for comparison.

Optimization of Treatment for Disinfection Byproducts Control

60

As can be seen in Figure 3 there was only a minimal reduction in the UV254 reading obtained from the 45 second PAC treatment. During the experimental phase of the jar testing dosages from 1 mg/L to 6 mg/L were used with increasing contact times and mixing intensities to optimize this chemical’s properties. 3) Coagulant Dosages It was found that the operators seldom varied the coagulant dosage from the average; therefore this was used in the initial jar testing. The reasoning behind the maintenance of a constant coagulant dosage is due to the previous primary indicator of optimum treatment being based on finished water turbidity. Traditionally treatment plant operators would find a dosage that provided adequate treatment for turbidity, add a small amount extra to guard against intermittent excursions in raw water quality and leave it there. Recent changes in regulations requiring the control of organic compounds in the water changed the primary treatment indicator from turbidity to total organic carbon (TOC). In order to control TOC a process referred to as Enhanced Coagulation is most commonly used (Enhanced coagulation performance requirements, State of New York. (2006)). Shown in Table 9, this dosage is 32 mg/L. 4) Softening Chemical Dosages The most frequently varied chemical in a surface water treatment plant is the lime dosage. The reason for this is that the pH of the water in warmer months can vary significantly from day to day and during the course of the day. The reason for this

Optimization of Treatment for Disinfection Byproducts Control

61

variation is the intensity of the sunlight and its effect on the photosynthesis activity of the algae in the water. In the presence of sunlight the algae convert carbon dioxide to sugar and release excess oxygen increasing the pH. During the night or on days with low levels of sunlight available the algae use the dissolved oxygen and release carbon dioxide lowering the pH. Other problems encountered are caused by the variability of the lime when small quantities are mixed for application in jar testing. High pH compounds are subject to degradation from exposure to the atmosphere or from the small amounts of carbon dioxide present in the water used to make up the solution. Additionally, pebble lime has variable amounts of grit (sand) present that affect the overall purity of the sample being used. Therefore, all six jars are used and are dosed such that the average dosage should be present in the middle jar. Samples are taken from each jar and the dosage calculations are corrected based on the observed alkalinity and pH. The other softening chemical used is soda ash (sodium carbonate). Sodium carbonate is used because the sodium ion will exchange with the calcium and magnesium associated with the non-carbonate anions. These compounds are referred to as permanent or noncarbonate hardness. In this exchange the calcium and magnesium attach to the carbonate anion of the sodium carbonate allowing them to be precipitated by the lime. From day to day, the non-carbonate hardness varies only slightly from the average. It is most heavily impacted by the coagulant dosage. This is especially true when enhanced coagulation is used. Sodium carbonate is applied stoichiometrically for the amount of non-carbonate hardness that is to be removed and proportionately to the amount of coagulant used to counter the effects of the elevated coagulant dosages.

Optimization of Treatment for Disinfection Byproducts Control

62

In the initial tests the average dosage was applied yielding results within an acceptable distance of the water being treated in the plant that day. As shown in Table 9 the average dosage of sodium carbonate is 7.8 mg/L. 5) Simulated Recarbonation (pH Adjustment) Simulated recarbonation is performed by adding dilute sulfuric acid to the softened water to adjust the pH and alkalinity prior to performing the filterability test. Sulfuric acid is used in place of carbon dioxide due to the impracticality of dissolving precise amounts of gaseous carbon dioxide into the water on a bench scale. As shown in Table 9, the average carbon dioxide dosage is 17.6 mg/L. Both sulfuric acid and carbon dioxide carry two hydrogen ions. Therefore, a sulfuric acid dosage can be calculated based on molecular weights. This also allows the quality of the water to be as close to that being applied to the filters in the plant as is possible on a bench scale. Filterability is tested by comparing the time required for equal amounts of a sample of treated water and distilled water to pass through a 0.45 μm glass fiber filter. The ratio formed by these filtration times is the filterability index. This allows the quality of the water being produced under test conditions to be directly compared to the quality of the water being produced by the plant. 6) Phosphate Dosages Phosphates are added to the water to inhibit the corrosive effects of the water on the metal components of the treatment plant and the distribution system. It is not possible to

Optimization of Treatment for Disinfection Byproducts Control

63

analyze the effects of corrosion and the need for a specific amount of phosphorous in short term bench scale testing. Therefore, this chemical cannot be evaluated in this study. 7) Disinfection and Simulated Distribution System Testing Simulated distribution system testing is used to estimate the potential of the water produced in the jars to form THM and HAA5 compounds. Since the levels of these compounds are tested for on a quarterly basis by the operators, it can be assumed that water produced in the jars of similar quality to that produced by the plant will have a similar THM formation potential. Additionally, during the experimental phase of the jar testing this allows the formation potential of the water produced in the jars using different treatment strategies to be directly compared to that produced by the treatment plant with a relatively high level of confidence. The method used to accomplish this is to take the water produced in the jars and disinfect it with liquid chlorine (bleach) at the same dosage being applied in the treatment plant. The water is then placed in sealed glass containers. The containers are placed in a dark location with an ambient temperature similar to that of the water in the distribution system and allowed to remain there for 72 hours. B) Experimental Phase (Chemical) During this phase the dosages of the chemicals presently in use were varied and chemicals not presently in use were tested. Curves showing the effect these variations had on the primary parameters and in particular the levels of organic compounds were developed from those tests. 1) Modification to Chemical Dosages

Optimization of Treatment for Disinfection Byproducts Control

64

Figure 5 shows a correlation between the results of UV254 analyses and the amount of Total Organic Carbon (TOC) present in the same sample. This correlation is possible because research has shown that there is a direct relationship between the amount of light absorbed at the wave length of 254 nanometers and the concentration of specific organic compounds present. This correlation was necessary to be able to produce useful results from UV254 analyses of the water. This is useful because UV254 samples can be taken and analyzed immediately in the plant laboratory. This allows the plant staff to make changes in treatment and be able to quantify the effect of changes in both the raw water and the treated water. Figure 6 illustrates the relationship found between the percentage of UV254 reduction and the amount of coagulant applied to the water. Although this chart does not allow the precise determination of a chemical dosage, it does illustrate that a significant portion of the organic compounds present in the raw water can be removed by enhanced coagulation (elevated dosages of coagulant). However, this chart would allow the operators to estimate an initial coagulant dosage on a given day based on the characteristics of the raw water. Figure 7 presents a correlation between coagulant dosage, pH and the percentage of UV254 reduction. This chart clearly indicates a direct correlation between pH levels greater than 10.7 and the TOC removal efficiency. Research has shown that this may be in part due to the production of magnesium hydroxide in the water during softening. (Mustafa M. Bob, B.S., M.S. (2003)) a) Pretreatment Chemical Dosages

Optimization of Treatment for Disinfection Byproducts Control

65

PAC was the only pretreatment chemical being fed at the time of this study and it was being injected into the raw water pipe approximately 45 seconds ahead of the application of the coagulant and the softening chemicals. In the initial phase of the jar testing the average PAC dosage was used at the calculated contact time. To determine an adequate contact time this dosage was used with six different contact times from 10 minutes to 60 minutes. It was found that UV254 readings steadily decrease up to 30 minutes but no discernable decreases were seen at times greater than 30 minutes. Using this information dosages from 3 mg/L to 13 mg/L were applied to the jars. The results of this can be found in Table 8. Here it can be seen that a dosage of 7 mg/L in contact with the water for 30 minutes produced the greatest reduction in the UV254 reading. This same procedure was repeated using potassium permanganate. The results of this can be found in Table 9. Here dosages of 1.5 mg/L to 9 mg/L were used. From that table it can be seen that a dosage of 4.5 mg/L yielded the best UV254 reduction. With both of the treatment chemicals it can be seen that a curve was formed as previously discussed indicating a point of most effective treatment. Here dosages above and below that point were noticeably counter productive. Also as previously discussed, the contact time reached a sharp point of diminishing return at 30 minutes. i) Sequencing of Chemical Additions Since these two chemicals act in completely different ways it was assumed that the sequence of their addition may have an impact on their effectiveness. To determine this, two jars were set up using these optimal dosages. On the first jar the carbon was added first and on the second jar the potassium permanganate was added first followed by the other chemical at 30 minutes.

Optimization of Treatment for Disinfection Byproducts Control

66

Figure 10 shows the effect proper dosages of PAC and potassium permanganate when their additions are properly sequenced. This should be compared with the UV254 reading in Figure 4 where the baseline using PAC treatment only for 45 seconds was established. The results of this showed that potassium permanganate followed by PAC provided a level of treatment approximately 20 percent better than PAC followed by potassium permanganate. The assumption here is that the potassium permanganate conditioned the organic compounds by oxidation. This oxidation modifies the structure and charge on the molecules and appears to make certain compounds more amenable to adsorption by the PAC. b) Coagulant Dosages Coagulant dosages were used that varied from 45 mg/L to 95 mg/L. These dosages were tried in both single stage and two stage (split) treatment. Figures 11 and 12 show the results of these trials. There were two treatment methods tested during these trials. The first was to simulate the sequential application of the coagulant and the softening chemicals in one basin. The use of the second basin for settling only would be simulated by extending the settling time. The second was to simulate splitting the treatment between the two basins. In the first basin only the coagulant would be applied for treatment. In theory the majority of the colloidal matter and larger organic compounds would be removed in the first (west) clarifier. This would eliminate their presence when the softening chemicals are added, thereby eliminating the possibility of them engaging in undesirable reactions with the softening chemicals.

Optimization of Treatment for Disinfection Byproducts Control

67

In the second basin a second smaller dose of coagulant would be applied to aid in floc formation for turbidity control prior to the application of the softening chemicals. The softening chemicals would then be applied at levels sufficient for Enhanced Softening where the precipitation of magnesium compounds is forced (Mustafa M. Bob, B.S., M.S. (2003)). i) Single Basin Treatment It can be seen in Figure 11 where all of the treatment would take place in one basin and at a maximum coagulant dosage of 95 mg/L the UV254 was only reduced to 0.055. There was no apparent curve to the effectiveness of the treatment. The curve does indicate that the effectiveness of treatment did increase directly with the coagulant dosage. However, when coagulant dosages reach these levels the coagulant can have a significant impact on the aesthetic quality of the water by the addition of chlorides and non-carbonate hardness. Additionally, these dissolved solids and non-carbonate hardness tends to make the water more corrosive (Edwards, M., Scardina, P. (2006)). Although this would be considered an effective treatment for organics, the resulting finished water would not likely meet the aesthetic standards of the final consumer. ii) Split Treatment Split treatment pertains to the operation of the clarifiers and is an extension of the concept of providing each treatment chemical a separate place to perform its intended purpose. To accomplish this in the treatment plant, the clarifiers are operated in series. The design of the treatment plant has the elevation of the west clarifier 18 inches above the east clarifier. Additionally, the flumes connecting the clarifiers were designed with a

Optimization of Treatment for Disinfection Byproducts Control

68

double pass configuration. This design allows the raw water flow to be either split equally between the clarifiers or to be sent to the west clarifier and the effluent of the west clarifier sent to the inlet of the east clarifier. In split treatment only coagulant is applied to the water in the first (west) clarifier after pretreatment. The operation of the clarifier would be optimized around the solids production under these conditions. At the second (east) clarifier a dosage of coagulant sufficient to initiate coagulation would be applied at the east rapid mixer. The floc produced from this coagulation will provide a sturdy substrate for the precipitates formed during softening to cling to, helping to reduce the effluent turbidity of the clarifier. In the mixing chamber of the second (east) clarifier the softening chemicals, lime and soda ash would be applied. In order to simulate this in jar testing, each jar was allowed to settle after the initial treatment with coagulant. The clear water was decanted from the jars, the settled solids removed and the decanted water placed back in the jars. A second dose of coagulant was then applied prior to the softening chemicals. This procedure allowed the jar test to be as closely representative to the full scale process as possible. It can be seen from the curve in Figure 12 that at all coagulant dosages the resulting UV254 reading was lower than the reading obtained with the highest coagulant dosage use in the single basin trial (See Figure 11). This bears out the theorized effectiveness of split treatment. The curve formed indicates that optimum treatment under these conditions was obtained at a coagulant dosage of 65 mg/L. Another observation can be made here. In split treatment the optimum treatment was obtained at a lower coagulant dosage meaning that overall chemical costs will also be lower. In addition to a lower coagulant cost, the need for extra lime and soda ash used to

Optimization of Treatment for Disinfection Byproducts Control

69

remove the additional non-carbonate hardness produced by the coagulant is also reduced proportionately. Additionally, there is a secondary cost reduction. This is from the costs associated with the handling and disposal of the solids that would have been produced by the additional coagulant, lime and soda ash. Although not often viewed as a direct cost of treatment, the storage of these solids requires large tracts of land and is often subject to discharge permitting where the water decanted from the solids is routed to a public waterway. c) Softening Chemical Dosages The dosage of softening chemicals used was based upon the amount of alkalinity that was desired in the finished water. Allowances were made for the additional coagulant used and the non-carbonate hardness that was added by the coagulant. In normal precipitative softening the lime is applied at a ratio of approximately 1.0 mg/L per each 1.25 mg/L of hardness expressed as calcium carbonate that is desired to be removed. (Operator Training Committee of Ohio. (OTCO). (2007)). When enhanced softening is used, lime is applied at a dosage substantially greater than that normally used. In enhanced softening the lime is applied at a dosage that will remove the calcium hardness down to its lower solubility limit of approximately 25 mg/L (depending upon temperature). At this point the pH rises to pH 10.5 to pH 10.9 where the magnesium alkalinity is converted to magnesium hydroxide and precipitation of the magnesium occurs.

Optimization of Treatment for Disinfection Byproducts Control

70

After softening to this level it is necessary to restore the alkalinity to the water. This is accomplished by adding additional lime at the previously noted ratio to obtain the desired finished water alkalinity after recarbonation. All of the softening chemical trials were initially performed using the same coagulant dosage of 45 mg/L in all jars simulating the west clarifier and 20 mg/L to simulate the dosage to be applied to the east clarifier. This was considered the best approach because the dosage to the east clarifier was high enough to produce the level of coagulation desired but not so high as to mask the effects of the higher softening chemical dosages. In Figure 13 it was found through serendipity when a jar was accidentally overdose with lime the pH where magnesium precipitation occurred and the effect this precipitation had on the organic content of the water became visible. The effects of this unintentional increase in the pH early in the jar testing align with the research of Bob M. Mustafa (Mustafa M. Bob, B.S., M.S. (2003)) on this subject. In his research Mr. Mustafa learned that as the magnesium hydroxide in the water was incrementally increased to the saturation level, the organic carbon content fell. Further trials were conducted based on the elevated pH until it was determined that a pH of 10.9 provided optimum treatment. When the pH was elevated significantly above pH 10.9 it was found that the level of TOC removal decreased. It is theorized that saponification of a portion of the organic compounds was occurring. d) Disinfection, Simulated Distribution System Testing A simulated distribution test was performed on the water from all six jars of the final trial using the method previously described. The results of this test are shown in Table 11.

Optimization of Treatment for Disinfection Byproducts Control

71

The data presented in Table 11 indicates that compliance with the MCL for TTHM’s of 80 micrograms per liter is achievable using the treatment method employed in the jar testing. 2) Compilation and Comparison The goal of the jar testing was to determine if it was possible to remove enough of the naturally occurring organic matter (NOM) from the water to meet the MCL for TTHM contaminants. The data presented in Table 11 and Table 12 present a comparison of the TTHM results obtained from simulated distribution system testing of the water treated under experimental conditions in the jars (Table 11) and the TTHM results obtained from the historical data at the treatment plant (Table 12). It could be assumed with a high degree of confidence that if the treatment strategy used in the jar test trials was to be applied to the full scale operations of the treatment plant, a significant improvement in the resulting distribution system TTHM results should be expected. 3) Modifications to Unit Processes During optimization work the simulation of different equipment configurations and sizes can be experimented with. This experimentation allows for the identification of problems in individual processes and to determine if the addition or removal of specific processes and associated infrastructure are needed. a) Addition of Reaction (Contact) Basins

Optimization of Treatment for Disinfection Byproducts Control

72

The jar test trials revealed that the application of both PAC and potassium permanganate proved beneficial. Additionally, it was determined that each of these chemicals had optimum dosages, an optimum application sequence in relationship to each other and required adequate time to react. As the PAC was being used, there was insufficient mixing to disperse it throughout the raw water stream and insufficient time for it to react prior to the application of other treatment chemicals that rendered it ineffective for optimum TOC removal. It was also determined that the application of potassium permanganate was beneficial at an optimum dosage with an adequate reaction time. To accomplish this, new basins had to be designed to provide both adequate detention time and sufficient mixing to assure the chemicals were dispersed throughout the process stream. A chemical feed system for potassium permanganate was installed. Two basins were constructed to allow for the sequential application of PAC and potassium permanganate to the raw water prior to the water entering the treatment plant. Each basin provides mixing of approximately 100 G and a 30 minute detention time. b) Modifications to Rapid Mix Basins Required The rapid mix basins were redesigned to provide the characteristics previously discussed in Section II, A, 2 c. The cross sectional area of each basin was reduced from 16 feet square to 9.0 feet square. This substantial reduction in the cross sectional area combined with more aggressive mixing reduces the possibility of water passing through the basin without being treated and assures that some of the mixing energy will be transferred to all of the water in the vicinity of the mixer. The increased velocity of the water flowing

Optimization of Treatment for Disinfection Byproducts Control

73

through the basin transports the water away from the region of the greatest turbulence prior to the formation of floc particles, averting the chance of those particles being sheared within the basin. Moving the coagulant injection point to a location just above the eye of the impeller combined with more aggressive mixing assures that the coagulant is properly dispersed throughout the process stream. c) Separation of Treatment between Clarifiers The separation of treatment is providing a vessel where each treatment chemical may be applied and provided with sufficient time in that vessel to fully react. During the jar test trials the traditional method of treatment being employed at the treatment plant of performing both coagulation treatment and softening treatment in the same basin was simulated and compared to the potential gains seen by utilizing the capabilities designed into the treatment plant and separating coagulation treatment from softening treatment. As was previously explained in Section III, B, 1, b, ii, the treatment obtained by optimizing coagulation treatment in the west basin followed by optimizing softening treatment in the east basin provided an exceptional advantage over the method traditionally used. To fully utilize this method all of the operating parameters of each basin and the chemical dosages were optimized based upon the results of UV254 analyses of the effluent from each stage of treatment. Chemical dosages, sludge recirculation rates, waste sludge removal rates and mixing intensities were varied individually until the greatest reductions in UV254 absorbance were obtained. This procedure initially required substantial operator involvement but once the indicating water quality characteristics were identified; operations at these points became routine.

Optimization of Treatment for Disinfection Byproducts Control

74

However, it must be noted that the allowable deviation from these points became substantially smaller and treatment less tolerant of variations in raw water quality. d) Modifications to Sludge Recirculation Piping on Clarifiers A solids contact clarifier his referred to as such because a portion of the previously precipitated material is reintroduced to the reaction zone. The type of clarifier used here is a Walker Process unit. Walker Process clarifiers use a small pump that draws suction from the sludge removal piping and transfers it to the reaction zone. When the treatment plant was originally constructed the recirculated sludge was delivered to the inlet side of the rapid mix chambers. This method of recirculation led to many operational and water quality problems. Of these was that the sludge had a very high pH and is highly precipitative. This led to significant scale build up in the basins, flumes and on the mixer impellers. Operationally, the sludge was routed through the high intensity mixer which sheared the large heavy particles into small colloids. The problems caused by this were significant enough to lead the operators to abandon recirculation. As was found by Mr. Mustafa (Mustafa M. Bob, B.S., M.S. (2003)) there was an optimum level of treatment obtained by closely controlling the concentration of recirculated solids in the reaction zone. This was also found to be true here. In order to obtain this the sludge recirculation pipe discharge was removed from the influent flume and relocated to the clarifier mixing zone. Here the recirculated solids are gently mixed with the newly formed solids providing more complete usage of any unreacted chemical and provides a substrate for the newly formed particles to cling to.

Optimization of Treatment for Disinfection Byproducts Control

75

e) Control of Solids in the Clarifier Slurry Pool and Reaction Zone Previous to this study there were no efforts made to monitor or optimize the levels of solids recirculated or retained in the clarifiers. Recirculation had been abandoned because of its undesired effects on water quality and treatment and the slurry pool was removed in its entirety every day eliminating the presence of solids for recirculation. However, once the physical modifications to the clarifier were completed, these parameters were noted by the operators as having a significant impact on treatment. As detailed in the OTCO training manual on solids contact clarifier operation, the solids concentration in both the slurry pool (sludge zone) and the reaction zone of the clarifier can be regulated by measuring the solids concentration volumetrically with a 100 ml graduated cylinder. (Operator Training Committee of Ohio. (OTCO). (2004). Solids Contact Clarifiers Optimizing Performance. (2004)). Optimizing these concentrations through regular blow-off cycles and recirculation rates are key to proper operation of the clarifiers. Normally it will be found that optimum operation will be achieved with a concentration of solids in the slurry pool between 75 percent and 90 percent by volume and concentrations in the reaction zone between 10 percent and 20 percent by volume. The exact concentrations necessary will vary seasonally and with raw water quality.

Optimization of Treatment for Disinfection Byproducts Control

76

IV) The Human Factor It must be realized that the treatment plants are operated by people and that optimizing the physical treatment plant can only be as effective as the understanding the people operating it have of the process. (Cavalier, Julia (2007)) Simply stated: If it is not possible to get those responsible for the operation of the treatment plant to implement the findings of the study, nothing has been gained. The Human Factors fall into three primary categories: 1) Education and Training 2) Attitude 3) Taking ownership Education and training removes the mystery surrounding what is being done. It establishes clear goals and provides both the knowledge and the means for the operators to achieve those goals. This knowledge provides them with the ability to monitor and maintain the process. This in turn gives the operators confidence in their decision making abilities. Taking ownership is in essence taking responsibility. All three of these factors are closely interdependent. The biggest obstacle found with the operators is that they often do not understand the relevance of many of the changes being proposed and/or made.

Optimization of Treatment for Disinfection Byproducts Control

77

One of the major complaints heard when going into a plant is that the engineers tend to shroud what they are doing in mystery and evade questions. In many instances the engineers will limit their interaction with personnel to those in management only. This in and of itself tends to alienate the operators and establishes an adversarial situation with the staff members most responsible for the day to day operation of the plant processes. Throughout this study every effort was made to get the operators involved at every step. Working from the assumption that by nature, most people are curious; and that curiosity can be harnessed. Through this involvement, the operators were not only able to observe the techniques and procedures being employed, but encouraged to participate and make inquiries. To aid them materials such as the OTCO training manuals and links to most of the references used in this paper were provided. Though this process the work done here not only proved to be successful, but sustainable.

Optimization of Treatment for Disinfection Byproducts Control

78

V) Summary A) Results of Optimization The results of the optimization work conducted here are shown in Table 13. The values presented here are averages of the historical data reviewed during the study and averages of the same parameters during the following year after completion of the improvements and implementation of the findings from the study. The values of greatest significance are those of the THM Formation Potential (TTHMFP) and the measured THM concentrations found in the distribution system. The TTHMFP is measured by dosing a sample of the water with 100 mg/L of hypochlorite ion (Standard Methods for the Examination of Water and Wastewater. (2005)). A dosage of this strength is provided to eliminate the possibility of the hypochlorite ion as the limiting reactant. This value represents the maximum level of THM contaminants that can be formed from the organic matter present in the water. From the table it can be seen that the present TTHMFP (125 mg/L) is not significantly higher than the running annual average of the sample sites before the study (120 mg/L). This reduction in organic matter allowed the running annual average of the same sample sites to be reduced to 39 mg/L. 1) Summary of Changes a) Unit processes Changes to the unit processes are the changes made to the physical plant itself.

Optimization of Treatment for Disinfection Byproducts Control

79

Addition of pretreatment basins Two pretreatment basins were installed hydraulically between the reservoir and the clarifier influent channels. Each basin provides 30 minutes of detention time at the treatment plant rated capacity of 3.0 MGD. Each basin is equipped with a mixer capable of providing 100 G of mixing energy to the water at the coldest anticipated temperature. The configuration of the piping and valves serving the basins is such that the basins can be operated in series, parallel or singularly. Rapid Mixer Modifications The cross sectional area of each basin was reduced from 16 feet square to 9.0 feet square. This change accomplished several goals. The first was to reduce the volume of water under the immediate influence of the impeller. The influence of the impeller (velocity of the turbulence) decreases proportionally as the distance from it increases. This assured that no water entering the basin could pass through without becoming entrained in the turbulence created by the impeller The coagulant injection point was relocated to just above the eye of the impeller. A radial type of impeller was selected because it creates a large vortex near its eye. The coagulant is entrained in this vortex, mixed thoroughly with the water which is accelerated off the tip of the impeller blade at a relatively high velocity assuring that the coagulant will be carried to walls of the basin and mixed with the water passing through prior to being able to react. Clarifier Modifications Mixers used to circulate water between the flocculation zone and the mixing zone to provide low G-value turbulence for the mixing zone were repaired and placed back into service.

Optimization of Treatment for Disinfection Byproducts Control

80

Sludge recirculation piping discharge was moved from the influent flume ahead of the rapid mixers to the mixing zone of the clarifiers. The gate valves in the influent flume were reconfigured to produce a series flow through the clarifiers with the west clarifier being the first in the series. Piping was added to allow softening chemicals to be added to the mixing zones instead of the influent channels. Recarbonation A roof was constructed over the recarbonation basin supported by side walls that have no windows. This modification removed sunlight penetration into the basin and will prevent the growth of algae. This in turn will eliminate food sources for bacteria that may have survived the previous treatment processes. Clearwell Baffling Partition curtains were added to the clearwells to prevent the water entering them for disinfectant contact time and storage from finding a short circuit path to the high service pump suction piping. b) Chemical Treatments Coagulant The coagulant dosage was increased from 32 mg/L to a combined dosage between the two clarifiers of 65 mg/L. This dosage is variable based upon the level of treatment the operators deem necessary to produce the desired quality of finished water.

Optimization of Treatment for Disinfection Byproducts Control

81

Lime The lime dosage did not substantially change, however is also variable to that required to produce the desired 10.9 pH during softening. Soda Ash The soda ash dosage increased proportional to the increase in coagulant to offset the non-carbonate hardness added by the coagulant. This dosage will be varied directly with increases or decreases in the coagulant dosage. Potassium Permanganate Potassium permanganate treatment was added. The optimal dosage of potassium permanganate was found to be 4.5 mg/L for the water being tested that day. The operators regularly check the demand for potassium permanganate and adjust the dosage as required. The effectiveness is measured through UV254 analysis. PAC PAC treatment was reinstituted and moved to the pretreatment basins. The optimal dosage was found to be 7 mg/L for the water being treated that day. As with the potassium permanganate, the operators regularly monitor the effectiveness of the chemical and adjust the dosage accordingly.

Optimization of Treatment for Disinfection Byproducts Control

82

VI) Looking Forward The intent of this paper was to demonstrate that a conventional surface water treatment plant can produce water that will meet both regulated water quality standards while still being aesthetically pleasing to those who consume the water. The techniques employed in this study were based on the findings of the authors of the references listed and the works of many others whose works these studies evolved from. These techniques were modified as required to suit the individual characteristics of the treatment plant being studied. Some of the techniques were modified and those modifications need to be published when correlations are further developed to expand the knowledge base for those responsible for supplying us with a continuous safe supply of drinking water.

Optimization of Treatment for Disinfection Byproducts Control

83

References Basuray, Sagnik and Chia Chang, Hsueh, (2007). Center for Microfluidics and Medical Diagnostics. Department of Chemical and Biomolecular Engineering: University of Notre Dame. Notre Dame. Indiana 46556. USA. Received 15 March 2007; Published 25 June 2007. Retrieved April 1, 2008 from: http://www.nd.edu/~changlab/Sagnik.pdf Cavalier, Julia (2007). Disinfection Byproducts Optimization. The Human Factor. – State of North Carolina. Retrieved April 1, 2008 from: http://www.deh.enr.state.nc.us /pws/AWOP/DisinfectionByproducts Optimization. pdf Chloral Hydrate in Drinking Water. Prepared by the Federal-Provincial-Territorial Committee on Drinking Water. (February 21, 2007). Retrieved April 1, 2008 from: http://www.hcsc.gc.ca/ewhsemt/alt_formats /hecssesc/pdf/pubs/watereau/consultation/chloral_hydrate/chloral-hydrate_e.pdf Edwards, M., Scardina, P. (2006). Enhanced Coagulation Impacts On Water Treatment Plant Infrastructure. AWWA Research Foundation (2006) Enhanced coagulation performance requirements. Required Removal of TOC by Enhanced Coagulation for Surface Water. Systems or Systems Using Ground Water Under the Direct Influence of Surface. Water Treatment Plants that Use Conventional Filtration Treatment. State of New York. (2006). Retrieved April 1, 2008 from:

Optimization of Treatment for Disinfection Byproducts Control

84

http://w3.health.state.ny.us/dbspace/NYCRR10.nsf/56cf2e25d626f9f7852565 38006c3ed7/7e4a0ab54a78f54085256e9000674b54?OpenDocument Everything You Want To Know About Coagulation and Flocculation. Zeta-Meter Inc. Fourth Edition. Zeta-Meter Inc., 765 Middlebrook Ave. Staunton, VA. 24402 Formation of Chlorinated Organics in Drinking Water (2005). Istanbul (Turkey) and Salerno (Italy), L. Rizzo 1 1: University of Salerno, Department of Civil Engineering; H. Selcuk 2 84084 Fisciano: (SA). Italy; A. Nikolaou 3: Pamukkale University, Environmental Engineering Department; V. Belgiorno 1 KinikliPamukkale. Turkey; M. Bekbolet 4 3: University of the Aegean, Department of Environmental Studies; S. Meric 1: * Water and Air Quality Laboratory, University Hill, 81100 Mytilene, Greece, 4; Bogazici University, Institute of Environmental Sciences, 0815 Bebek (Istanbul), Turkey, Retrieved April 1, 2008 from http://www.gnest.org/journal/Vol7_No1 /7_1_Rizzo-317_95-105.pdf García, Indiana (2005). Removal of Natural Organic Matter By Enhanced Coagulation In Nicaragua. Department of Chemical Engineering and Technology. Royal Institute of Technology, Stockholm, Sweden, December (2005). Retrieved April 1, 2008 from: http://www.diva-portal.org/diva/getDocument? urn_nbn_se_kth_diva-586-2__fulltext.pdf Gaulinger, Siegfried. (2007) Coagulation Pre-Treatment for Microfiltration with Ceramic Membranes. (2007). TECHNEAU is an Integrated Project Funded by the European Commission under the Sustainable Development, Global Change and Ecosystems Thematic Priority. Retrieved April 1, 2008 from: http://www.techneau.org/fileadmin/files/Publications /Publications /Deliverables/D2.3.2-1.pdf

Optimization of Treatment for Disinfection Byproducts Control

85

Heinonen, Pekka and Pisto, Sannimaria -Lopez (2007). Optimization of Chemical Water Treatment Process For Water Treatment Plants. Tempere (2007). Tampere Polytechnic, Environmental Engineering. Retrieved April 1, 2008 https://oa.doria.fi/bitstream/handle/10024/7027/Heinonen.Pekka.PistoLopez. Sannimaria.pdf?sequence=1 Hill, Fred (2007). Total System Optimization. State of North Carolina. Retrieved April 1, 2008 from: http://www.deh.enr.state.nc.us/pws/AWOP/TotalSystemOptimization.pdf Mustafa M. Bob, B.S., M.S.(2003) Enhanced Removal Of Natural Organic Matter During Lime-Soda Softening. April 1, 2008 from: http://www.ohiolink.edu/etd/sendpdf.cgi/Bob%20Mustafa%20M.pdf?osu1047486107 Operator Training Committee of Ohio. (OTCO). (2004). Solids Contact Clarifiers. Optimizing Performance. (2004). Operator Training Committee of Ohio, Inc. 3972 Indianola Av. Columbus, OH 43214-3158. Operator Training Committee of Ohio. (OTCO). (2007). Precipitative Softening. OperatorTraining Committee of Ohio, Inc. 3972 Indianola Ave. Columbus, OH 43214-3158. Pascal Roche, Benanou, David. (2006). Impact Of Chlorination On The Formation Of Odour Compounds And Their Precursors In Treatment Of Drinking Water.: TECHNEAU, TECHNEAU is an Integrated Project Funded by the European Commission under the Sixth Framework Programme, Sustainable Development,

Optimization of Treatment for Disinfection Byproducts Control

86

Global Change and Ecosystems Thematic Priority. Retrieved April 1, 2008 from: http://www.techneau.org/fileadmin/files /Publications/Publications/Deliverables/D5.3.8.pdf Perišio, M. (2006). NOM and Arsenic Removal from Natural Water by Enhanced Coagulation. E-Water Official Publication of the European Water Association (EWA) (2006). Retrieved April 1, 2008 from: http://www.ewaonline.de/journal/2006_08.pdf Prashant, Kumar (2003). Trihalomethanes - Causes, Effects and Remedy. (Entry No. 2003 CEV 0025). Retrieved April 1, 2008 from: http://people.pwf.cam.ac.uk/pp286/TRIHALOMETHANES%20%20CAUSES,%20EFFECTS%20AND%20REMEDY.pdf Recommended Standards for Water Works, (2003) Retrieved April 1, 2008 from: http://www.dutchessny.gov/countygov/departments/health/reports/hd10stateprefac e.pdf Standard Methods for the Examination of Water and Wastewater. (2005). 21st Edition: American Public Health Association. (2005). Method 5910B UV 254 Absorbance.. United States Geological Survey Web Site http://waterdata.usgs.gov/oh/nwis/monthly/?referred_module=sw&site_no=04189 000&por_04189000_3=707503,00060,3,1923-10,200709&format=html_table&date_format=YYYY-MMDD&rdb_compression=file&submitted_form=parameter_selection_list

Optimization of Treatment for Disinfection Byproducts Control

87

Weeks, Daniel P., PhD (2003).. Pushing Electrons, A Guide for Students of Organic Chemistry, 3rd Edition, Saunders College Publishing. Pg. 36. Homolytic covalent bond breakage. World Health Organization (2004). Disinfectants and Disinfectant By-Products. First draft prepared by G. Amy: University of Colorado, Boulder, Colorado, USA; R. Bull, Battelle Pacific Northwest Laboratory, Richland, Washington, USA; G. F. Craun, Gunther F. Craun and, Associates, Staunton, Virginia, USA; R.A. Pegram, US, Environmental Protection Agency, Research Triangle Park, North, Carolina, USA; and M. Siddiqui, University of Colorado, Boulder, Colorado, USA, This web versions have been updated to incorporate the corrigenda published by November 30, 2004. Retrieved April 1, 2008 from: http://whqlibdoc.who.int/ehc /WHO_EHC_216.pdf .

Optimization of Treatment for Disinfection Byproducts Control

Appendices Abbreviations and Acronyms Tables Figures

88

90 92 107

Optimization of Treatment for Disinfection Byproducts Control

Acronyms and Abbreviations AWWA

American Water Works Association

CCPP

Calcium Carbonate Precipitation Potential

cfs CT

cubic feet per second Concentration times contact Time

DIC

Dissolved Inorganic Carbon

DBP D/DBPR ES EPA EVF ft2 ft3 gpd gph gpm gps HAA5 IOC’s LI LCR MCL MG MGD mg/L MRDL NA NC NOM NPDES NTU ODNR PAC pC/L SCADA SOC’s SWTR TCR TDS THM THMFP TOC

disinfection byproduct Disinfection/Disinfection Byproducts Rule Effective Size Environmental Protection Agency Effective Volume Factor square Feet cubic Feet gallons per day gallons per hour gallons per minute gallons per second Halo-aceticacid Inorganic Chemicals Langelier Index Lead and Copper Rule Maximum Contaminant Level million gallons Million Gallons Per Day milligram per liter Maximum Residual Disinfectant Level Not Applicable no current MCL exists Naturally Occurring Organic Matter National Pollutant Discharge Elimination System Nephelometric turbidity units Ohio Department of Natural Resources powdered activated carbon picoCurie per Liter Supervisory Control and Data Acquisition Synthetic Organic Chemicals Surface Water Treatment Rule Total Coliform Rule Total Dissolved Solids Trihalomethane THM formation potential Total Organic Carbon

89

Optimization of Treatment for Disinfection Byproducts Control

UC USGS VOC’s WOR µg/L

Uniformity Coefficient United States Geologic Survey Volatile Organic Chemicals Weir Overflow Rate micrograms per liter

90

Optimization of Treatment for Disinfection Byproducts Control

91

Tables Table 1 Average Water Quality

93

Table 2 Existing Clarifier Design Characteristics (Each Clarifier At Design Flow)

95

Table 3 Filter Specifications

96

Table 4 Existing River Pumps

97

Table 5 Typical Backwash Profile Criteria

98

Table 6 Turbidity Measurements From A Typical Backwash Cycle

99

Table 7 Turbidity Measurements From A Simulated Filter-To-Waste Period

100

Table 8 Existing High Service Pumps

101

Table 9 Existing Chemical Feed Applications

102

Table 10 Potential Results Of Softening Adjustments On Recarbonated Water Quality

103

Table 11 Results Of Simulated Distribution Test

104

Table 12 Average Distribution System TTHM Results

105

Table 13 Water Quality Before And After Optimization

106

Optimization of Treatment for Disinfection Byproducts Control

TABLE 1 AVERAGE WATER QUALITY All results are mg/L concentrations unless noted otherwise. Parameter Raw Water Finished Water o Temperature C 16 NA pH, s.u. 8.6 8.8 Turbidity, NTU 5.3 0.09 Total Alkalinity (as CaCO3) 157 38 Phenol Alkalinity (as CaCO3) 14 3 Hardness (as CaCO3) 277 165 Calcium (as CaCO3) 177 123 Magnesium (as CaCO3) 100 42 Nitrate 2.0 2.0 Fluoride 0.48 1.01 Phosphate (as P) NA 0.17 Free chlorine NA 1.85 Total chlorine NA 2.15 Trihalomethanes 0.319 (2) 0.082 Total Organic Carbon (TOC) 6.5 N/A Haloacetic Acids NA 0.021 Arsenic NA <0.003 Antimony NA <0.003 Barium NA 0.033 Beryllium NA <0.005 Cadmium NA <0.005 Chromium NA <0.005 Cyanide NA <0.0005 Mercury NA <0.0002 Nickel NA <0.01 Selenium NA <0.003 Sulfate NA 99 Thallium NA <0.001 Gross Alpha (3) NA <4 (3) Gross Beta NA 7 Total dissolved solids 350 260 Langelier Index 0.91 0.38 CCPP, (as CaCO3) (4) 19.2 2.4 (5) DIC, (as C) 37.1 8.8 (1)

Maximum Contaminant Level (MCL) for drinking water

92

MCL (1) NC NC 0.5 NC NC NC NC NC 10 4.0 NC 4.0 4.0 0.100 NC NC 0.05 0.006 2.0 0.004 0.005 0.100 0.2 0.002 0.1 0.05 NC 0.002 15 50 NC NC NC NC

Optimization of Treatment for Disinfection Byproducts Control

(2)

THM formation potential (THMFP) picoCurie/L radiological analysis (4) Calcium Carbonate Precipitation Potential (5) Dissolved Inorganic Carbonate, as carbon NA - not available NC - no MCL currently exists (3)

93

Optimization of Treatment for Disinfection Byproducts Control

TABLE 2 EXISTING CLARIFIER DESIGN CHARACTERISTICS (EACH CLARIFIER AT DESIGN FLOW) Characteristic Parallel Flow Volume, gallons 347,000 Flocculation volume, gallons 46,900 Sedimentation volume, gallons 300,100 Flocculation time, minutes 45 Sedimentation time, hours 4.8 Up-flow rate, gpm/ft2 0.44 Weir overflow rate, gpm/ft 1.8 Detention time, hours 5.6 Total flocculation time, minutes 45 Total sedimentation time, hours 4.8 Total detention time, hours 5.6

Series Flow 347,000 46,900 300,100 22.5 2.4 0.88 3.6 2.8

94

Optimization of Treatment for Disinfection Byproducts Control

95

TABLE 3 FILTER SPECIFICATIONS Number of Filters Dimensions: Surface Area: Rated Capacity: Gravel: (depth) Sand: (depth) Effective Grain Size (ES) (mm) Grain Size Uniformity Coefficient (UC)

3 18.5 ft. x 19 ft. (each, dual cells) 351.5 ft2 (each) 1.0 MGD (each at 2 gpm/ft2) 12-inches 16.5-inches 0.42 2.06

Anthracite: (depth) Effective Grain Size (ES) (mm) Grain Size Uniformity Coefficient (UC)

7.5 inches 0.39 3.85

Average filter run time:

86 hours

Backwash Pump: Backwash Rate:

5,277 gpm at 25 ft. TDH 15.0 gpm/ft2 at maximum flow

Optimization of Treatment for Disinfection Byproducts Control

TABLE 4 EXISTING RIVER PUMPS Pump Capacity, gpm 1 2,300 2 2,400 3 4,600

Horsepower 90 40 88

Head, ft. 42 not available 44

96

Optimization of Treatment for Disinfection Byproducts Control

TABLE 5 TYPICAL BACKWASH PROFILE CRITERIA Turbidity, NTU/100 grams Filter Media Condition 30 to 60 Clean 60 to 120 Slightly Dirty >120 Dirty >300 Mud ball problem >2,000 Extremely Dirty

97

Optimization of Treatment for Disinfection Byproducts Control

98

TABLE 6 TURBIDITY MEASUREMENTS FROM A TYPICAL BACKWASH CYCLE Time Interval Turbidity, NTU Time Interval Turbidity, NTU 0.5 minutes 35 5.5 minutes 5.3 1.0 minutes 33 6.0 minutes 4.7 1.5 minutes 43 6.5 minutes 3.9 2.0 minutes 63 7.0 minutes 3.6 2.5 minutes 14 7.5 minutes 3.1 3.0 minutes 14 8.0 minutes 2.6 3.5 minutes 13 8.5 minutes 2.4 4.0 minutes 12 9.0 minutes 2.1 (1) 4.5 minutes 7.3 9.5 minutes 5.0 minutes 7.9 10.0 minutes (1) Indicates the end of the backwash period

Optimization of Treatment for Disinfection Byproducts Control

99

TABLE 7 TURBIDITY MEASUREMENTS FROM A SIMULATED FILTER-TO-WASTE PERIOD Time interval Turbidity, NTU Time interval Turbidity, NTU 1 minute 0.12 11 minutes 0.18 2 minutes 0.12 12 minutes 0.18 3 minutes 0.12 13 minutes 0.18 4 minutes 0.12 14 minutes 0.18 5 minutes 0.12 15 minutes 0.18 6 minutes 0.12 16 minutes 0.18 7 minutes 0.13 17 minutes 0.19 8 minutes 0.15 18 minutes 0.20 9 minutes 0.16 19 minutes 0.20 10 minutes 0.18 20 minutes 0.19

Optimization of Treatment for Disinfection Byproducts Control

TABLE 8 EXISTING HIGH SERVICE PUMPS Pump Capacity, gpm Horsepower 1 700 60 2 1,400 100 3 2,100 150 Total* 2,100 * Total capacity with the largest pump out of service.

Head, feet 167 167 167

100

Optimization of Treatment for Disinfection Byproducts Control

TABLE 9 EXISTING CHEMICAL FEED APPLICATIONS Feed System Typical Dosage, mg/L Activated carbon 2.9 Ferric chloride 32.0 Lime 139 Soda ash 7.8 Carbon dioxide 17.6 Fluoride Sodium hexametaphosphate Chlorine

0.62 0.64 4.72

101

Application Point Raw water ahead of flume Raw water at in-line mixer Raw water flume Raw water flume Recarbonation basin mixing chamber Settled water flume Settled water flume Clearwell influent piping

Optimization of Treatment for Disinfection Byproducts Control

TABLE 10 POTENTIAL RESULTS OF SOFTENING ADJUSTMENTS ON RECARBONATED WATER QUALITY (all results in mg/L concentration except as noted) Current Softening Parameter Treatment Run 1 Water pH, s.u. 9.14 Total alkalinity, as CaCO3 49 Phenol alkalinity, as CaCO3 14 Total hardness, as CaCO3 221 Calcium, as CaCO3 203 Magnesium, as CaCO3 18 Bicarbonate alkalinity 21 Carbonate alkalinity 28 Solids production, gpd 15,100 CCPP (1) 7.8 Langelier Index 1.04 Lime dosage 139 Soda ash dosage 7.8 Carbon dioxide dosage 17.8 Probable treatment costs (2) $113/day (1) Calcium carbonate precipitation potential (2)

Run 2 9.27 52 13 219 126 92 26 26 9,750 9.7 0.95 69 21 15.2 $90/day

Costs include average chemical and solids handling costs

102

Softening Run 3 9.25 56 13 207 95 112 30 26 10,200 9.4 0.83 61 41 15.7 $113/day

Optimization of Treatment for Disinfection Byproducts Control

TABLE 11 RESULTS OF SIMULATED DISTRIBUTION TEST COAG. DOSAGE 15 25 35 45 TOC, filt 3.8 3.5 3.4 3.1 MCAA 1.28 1.20 1.18 1.10 DCAA 12.95 11.86 11.12 10.35 TCAA 14.16 12.70 11.73 10.74 MBAA 0.06 0.05 0.05 0.05 DBAA 0.08 0.08 0.08 0.08 THAA5 28.5 25.9 24.2 22.3 pH 9.44 9.44 9.44 9.44 Bromide 0.005 0.005 0.005 0.005 o Water C 10 10 10 10 Chlorine 3.1 3.1 3.1 3.1 UV254 0.059 0.054 0.050 0.047 t(hrs)HAA 3 3 3 3 t(hrs)THM 102 102 102 102 TTHM4 49.2 45.6 43.5 40.7 TTHM MCL = 80

55 3.2 1.12 9.96 10.23 0.05 0.08 21.4 9.44 0.005 10 3.1 0.044 3 102 40.1

103

65 3.2 1.12 9.46 9.60 0.05 0.07 20.3 9.44 0.005 10 3.1 0.041 3 102 38.8

Optimization of Treatment for Disinfection Byproducts Control

104

TABLE 12 AVERAGE DISTRIBUTION SYSTEM TTHM RESULTS Location Location Average DATE 1 2 Location 3 Location 4 Per Date 1/14/2002 99.2 4/22/2002 137 7/17/2002 196.0 196.0 10/18/2002 128.0 128.0 1/15/2003 98.2 98.2 4/14/2003 55.9 58.0 78.1 103.0 73.8 7/17/2003 122.0 148.0 126.0 142.0 134.5 11/19/2003 71.5 95.2 102.0 118.0 96.7 AVERAGE 83.1 100.4 102.0 130.9 120.4

Optimization of Treatment for Disinfection Byproducts Control

TABLE 13 WATER QUALITY BEFORE AND AFTER OPTIMIZATION Parameter Before Study Settled Water TOC 4.1 Settled Water UV254 0.094 TOC Reduction (%) 43.8 UV254 Reduction (%) 37 Turbidity (NTU) 0.09 Finished TOC 2.6 Finished UV254 0.045 Overall TOC Reduction (%) 49 Overall UV254 Reduction (%) 63 THM Formation Potential (mg/L) 368 THM (mg/L) 120.4

105

After Study 2.4 .034 59 64 .04 1.9 0.021 74 93 125 39

Optimization of Treatment for Disinfection Byproducts Control

106

Figures Figure 1 Filter #2 Particle Retention Profile

108

Figure 2 Backwashed Filter Particle Profile

109

Figure 3 Jar Test G Value Graph

110

Figure 4 Initial Carbon Baseline Application

111

Figure 5 UV254 Reading vs. TTHM

112

Figure 6 UV254 Reading vs. Ferric Chloride Dosage

113

Figure 7 UV254 Reading vs. pH

114

Figure 8 UV254 Reading vs. PAC Dosage

115

Figure 9 UV254 Reading vs. Potassium Permanganate Dosage

116

Figure 10 UV254 Reading with Optimum Carbon and Permanganate Dosages

117

Figure 11 UV254 Reading Single Stage Ferric Chloride Treatment

118

Figure 12 UV254 Reading Two Stage (Split) Ferric Chloride Treatment with Enhanced Softening

119

Figure 13 UV 254 vs pH

120

Optimization of Treatment for Disinfection Byproducts Control

107

FIGURE 1 FILTER #2 PARTICLE RETENTION PROFILE 0 0 5 10 15 20 BED DEPTH, inches

25 30 35

NTU/100 grams 50

100

150

200

Optimization of Treatment for Disinfection Byproducts Control

108

FIGURE 2 BACKWASHED FILTER PARTICLE PROFILE NTU/100 grams 0 0 5 10 15 20 BED DEPTH, inches

25 30 35

20

40

60

80

100

120

140

160

180

Optimization of Treatment for Disinfection Byproducts Control

FIGURE 3 JAR TEST G VALUE GRAPH (Phips-Bird Jar Test Machine Operators Manual)

109

Optimization of Treatment for Disinfection Byproducts Control

110

FIGURE 4 INITIAL CARBON BASELINE APPLICATION CHEMICAL DATA

Ferric Chloride

Polymer

NaOH

KMnO4

Carbon

Specific gravity Percent dry chemical

100

Chemical added, mL or grams

2 grams

Stock concentration, mg/L

2,000

Solution volume, mL TEST CONDITIONS Rapid Mixing

STIRRER RPM 100

DURATION 45 seconds

Flocculation Settling

0

Filtered distilled water time Temperature °C

pH

1,000 Simulated Conditions G - 100 sec-1

minutes

G-

sec-1

minutes

Coag.

gpmft2

seconds

Soft.

gpm/ft2

RAW WATER CHARACTERISTICS Turbidit Alkalinity Hardness y Color

TOC

DOC

T HMF P

UV254

Filtered 0.185

POC

8.5 Filterability Index

Calcium

Magnesiu m

Iron

Manganes e

0.238 JAR NUMBER Raw water volume, mL Carbon dosage, mg/L

1

2

2,000

2,000

3

TOC, mg/L UV254, cm-1 Filtered water time Filterability Index NOTES : Baseline carbon application

0.16

3 2,00 0

4

5

6

2,000

2,000

2,000

Optimization of Treatment for Disinfection Byproducts Control

111

FIGURE 5 UV254 READING vs. TOC

TOC, mg/L

CORRELATION BETWEEN UV254 AND TOC

6.0 5.8 5.6 5.4 5.2 5.0 4.8 4.6 4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8 2 y = 533.61x - 19.295x + 3.005 2.6 2 2.4 R = 0.9921 2.2 2.0 1.8 0.025 0.030 0.035 0.040 0.045 0.050 0.055 0.060 0.065 0.070 0.075 0.080 0.085 0.090 0.095 -1

UV254, cm

Optimization of Treatment for Disinfection Byproducts Control

112

FIGURE 6 UV254 READING vs. FERRIC CHLORIDE DOSAGE UV254 vs FERRIC CHLORIDE DOSE 0.085 0.080 y = -1E-06x2 - 0.0003x + 0.0904

0.075

R2 = 0.9835

0.070 0.065 0.060

UV254, cm-1

0.055 0.050 0.045 0.040 0.035 0.030 0.025 0.020 0.015 0.010 0.005 0.000 35

45

55

65

75

85

95

105

115

125

135

Ferric Chloride Dosage, mg/L

145

155

165

175

185

195

Optimization of Treatment for Disinfection Byproducts Control

113

FIGURE 7 UV254 READING vs. pH UV254 vs. pH ENHANCED SOFTENING 0.055

0.050

3

2

y = 0.0458x - 1.459x + 15.474x - 54.549 2

R = 0.9668

UV254, cm-1

0.045

0.040

0.035

0.030

0.025

0.020 10.0

10.2

10.4

10.6

10.8 Water pH

11.0

11.2

11.4

Optimization of Treatment for Disinfection Byproducts Control

114

FIGURE 8 UV254 READING vs. PAC DOSAGE Carbon Dosage vs UV254

y = 4E-05x3 - 0.0007x2 + 0.0042x + 0.0958 R2 = 0.9946

0.114

0.112

UV254

0.11

0.108

0.106

0.104

0.102 0

2

4

6

8 Carbon Dosage

10

12

14

Optimization of Treatment for Disinfection Byproducts Control

115

FIGURE 9 UV254 READING vs. POTASSIUM PERMANGANATE DOSAGE y = 9E-05x3 - 0.001x2 + 0.0038x + 0.0993

UV254 vs KMnO4 Dosage

R2 = 0.9946

0.114 0.112

UV254

0.11 0.108 0.106 0.104 0.102 0

1

2

3

4

5 KMnO4 Dosage

6

7

8

9

10

Optimization of Treatment for Disinfection Byproducts Control

116

FIGURE 10 UV254 READING WITH OPTIMUM CARBON AND PERMANGANATE DOSAGES

UV 254 Optimum PAC and Permanganate y = 6E-08x3 - 1E-05x2 - 0.0002x + 0.1149 R2 = 0.9957

0.1 0.09 0.08 0.07 UV254

0.06 0.05 0.04 0.03 0.02 0.01 0 Time

Optimization of Treatment for Disinfection Byproducts Control

117

FIGURE 11 UV254 READING SINGLE STAGE FERRIC CHLORIDE TREATMENT UV 254 vs Ferric Chloride Dosage

3

2

y = -3E-07x + 6E-05x - 0.0055x + 0.2354 2 R = 0.9944

0.1 0.09 0.08 0.07 UV254

0.06 0.05 0.04 0.03 0.02 0.01 0 0

10

20

30

40

50 Ferric Dosage

60

70

80

90

100

Optimization of Treatment for Disinfection Byproducts Control

118

FIGURE 12 UV254 READING TWO STAGE (SPLIT) FERRIC CHLORIDE TREATMENT WITH ENHANCED SOFTENING

Split Treatment UV254

3

2

y = -8E-08x + 2E-05x - 0.0019x + 0.0827 2 R = 0.9357

0.0365 0.036 0.0355

UV254

0.035 0.0345 0.034 0.0335 0.033 0.0325 0.032 0.0315 0

10

20

30

40

50

60

Ferric Dosage

70

80

90

100

Optimization of Treatment for Disinfection Byproducts Control

119

FIGURE 13 UV 254 vs pH UV 254 Reduction vs. pH and Coagulant Dosage 60.00

10.85 10.8 10.75

40.00

10.7

30.00

10.65 10.6

20.00

10.55 10.00

10.5

0.00

10.45 30

32

34

36

38

40

Coagulant Dosage mg/L Ferric Dose

pH

42

44

46

pH

Percent UV 254 Reduction

50.00