|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001790611
007 cr mnu|||uuuuu
008 070622s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001517
Aguinaldo, Jorge T.
Precipitative softening and ultrafiltration treatment of beverage water
h [electronic resource] /
by Jorge T. Aguinaldo.
[Tampa, Fla] :
b University of South Florida,
ABSTRACT: Lime softening, chlorination, clarification and filtration have been long recognized treatment processes for beverage water specifically the carbonated soft drink (CSD) because it provides consistent water quality required for bottling plants, however these processes are becoming uneconomical and causes more problems than the benefits they offer. These processes require very large foot print, occupy large plant volume, and generate large volume of sludge which causes disposal problems. Chlorination produces trihalomethanes (THMs) and other by-products which are detrimental to health and imparts tastes to the final products. Using the newly developed submerged spiral wound ultrafiltration membranes in conjunction with lime softening may replace the conventional lime softening, clarification and filtration processes. This research was conducted to demonstrate the feasibility of integrating immersed ultrafiltration (UF) membrane with lime softening. The objectives of thi s research was to achieve the water quality required by the CSD bottlers; determine the relationships of operating parameters such as pH and membrane flux with trans-membrane pressure (TMP), and membrane permeability; determine the optimum dosage of lime; evaluate the operating parameters as basis for the design and construction of the full scale plant; and predict themembrane cleaning intervals. A pilot unit consisting of lime reactor and UF system was designed and built for this research. The pilot unit was operated at various pH ranging from 7.3 to 11.2 and at membrane flux rates of 15, 30 and 45 gfd. The pilot unit was also operated at the CSD bottler's operating conditions which is pH 9.8 at flux of 30 gfd. The pilot unit operated for a total of 1800 hours. The raw water source was from city water supply. The filtrate from the pilot unit achieved alkalinity reduction to 20 to 30 mg/L preferred by CSD bottlers, with lime dosage close to the calculated value. The filtrate turbidity^ during the test was consistently within 0.4 to 0.5 NTU. The TMP values obtained during the test ranges from 0.1 to 2.5 psi, while the permeability values ranges from 18.19 to 29.6 gfd/psi. The increase in flux results to corresponding increase in TMP, and increase in operating pH, increases the rate of TMP. Permeability decreases with increasing operating pH. The TOC reduction ranges from 2.6 % to 15.8% with increasing operating pH. No scaling of the UF membranes was observed during the test. Thirty days UF membrane cleaning interval was predicted. The results from this research can use as the basis of designing and operating a full scale Lime Softening UF Treatment Plant.
Thesis (M.A.)--University of South Florida, 2006.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 60 pages.
Adviser: Robert P. Carnahan, Ph.D.
Membrane water treatment.
Carbonated soft drinks.
x Environmental Engineering
t USF Electronic Theses and Dissertations.
Precipitative Softening and Ultrafiltration Treatment of Beverage Water by Jorge T. Aguinaldo A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Environmental Engineering Department of Civil and Environmental Engineering College of Engineering University of South Florida Major Professor: Robert P. Carnahan, Ph.D. Marilyn Barger, Ph.D. Daniel H. Yeh, Ph.D. Date of Approval: April 5, 2006 Keywords: lime softening, membrane water tr eatment, carbonated soft drinks, alkalinity reduction, immersed membrane Copyright 2006, Jorge T. Aguinaldo
Acknowledgements I would like to thank Dr. Robert P. Carnahan, Ph.D, my Major Professor who are among the few who believed in my idea, supported and encouraged me throughout my research. I would like to thank and acknowledge my committee members, Dr. Marilyn Barger, Ph.D and Dr Daniel H.Yeh, Ph.D for the valuable time they spent reviewing my thesis a nd providing very important comments and suggestions. I would like to thank the mana gement, and staff of Doosan Hydro Technology Inc.: Mr. Ali Kalantar, CEO; Mr Del Martinez, Direct or of Sales, who shared his experiences on lime soda softening and my coordinator with CSD Bottlers; Mr. Dalmacio Aviso, Manufacturing Manage r, who built the pilot unit; Mr. Carlo Canezo, Engineering Trainee who helped in the manufacture and operation of the pilot unit; Ms. Sarah Vinson, Technology & De velopment Coordinator, who helped in formatting the manuscript; and Dr. Silv ana Ghiu, Ph.D, Process & Development Engineer, who reviewed the final manuscr ipt and provided valuable suggestions. I would like also to thank Mr. Ben Goulds and Mr. Mi chael Snodgrass, both from Trisep Corporation for the information a nd data they provided on the Immersed Spirasep UF membrane.
i Table of Contents List of Tables iii List of Figures iv Abstract v Chapter One: Introduction 1 Chapter Two: Background 7 2.1 Lime Softening 7 2.2 Limitations/Problems Associated With Lime Softening 9 2.3 Precipitative Softening 10 2.4 Ultrafiltration 14 Chapter Three: Materials and Methods 16 3.1 Experimental Plan 16 3.2 Pilot Lime Softening Ultrafiltration Unit 17 3.2.1 Lime Reactor 17 3.2.2 SpiraSep Ultrafiltration Membrane 18 3.2.3 Pilot Lime Softening Ultrafiltration Control Description 19 3.3 Chemicals 26 3.4 Experimental Procedures 27 3.5 Analytical Procedures 29 3.5.1 pH and Temperature 30 3.5.2 Alkalinity 30 3.5.3 Calcium and Magnesium Hardness 30 3.5.4 Turbidity 30 3.5.5 Total Suspended Solids 31 3.5.6 Total Organic Carbon 31 Chapter Four: Results and Discussions 32 4.1 Initial Operating Conditions Without Chemical Addition 32 4.2 Operation at Varying pH and Flux 34 4.3 Operation at CSD Bottler Plant Conditions 39
ii Chapter Five: Summary and Conclusions 42 5.1 Alkalinity Reduction 42 5.2 UF Filtrate Turbidity 42 5.3 Trans-membrane Pressure (TMP) vs. pH and Flux 43 5.4 Permeability 43 5.5 Total Organic Carbon (TOC) 45 5.6 Hardness Reduction 46 5.7 Operating Flux 46 5.8 Chlorination 47 5.9 Benefits of the Lime Softening Ultrafiltration (LSUF) Process to CSD Bottler 47 References 49 Appendices 51 Appendix A: Pilot Unit Equipment Description 52 Appendix B: SpiraSep Trans-membrane Pressure (TMP) Measurements 58
iii List of Tables Table 1 Selected Contaminants Limits from the National Primary and Secondary Drinking Water Standards (EPA, 2003) 3 Table 2 CSD Bottlers Water Quality Survey 4 Table 3 Canadian Water Quality Guidelines for Carbonated Beverage 5 Table 4 Raw Water Analysis 34 Table 5 Average TMP Values Before and After UF Backflushing at Various Flux Values 35 Table 6 Flux vs. Permeability at Various Operating pH 35 Table 7 Analysis of Water Samples at Various Operating Conditions 38 Table 8 Analysis of the Filtrate by CSD Bottler 40 Table 9 Average Suspended Solids Concentrations at the Membrane Reactor 41
iv List of Figures Figure 1 Spiral Wound Membrane 15 Figure 2 SpiraSep Immersed UF Membrane Configuration 19 Figure 3 SpiraSep UF Membrane in Backflushing Mode 21 Figure 4 SpiraSep UF Membrane Air Scour 22 Figure 5 UF System during Filtration 22 Figure 6 UF System during Backflushing 23 Figure 7 Process Flow Diagram of the Pilot Unit 25 Figure 8 Permeability Profile at Various Operating Conditions 36 Figure 9 TMP Profile at Various Operating Conditions 37 Figure 10 Permeability Profile at CSD Bottler Operating Conditions 40 Figure 11 TMP Profile at CSD Bottler Operating Conditions 41 Figure 12 Permeability vs. pH at Various Flux Rates 44 Figure 13 TMP vs. pH at Various Flux Rates 44 Figure 14 Permeability vs. Flux at Various Operating pH 45 Figure 15 TMP vs. Flux at Various Operating pH 45
v Precipitative Softening and Ultrafiltration Treatment of Beverage Water Jorge T. Aguinaldo ABSTRACT Lime softening, chlorination, clarificati on and filtration have been long recognized treatment processes for beverage water specifi cally the carbonated soft drink (CSD) because it provides consistent water qua lity required for bottling plants, however these processes are becoming uneconomical and causes more problems than the benefits they offer. These processes require very large foot print, occ upy large plant volume, and generate large volume of sludge which causes disposal problem s. Chlorination produces trihalomethanes (THMs) and other by-products which are detrimenta l to health and imparts tastes to the final products. Using the newly developed submerge d spiral wound ultrafiltration membranes in conjunction with lime softening may replace th e conventional lime softening, clarification and filtration processes. This research was conducted to demonstrate the feasibility of integrating immersed ultrafiltration (UF) membrane with lime soften ing. The objectives of this research was to achieve the water quality required by the CSD bottlers; determine the relationships of operating parameters such as pH and membrane flux with trans-membrane pressure (TMP), and membrane permeability; determine the optim um dosage of lime; evaluate the operating parameters as basis for the design and constr uction of the full scale plant; and predict the membrane cleaning intervals.
vi A pilot unit consisting of lime reactor and UF system was designed and built for this research. The pilot unit was operated at various pH ranging from 7.3 to 11.2 and at membrane flux rates of 15, 30 and 45 gfd. The pilot unit was also operated at the CSD bottlers operating conditions which is pH 9.8 at flux of 30 gfd. The pilot unit operated for a total of 1800 hours. The raw water source was from city water supply. The filtrate from the pilot unit achieve d alkalinity reduction to 20 to 30 mg/L preferred by CSD bottlers, with lime dosage cl ose to the calculated value. The filtrate turbidity during the test was consistently w ithin 0.4 to 0.5 NTU. The TMP values obtained during the test ranges from 0.1 to 2.5 psi, while the permeability values ranges from 18.19 to 29.6 gfd/psi. The increase in flux results to co rresponding increase in TMP, and increase in operating pH, increases the rate of TMP. Pe rmeability decreases with increasing operating pH. The TOC reduction ranges from 2.6 % to 15.8% with increasing operating pH. No scaling of the UF membranes was observed during the test. Thirty days UF membrane cleaning interval was predicted. The results fr om this research can use as the basis of designing and operating a full scale Lime Softening UF Treatment Plant.
1 Chapter One Introduction The ingredients used in carbonated soft drinks (CSDs) including water are approved and closely regulated by the US Food and Drug Administration (FDA), but there are no defined water quality standards as long as it meets the federal and local drinking quality standards. The source water for soft drink manufacture is typically the municipal water supply, and at minimum it should comply with the primary and secondary National Drinking Water Standards. The municipal water supply however vary from one area to another and may not be able to provide consistent quality required for soft drink manufacture, therefore additional treatment is necessary. Most of the impurities that concerns the carbonated soft drink bottlers are those that affect the appearance and flavor of the product. The im portant ingredients of CSDs, aside from water are sugar, flavors and carbon dioxide. Carbon dioxide is the essential characterizing ingredient in all soft dri nks, the tingly fizz which gives a refreshing taste. When CO2 is dissolved in water, it imparts a unique taste. Natural carbonated or effervescent mineral water was popular because the minerals dissolved in water were believed to have beneficial medical properties. By 1800, artificial effervescent mineral water were introduced in Europe and North America. Then the innovative step of adding flavors to these popular soda water gave bi rth to the soft drink beverage we enjoy today.
2 Originally, carbon dioxide was made from sodium salts and the carbonated beverage became known as soda water (American Beverage Association, 2005). Lime softening is the most common water treatment process in CSD bottling plants. The typical water treatment process includes pr e-chlorination, lime softening with ferric salt dosage, media filtration or manganese greensand filtration. The addition of coagulants, such as ferric salts in lime softening process prom otes better sludge settling and also can reduce organic matter in the raw water. The unit processes above when accompanied by super chlorination followed by activated carbon f ilter and polishing filter comprise the conventional system for CSD product water (Morelli 1994). Lime softening has been the choice of bottlers because it provides consistent water quality suitable for bottling operations, regardless of the raw water quality. Recently, many bottling plants are replacing the lime-soda softening with other processes such as reverse osmosis, microfiltration and/or ultrafiltration. These processes, in most cases provide treated water that meets the quality requirements of the bottling. However, there are cases that lime softening can not just be replaced by reverse osmosis, especially when the high con centration of hardness in the raw water limits the recovery in the RO system. RO is excelle nt in reducing total dissolved solids, hardness and alkalinity in raw water, but it requires pre-treatment such as media filter or membrane microfiltration or ultrafiltration. The major CSD bottlers require the raw water feed to the RO system to be chlorinated to prevent bi ological fouling of the RO membranes. The drawback of chlorination of RO feed water is the breakdown of organic matter into smaller molecules forming trihalomethanes (THMs), wh ich are not rejected by the RO membranes.
3 The activated carbon, as part of the process removes residual chlorine and most of the organic matter that may impart off-taste and odor in the final product. Table 1 Selected Contaminants Limits in the National Primary and Secondary Drinking Water Standards (EPA, 2003) Primary Drinking Water Standards Turbidity: < 1 NTU or < 0.3 NTU in 95% of daily sampling in a month Chlorine (as Cl2): MRDL = 4.0 mg/L Cu: 1.0 mg/L Pb: 0.015 mg/L As: 0.010 mg/L F: 4.0 mg/L Hg: 0.002 mg/L NO2 (as N): 10 mg/L NO3 (as N): 1.0 mg/L TTHMs: 0.08 mg/L Secondary Drinking Water Standards Aluminum: 0.05 to 0.2 mg/L Chloride: 250 mg/L Sulfate: 250 mg/L Color: 15 (CU) Foaming Agents: 2.0 mg/L Iron: 0.3 mg/L Manganese: 0.05 mg/L Odor: 3 threshold number pH: 6.5 8.5 TDS: 500 mg/L In US there is no standard water quality specifically for CSD, because every bottler has its own quality control requirements. At minimum, the water supply to CSD bottling plant should meet the National Primary and Secondary Drinking Water Standards as shown in Table 1.
4 In 1958, the Society of Soft Drink Technologists carried out survey among bottlers on the quality of water they require for their plant (Morelli 1994). The result of the survey is shown in Table 2. In Canada, the Agriculture and Agri-Food Canada issued a water quality guideline (Agriculture and Agri-Food Canada 2000) for food and beverage industry which specifically includes carbonated beverages as shown in Table 3. In the CDS beverage industry variati ons in taste could be caused by the variations in the alkalinity of the product water. Lime so ftening seems to be the only treatment process that can provide consistent quality of treated water. Lime softening primarily will reduce and/or maintain alkalinity in the treat ed water to less than 50 mg/L as CaCO3 and a pH range of 8 to 9, however most CSD bottler operato rs are aiming for 20 to 30 mg/L alkalinity because it provides better yield (or less rejecti on of final product due to off taste). Another criterion is the hydroxide concentration whic h should be between 2 to 7 mg/L as CaCO3 based on calculation using Phenolphthalein and Methyl Orange Alkalinity values. Table 2 CSD Bottlers Water Quality Survey (Morelli 1994) Max. Min. Avg. Median Avg. Turbidity, NTU 10 0 2.3 2.0 Color, CU 20 0 4.8 3.5 Organic, Matter, ppm 5 0 0.4 0 Taste & Odor 0 0 0 0 Chlorine, ppm 0.2 0 0.03 0 Alkalinity, ppm CaCO3 130 0 70 50 Sulphates, ppm 900 0 240 225 Chlorides, ppm 525 0 210 225 Iron & Manganese, ppm 1.8 0 0.4 0.1 Copper, ppm 0.05 0 --Calcium, ppm 500 25 182 150 Magnesium, ppm 650 0 160 80 Sodium, ppm 900 500 --
5 Table 3 Canadian Water Quality Guidelines for Carbonated Beverage (AAFC, 2000) pH < 6.9 Color < 10 Hazen Units Turbidity 1 2 NTU Taste, Odor N.D. TDS < 850 mg/L Iron < 0.1 mg/L Manganese < 0.1 mg/L Carbonate < 5 mg/L Sulphate < 200 mg/L Chloride < 250 mg/L Fluoride 0.2 to 1.0 mg/L Hardness 200 to 250 mg/L Alkalinity 50 to 128 mg/L The CSD bottlers apply the multi-barrier concept in treating raw water into product water. Shachman (2004) defines multi barrier system as an orderly series of reliable processes that, in a complementary and incremental manner, completely removes or reduces targeted raw water adverse quality factors to acceptable levels, at lowest practical cost. To apply this concept, many CSD bottlers are incorporating membrane treatment processes, such as ultrafiltration, microfiltration, nanofiltration and reverse osmosis in their existing processes. In many cases, the membrane processes alone cannot provide the required product water quality. It is common to find membrane treatment after lime softeners. Talking to CSD quality personnel and plant operators, the majority expressed desire to simplify the lime softening and membrane processes, possibly to combine both processes. It is common for UF systems in CSD bottling plants to dose coagulant, such as ferric sulfate or ferric chloride. A novel approach is to dose lime to achieve softening. The application of lime for softening is not
6 the same as dosing ferric salts. By combini ng both the lime dosing and membrane treatment, it will be possible to reduce the lime dosage and sludge production, and achieve the desired product water quality at reduced cost. The memb rane utrafiltration is a barrier that can physically prevent microorganism from passing through into the treated water. The existing lime softening facilities can integrated with ultrafiltration. Additional minor modification will increase the existing plants capacity. The purpose of this research is to demonstrate the feasibility of combining lime softening with membrane ultrafiltration to achieve the water quality required in the bottling process with minimum usage of chemicals and eliminating continuous chlorination of the raw water.
7 Chapter Two Background 2.1 Lime Softening Lime softening has been long recognized as an effective process to reduce calcium and magnesium hardness in water by adding CaO or Ca(OH)2(lime) and/or Na2CO3 (soda ash) to precipitate calcium as CaCO3 and magnesium as Mg(OH)2. It will also remove CO2 in the water. In addition to hardness, other impurities such as iron, manganese, fluoride, phosphates, heavy metals, silica, chloride and total dissolved solids in the water are also removed with the addition of lime alone or in combination with other chemicals such as alum, sodium silicate, ferric and ferrous salts, fl occulant, etc. The elevated pH required in the process also inactivates many microorgani sms. Lime softening has been known to remove natural organic matter (NOM) in water specifically trihalomethane (THM) precursors (Collins, Amy, and King 1985). Lime softening was found to remove significant fraction of fulvic acid extracted from gr ound water (Liao and Randke 1985), and the NOM removal was achieved by the adsorption onto calcium carbonate and magnesium hydroxide formed in the process. EPA (1999) recommendation to enhanced total organic carbon (TOC) removal using precipitativ e softening is to provide the conditions that favor the formation of magnesium hydroxide and small calcium carbonate particles. This can be achieved by elevating the pH to 10.8 or higher, delaying carbonate addition and sludge recycling.
8 The degree of precipitation of calcium, magnesium and other impurities depends on the operating pH. Soda ash is also added to precipitate non-carbonate hardness and to precipitate excess lime. Caustic soda is also added to adjust the operating pH and promote precipitation of calcium and magnesium. This process is often called caustic soda softening. This process is applicable if there is enough calcium in the raw water to complete the softening reactions. The typical reactions in lime, or similar precipitative softening processes are: Lime as CaO when water is added becomes Ca(OH)2 CO2 + Ca(CO)2 = CaCO3 + H2O At pH 9.5 or above the following reaction will occur: Ca(HCO3)2 + Ca(OH)2 = 2CaCO3 + 2H2O Mg(HCO3)2 + Ca(OH)2 = CaCO3 + Mg CO3 + 2H2O At pH 11 or above Mg CO3 + Ca(OH)2 = CaCO3 + Mg(OH)2 + Ca(OH)2 (excess) Reactions with soda ash Ca(OH)2 + Na2CO3 = CaCO3 + 2NaOH CaSO4 + Na2CO3 = CaCO3 + Na2SO4 CaCl2 + Na2CO3 = CaCO3 + 2NaCl Ca(NO3)2 + Na2CO3 = CaCO3 + 2NaNO3 MgSO4 + Na2CO3 + Ca(OH)2 = CaCO3 + Mg(OH)2 + Na2SO4 MgCl2 + Na2CO3 + Ca(OH)2 = CaCO3 + Mg(OH)2 + 2NaCl Mg(NO3)2 + Na2CO3 + Ca(OH)2 = CaCO3 + Mg(OH)2 + 2NaNO3
9 In lime softening, additional and/or excess chemicals are often added to increase the mass of sludge to promote settling. 2.2 Limitations/ Problems Associated with Lime Softening The lime softening although reliable and being used in the beverage industry for almost a century, has its limitations and problems. Some of the limitations and problems associated with lime softening are: Disposal of large amount of sludge generated by the process Requires larger foot plant print for the lime reactor, as well as the sludge handling equipment, lime preparation and storage facilities. Additional chemicals are required to promote settling of the sludge and solids. Requires media filtration after clarification. The lime softening plant should be continuously running and requires longer time to stabilize after start-up. The lime-soda softening is more expensive compared to other competing processes. There are very limited companies now specialized in the manufacture of lime softening systems.
10 2.3 Precipitative Softening The USEPA, acting on the 1986 Amendment to the Safe Drinking Water Act (SDWA), set maximum contaminant level goals (MCLGs) for a variety of contaminant that is present in drinking water. Th e disinfectants and disinfection byproducts (DBPs) are among the list of contaminants for regulated in the Disinfection Byproduct Rule (DBPR). USEPA developed treatment techni ques or a maximum contaminant level (MCL) that is as close to the MCLG as is feasible with the use of the best available technology (BAT). As part of the DBPR has USEPA, in cluded a treatment technique requirement to remove natural organic matter (NOM) which serves as the primary precursor for DBP formation. The goal of this pre-treatment t echnique is to provide additional removal of NOM, measured by total organic carbon (TOC). The USEPA Enhanced Coagulation and Enhanced Precipitative Softening Guidance Manual define enhanced coagulation as a term to represent the process of obtaining improved removal of DBP precursor by conventional treatment whereas enhanced softening refers to the process of obtaining improved removal of DBP precursors by precipitative softening. In the implementation of the enhanced coagulation and enhanced softening requires process modification in the existing plants and will have some impacts which may be either beneficial or detrimental. USEPA cited some of the impacts as: Inorganic constituents levels (manganese, aluminum, chloride and sodium) Corrosion control Disinfection Particle and pathogen removal
11 Residuals (handling, treatment, disposal) Operation and maintenance Recycle streams In addition to the above, from the beverage plant operators point of view, the impacts are: Maintaining treated water quality suitable for beverage bottling operations that are often to higher quality standard compared to the municipal drinking water quality. Operating costs Limited plant area to implement process modification Additional cost associated with the plant upgrade Precipitative softening specifically lime process, comes in various forms and variations. Humenick (1977) listed four process types, based on the amount of chemicals added: Single-stage lime process is used when the source water has high calcium, low magnesium carbonate hardness (usually less than 40 mg/L as CaCO3), and no noncarbonate hardness. Single-stage lime softening is not intended for magnesium hardness removal. Lime is added up stream of the reactor in a separate flash mixing chamber or into the reactor-clarifier. The pH of the water leaving flash mixer is about 10.2 to 10.5.
12 Excess lime process is used when the source water has high calcium, high magnesium hardness, and no noncarbonate hardness. Excess lime process can be single or in two stages. Excess lime is added to precipitate magnesium carbonate hardness as magnesium hydroxide. The pH of the water after flash mixing will be from 10.2 to 11.2. Above pH 10.2, causticity will be present. Single-stage lime-soda process is used when the source water has high calcium, low magnesium hardness, (usually less than 40 mg/L as CaCO3), and some calcium non-carbonate hardness. This is similar to the single-stage lime process, except that the soda ash is added for the removal of non-carbonate hardness. The s oda ash is added in the flash mixer or sequentially after the lime has been added. Excess lime-soda process is used when the source water has high calcium, high magnesium carbonate hardness and some non-carbonate hardness. The addition of soda ash in the excess lime process will allow removal of non-carbonate hardness, while removing calcium and magnesium hardness. Excess lime-soda process can be in one or two stages, however two stage process is common practice, because the soda ash added in the second stage will remove the excess lime. In addition to the above, other variations of lime softening include the following: Pellet softening (Van der Veen, C. & Graveland, A., 1988) uses fluidized bed of grains on which crystallization of CaCO3 takes place.
13 The softening reaction takes place in the presence of suspended bed of fine sand or crushed CaCO3 that acts as catalyst. Feed water and chemicals enter tangentially at the bottom of the pellet reactor chamber and mix immediately. The treated water rises through the reactor in swirling motion. The upward velocity is sufficient to keep the sand fluidized. The precipitated hardness particles attaches to the surface of the sand grains and the sand diameter increases. Large grains are continuously removed. Ultra high lime softening (Batchelor, B; Lasala, M. McDevitt, M; Peacock, E., 1991) is another variation of lime softening and is used when the source water has high calcium and magnesium hardness, and high silica concentration. Excess lime is added to the reactor to increase the operating pH to above 11. Ultra high lime softening is usually is a two stage process. Other modification of lime softening is the addition of caustic soda instead of lime to achieve the reaction pH. In all the processes above the softening is achieved by precipitation of CaCO3 and Mg(OH)2 at elevated pH, where the solubilities of CaCO3 and Mg(OH)2 are relatively low. The various process modifications in lime softening also enhances the removal of the precipitate through effective settling or, in case of pellet softening attachment to the fine sand grains
14 2.4 Ultrafiltration Ultrafiltration is a pressure driven membrane process, where the source water is passed through a membrane with nominal pore size of 0.01 to 0.1 m m, and suspended solids, colloidal particles, bacteria and other particles are retained. Ultrafiltration also removes high molecular weight organic matter. The typical ultrafiltration membranes have a typical molecular cut-off of 150,000 daltons (1 dalton or Da = 1/12 mass of one atom of Carbon-12), however through the addition of coagulants, it can effectively remove organic matter with molecular weight down to less than 20,000 daltons. The addition of coagulant in the form of ferric salts, poly aluminum chloride or alum is common in ultrafiltration process. The addition of lime in the feed of ultrafiltration membrane was never been reported in the literature, but there were published reports integrating pellet softening with UF membrane treatment (Li, Jian, and Liao, 2004). In most membrane processes especially in reverse osmosis and nanofiltration, CaCO3 scaling is a common problem. In treating hard water using ultrafiltration, the precipitation of CaCO3 can be a problem, especially occurring in capillary or small diameter tubular UF membranes. The development of the Spirasep UF membrane, which air-scoured immersed membranes in spiral configuration developed by Trisep, will minimized the build up of scale in the UF membrane surface. Compared to RO or NF, there is no change in salt concentration in the membrane surface, therefore formation of scale will be minimized. The Spirasep membrane is similar in appearance to 8 diameter x 40 length RO membrane, made of polyethersulfone, and with effective membrane area of 178 ft2. The operating pH is from 4 to 11 on continuous basis and pH of 2 to 12 for cleaning. The Spirasep membrane has chlorine tolerance of 2,000 mg/L.
15 In this research the manufacturers ope rating guidelines were strictly followed because the UF unit is a working commercial unit with single UF element and to limit the variables. Among the operating conditions maintained were the following: Continuous aeration at the recommende d aeration rate of 0.02 to 0.05 scfm per square feet of membrane area. Continuous aeration was recommended for water with high suspended solids concentration. Back flushing was set every 15 minutes with 30 seconds duration at the rate of 45 gfd. Trans-membrane pressure was define d and measured as per the membrane manufacturers guidelines as shown in Appendix B. Figure 1 Spiral Wound Membrane
16 Chapter Three Materials and Methods 3.1 Experimental Plan The objective of this experiment is to demonstrate the applicability of combining lime softening with ultrafiltration membrane to produce water that meets the beverage water quality. The specific objectives are: Reduce hardness in the feed water and, at the same time, maintaining alkalinity of the treated water to less than 50 mg/L as CaCO3 Determine the relationships of operati ng pH and membrane flux with transmembrane pressure and membrane permeability. Compare the lime dosage in this res earch with the lime dosage used in conventional lime softening plant, treating similar water source. Evaluate the operating parameters important in designing a full scale plant. These includes membrane flux rates, permeability, recovery, backwash intervals, cleaning intervals, and trans membrane pressure.
17 3.2 Pilot Lime Softening Ultrafiltration Unit The pilot lime softening ultrafiltration system was designed and built by Doosan Hydro Technology, Tampa, Florida. The details of the plant are described in Appendix A. The pilot unit is a full scale commercial operating plant with one SpiraSep UF membrane immersed in a reactor tank. 3.2.1 Lime Reactor The lime reactor is a polyethylene cyli ndrical conical bottom tank, with maximum capacity of 200 gallons, to allow 30 minutes retention at the maximum flow of 5.7gallons per minute (gpm). The tank was provided with disc harges at three different levels for the different flow rates. The elevation of the lime reactor is adjustable, in order to allow gravity flow into the membrane reaction tank. The lime solution or slurry was fed by a BLUEWHITE Model A-100N Peristaltic Metering Pum p, with a maximum capacity of 2.3 gallons per hour (gph). The flow rate of the me tering feed pump was controlled by the pH transmitter. The lime slurry or solution was fed to the incoming raw water into the mixing chamber which directed the flow to the bottom of the lime reactor. The mixing chamber was provided with a mixer (FPI Model PM1/20 PE) driven by a 1/20 horse power (hp) electric motor. The precipitate, or sludge, settles at the bottom of the tank. Sludge was expected to be carried over to the membrane reactor tank. The bottom of the lime reactor was provided with a connection for pumping out the sludge at scheduled interval. The pH sensor was installed at the inlet of the membrane reactor tank.
18 3.2.2 SpiraSep Ultrafiltration Membrane The SpiraSep UF membrane manufactured by Trisep Corp. of Goleta, California is an immersed, negative-pressure ultrafiltration process, which will remove suspended solids, turbidity, viruses, bacteria, and some organic compounds. A typical SpiraSep system consists of an array of spiral wound elements submerged inside a process tank. The membrane elements are attached to a manifold assembly, consisting of a central permeate header with an array of membrane permeate ports, which connects to the SpiraSep membrane. A vacuum is generated by the suction of a centrifugal pump, creating the necessary net drive pressure to pull water through the SpiraSep membrane. Air is bubbled up through each membrane element via bubble diffusers, creating tremendous shear forces on the membrane surface that remove any suspended solids. A small amount of a coagulant is injected into the process influent. The enhanced coagulation process will help reduce organic fouling and improve TOC and color reduction. Periodically (on a timed basis), permeate water is reversed through the membrane, or back flushed, to help further remove the accumulated suspended solids. This process also introduces a small amount of disinfectant to help control the microbial activity on the membrane surface. Concentrate is removed from the process tank, and is typically less than 10% of the influent rate. SpiraSep membranes can also be chemically cleaned through one of two processes: a periodic flux enhancement (PFE) or a flux recovery clean (FRC) procedure.
19 Figure 2 SpiraSep Immersed UF Membrane Configuration The pilot plant was manually controlled and operated with several automated Features, such as backwashing. Feed from a pressurized source is delivered to the UF system, and is controlled by a feed control valve. A blower is operated continuously to deliver pressurized atmospheric air to the membrane element. Membrane backwashing is controlled by a timer, and is performed on a timed basis. Membrane cleaning is operator initiated. 3.2.3 Pilot Lime Softening Ultrafiltration Process Control Description The feed water to the pilot unit was delivered to the lime reaction tank and was controlled by a control valve and rotameter. A sample line from the feed was connected to the in-line turbidity analyzer. Lime solution was added to the feed water at the flash mixing chamber. Lime was dosed by a peristaltic chemical dosing pump, drawing lime solution or slurry from a solution tank. The dosing rate of the chemical dosing
20 pump was controlled by the pre-set operating pH. The pH probe measures the pH of the water in the overflow. From the flash mixing chamber, water flows downward to the conical bottom of the lime reaction tank. A provision for another coagulant dosing was included, in the event that another coagulan t will be added in conjunction with or to supplement the lime. The CaCO3 and other precipitates settled in the conical bottom of the lime reactor tank and softened water overflowe d to the UF process or membrane tank. Carryover CaCO3 and/or precipitate were expected in the overflow. Feed to the ultrafiltration unit results in two streams: filtrate and concentrate. Feed was introduced to the membrane tank fr om the overflow in the lime reaction tank. Once feed water was introduced to the membra ne tank, the blower was turned on. The air flow was manually adjusted to provide the proper air flow rate to the element. The air flow rate was measured using a flow meter. The concentrate valve was set to obtain the proper concentrate flow rate. Once the membrane tank was completely filled, the Process Logic Controller (PLC) will start the filtrate pump and open the concentrate valve. The UF filtrate pump provides the necessary net drive pressure to force feed water through the membrane surface. A self-priming centrifugal pump generates a vacuum, typically less than -10 psi, drawing water through the UF membrane surface. Filtrate flow was manually set with a control valve but pump operation is controlled via the PLC.
21 Figure 3 SpiraSep UF Membrane in Backflushing Mode The filtrate pump flow rate was adjusted manually with the permeate control valve. The UF membrane was back flushed at set interval The water required for the membrane back flush was taken from the UF filtrate tank and pumped to the membranes using a separate backwash pump. The backwash pump reverses the flow of water through the UF membranes. A membrane back flush was performed every 15 minutes for 30 seconds and is automatically controlled by the PLC. Once filtrate production started, timers for the back flush frequency and Periodic Flux Enhancement (PFE) are started. The blower remains on running at the manually set value.
22 Figure 4 Spirasep UF Membrane Air Scour Figure 5 UF System During Filtration When a back flush sequence is started, the automatic feed valve was closed, and the filtrate pump and blower were automatically turned off (concentrate valve remains
23 open). UF filtrate water and chlorine were then backflushed through the membrane for a period of about 30 seconds. A Variable Fr equency Drive (VFD) adjusts the back flush pump speed, to the manually set value. Output of the metering pump was manually adjusted. Excess water introduced to the tank was rem oved via a tank overflow and/or concentrate line. Once the back flush sequence was co mpleted, the back flush pump and chlorine metering pump were automatically turned off. The blower was turned on and allowed to operate for 10 15 seconds before the filtrate pump was restarted and the feed valve opened to allow normal filtrate production. Figure 6 UF System During Backflushing The UF membrane was continuously aerated to prevent and minimize membrane fouling. A blower takes atmospheric air and bubbles them up through individual membrane module via an aeration disc. The blower was operated using a VFD, and the motor speed is set manually. The operation of the blower was controlled by the PLC.
24 Air was delivered to the UF membrane through a coarse bubble diffuser. The air diffuser was attached to an aeration pipe. The aeration pipe contains a manual flow control valve and air flow indicator to ensure proper air flow. Various chemicals were dosed for various system operations. Chlorine was dosed during each back flush, in addition to PFE and Clean-In-Place (CIP) processes. Sodium hydroxide was injected for just PFE and CIP processes. Citric acid was dosed for PFE and CIP processes. The flow rates of the chemi cal dosing pumps were set manually. Operation of the chemical dosing pumps during backwa sh, PFE, and CIP was controlled by the PLC. Operating performance can be optimized through the use of PFE. A chemical solution was backwashed through the membrane s in situ to perform a quick chemical treatment. This process was performed while the membrane tank was filled with process water, requiring approximately 20 30 minutes. This was done on a daily or every two days. When a PFE process was initiated, the feed valve was closed, and the filtrate pump and blower were turned off. UF filtrate and ch emicals were then automatically back flushed through the membranes while they are still immers ed in the feed water (i.e. membrane tank is not drained for this process). Excess wate r introduced to the tank was removed via a tank overflow and/or concentrate line. During membrane cleaning, a cleaning solution was back flushed through the membranes until the filtrate tank was completely filled. The membrane was statically soaked in the cleaning solution for approximately 4 8 hours. A CIP process is typically performed once every 3 months fo r municipal water treatment. Actual CIP frequency is determined through pilot testing and actual plant operation. CIP is a manual operation. In high suspended solids environmen t like in lime softening CIP every 2-3 weeks
25 is acceptable. The UF system is normally designed to allow the membrane elements cleaned in place in the membrane tank. UF filtrate a nd cleaning chemicals are back flushed through the membranes until the CIP tank is completely f illed. At the end of the chemical soak, the tank is drained and then refilled. Figure 7 Process Flow Diagram of Pilot Unit
26 3.3 Chemicals The chemicals used in the pilot test are: Hydrated Lime, Ca(OH)2, 93%, CAS 1305 78-8, technical grade Sodium Hypochlorite, NaOCl, 12% chlorine CAS 7681-52-9 Sodium Hydroxide, NaOH, 45% CAS 1310-73-2 Citric Acid Anhydrous 99.5%, C6H8O2 CAS 77-92-9 A 3.2 % lime slurry was prepared by adding 32.24 grams of hydrated lime (93% Ca(OH)2) per liter of water mixed into the slurry tank. The 3.2 % lime slurry has a specific gravity of 1.020 or 2.84 Baume, which will be verified using a Hydrometer (Cole Palmer Cat# C-08287-55, range SG 1.000 to 1.225, Baume 0 to 26 deg). The sodium hypochlorite (12% chlorine) was dosed at 10 mg/L during back flush and 100 mg/L during Periodic Flux Enhancement (PFE). The sodium hypochlorite solution for both the back flush and the PFE back flush were dosed by metering pumps drawing directly from the sodium hypochlorite container. The citric acid crystals was dissolved in water at 200 grams/L solution. From this stock solution, the citric acid was dosed directly to the PFE back flush line at rate of 2 l/h. during CIP. The citric acid was dosed to the CIP line at the rate of 20 l/h. Caustic soda, 45% solution was dosed at 0.1% or 1,000 ppm using chemical feed pump at a rate of 0.63 l/h drawing directly from the caustic soda container.
27 3.4 Experimental Procedures The pilot unit was initially operated for one week without any chemical addition to stabilize the flow and calibrate the instruments. After one week the pilot unit was operated for approximately one month with varying dosage of lime to determine the conditions that can provide the desired water quality. The pilot unit was operated for another month at the selected optimum operating conditions. The lime slurry was dosed by peristaltic pump (Blue White Model A1N30F-6T) with maximum capacity of 1.25 gph (4.73 lph). This pump is capable of delivering lime up to 346 mg/L when operating at flux of 15 gfd and 120 mg/L when operating at 45 gfd. Operating flux of 15 gfd was selected to be the starting flux, based on previous pilot testing using other coagulants such ferric chloride, ferric sulfate and alum. Trisep recommends the following sustainable flux rates: For municipal secondary effluent: 15 to 18 gfd; municipal drinking water: 25 gfd; landfill leachate (with chemical precipitation): 15 gfd. The flux will eventually increase to 30, and 45 gfd. Lime slurry was dosed to achieve pH of 8.3, 9.4, 10.6, and 11.2 at the lime reactor overflow or discharge to the membrane tank. Th e various phases of testing were performed at the following schedule: Day Flux (gfd) pH 1 15 Feed water pH 2 15 Feed water pH 3 30 Feed water pH 4 30 Feed water pH 5 45 Feed water pH 6 45 Feed water pH 7 15 8.3 8 15 8.3 9 30 8.3 10 30 8.3
28 11 45 8.3 12 45 8.3 13 15 9.4 14 15 9.4 15 30 9.4 16 30 9.4 17 45 9.4 18 45 9.4 19 15 10.6 20 15 10.6 21 30 10.6 22 30 10.6 23 45 10.6 24 45 10.6 25 15 11.2 26 15 11.2 27 30 11.2 28 30 11.2 29 45 11.2 30 45 11.2 The flux was set by controlling the flow through the filtrate pump through the adjustment of the filtrate control valve. Du ring the test the trans-membrane pressure (TMP) was monitored through a digital pressure indicator connected to a pressure transmitter installed at the manifold between the UF membrane filtrate discharge and the suction of the filtrate pump. The pilot testing log will include the following information: Date and time, actual flow rate reading, total flow (from flow totalizer), pH, temperature, raw water and filtrate turbidity, TMP or UF pump suction lin e pressure located at the same level as the water in the UF reactor tank. The net flow in each segment of test can be determined and used as basis of calculating the average permeability.
29 The flow is indicated by a SIGNET Model 8550 Flow Transmitter with digital flow indicator and totalizer, receiving signa l from a SIGNET Model 515 flow sensor. The pH is indicated by a SIGNET Model 8750 pH transmitter with digital pH and temperature indicator, receiving signal fr om a SIGNET Model 2754 pH probe. The TMP is measured by local mounted EFFECTOR pressure transmitter/ indicator. The turbidity is continuously monitored by HACH Model 1720D Low Range Process Turbidimeter, provided with sample connections to allow turbidity measurement of either the raw water or the filtrate. The permeability was plotted against elapsed time. The permeability was calculated as flux (in gfd) divided by the trans membrane pressure (psi). The permeability has a unit of gfd/psi. The TMP values were also plotted against time. Composite samples of feed and filtrate were taken daily and were analyzed for pH, alkalinity, calcium and magnesium hardness, conductivity, turbidity, and total organic carbon (TOC). Sample of the water in the membrane reactor was also taken for suspended solids analysis. 3.5 Analytical Procedures The analysis of the water samples were made following the EPA Methods and Standard Methods for the Examination of Water and Wastewater (APHA, AWWA, WEF, 1995). The water samples taken dur ing the test were sent to Severn Trent Laboratories, Inc. (STLI) in Tampa for analysis. STLI is EPA cer tified laboratory. Chemical analyses were also conducted on site using Hach test kits fo r verification and calibration of instruments.
30 Water analysis was also conducted in the nearby CSD bottlers laboratory, for comparison. Analysis was also done at the Ameraican Water Chemicals facilities. 3.5.1 pH and Temperature pH and temperature were directly measured using the installed pH analyzer (Signet 8750 ProcessPro pH Transmitter) with immersed probe (Signet 2754 pH probe). The immersed pH probe was calibrated with pH buffer kit (Signet PN 30700.390). The pH of the water samples were measured using portable pH meter (Hach SensION 1 Portable pH meter). The probe of the portable pH meter was calibrated using pH 4.01 and pH 10.0 buffer solutions (Hach PN#22834-49 and PN#22836-49). 3.5.2 Alkalinity Alkalinity was measured using SM18 2320 B. 3.5.3 Calcium and Magnesium Hardness The calcium and magnesium hardness were measured using EPA SW8466010B Inductive Coupled Plasma Atomic Emission SpectrometryTotal Recoverable. 3.5.4 Turbidity Turbidity was measured using a portable turbidimeter (Hach Model 2100 Series) calibrated with <0.1, 1, 20, 100 and 800 NTU stabilized formazin standards (Hach Calibration kit PN#26594-05) and EPA Method 180.1. Turbidity was also measured directly from the HACH Model 1720D Low Range Process Tu rbidimeter installed in the pilot unit.
31 3.5.5 Total Suspended Solids Total Suspended Solids was analyzed using EPA Method 160.2 The suspended solids analyzed was the calcium carbonate precipitate in the UF reactor tank. 3.5.6 Total Organic Carbon Total organic carbon was analyzed using EPA Method 415.1
32 Chapter Four Results and Discussions The pilot testing was conducted at Doosan Hydro Technology, Inc. facilities in Tampa, Florida. The pilot testing was divided into three phases. The first phase was to stabilized the flows and calibrate control valves and instruments. The first phase started on October 15, 2005, and was supposed to last one week, however it was extended by one more week, due to mechanical and in strument problems. The second phase was performed at varying flux and pH conditions It started on October 29, 2005 and lasted four weeks. The objective of the third phase was to simulate the operation in a CSD Bottler Plant, based on the data obtained from th e second phase. The third phase started on December 2, 2005 and ended on January 5, 2006. Th e source of feed water during the test was city of water supply. 4.1 Initial Operating Conditions Without Chemical Addition The purpose of running the pilot unit at different flux levels, without the addition of chemicals, is to determine the flow characteristics of the unit and to calibrate the instruments. Based on the UF membrane area of 178 ft2, the filtrate flow rates of 1.9, 3.7 and 5.6 gpm corresponded to flux values of approximately 15, 30 and 45 gfd. During the initial test run, it was noticed that display on the pressure indicator in the
33 suction line of the UF permeate pump was giving reading on increments of 0.5 psig and has to be replaced with a pressure indicator to provide reading down to 0.1 psig. pH and temperature were continuously displayed. The pressure measured on the UF membrane filtrate discharge and suction of the UF pe rmeate pump pipework is the trans-membrane pressure. The location of the pressure sensor wa s in the same level as the water level in the UF reactor tank as recommended by the membrane manufacturer. This eliminated the need for correcting for the difference in hydraulic heads. The vacuum pressure reading can be considered as the trans-membrane pressure. Controlling the filtrate flow with the manual ball valve at the discharge of the UF pump was difficult, especially at lower flow, and it was replaced with a more accurate globe valve. After the flow and pressure readings were stabilized, the pilot unit was operated with varying flows of 1.9 to 5.6 gpm. The back flushing was set every 15 minutes for duration of 30 seconds. It was expected that the TMP will increase prior to back flushing. During the initial run at 1.9 gpm, the TMP remained at -0.5 psi, before and after back flushing throughout the 2 days of operation. At the flow of 3.7 gpm, the TMP stayed consistently at -1.1 psi after back flushing, and the pressure before back flushing was -1.5 psi. When operating at 5.6 gpm, the TMP after back flushing was -1.7 psi and increased to -2.0 psi before back flushing. Water samples were taken for analysis. Raw water analysis is shown in Table 4. The average pH of the feed water is 7.3 and the water temperature ranges from 20 to 25oC. Chlorine was not dosed during back flushing and during PFE.
34 Table 4 Raw Water Analysis 10/15/05 11/12/05 12/10/05 pH 7.31 7.5 7.3 Alkalinity, mg/L CaCO3 76 70 75 TOC, mg/L 3.8 4.0 3.6 Ca, mg/L CaCO3 65 60 62 Mg, mg/L CaCO3 4.2 4.5 4.2 Turbidity, NTU 0.1 0.1 0.1 4.2 Operation at Varying Flux and pH The second phase of the pilot testing wa s the addition of lime to achieve operating pH values of 8.3, 9.4, 10.6 and 11.2, at flows of 1.9, 3.7 and 5.6 gpm (or flux of 15, 30 and 45 gfd). The pilot unit was operated continuous ly for 2 days for each flow condition. The pH was set to the desired operating pH and the chemical feed pump automatically dosed the required lime solution. The average TMP values before and after back flushing are shown in the Table 5. The flux and permeability values at different operating c onditions are shown in Table 6. The Permeability Profile at various ope rating conditions is shown in Figure 8. The permeability values range from 50% to 85% of the clean water permeability for SpiraSep UF membrane, which is 35 gfd/ psi. Figure 9 shows the TMP profile during the test. It can be observed, TMPs tends to increase with increasing flow (or flux) and operating pH.
35 Table 5 Average Vacuum Pressures or TMP Values in psi Before and After UF Back Flushing at Various Flux Values Flux Values pH 15 gfd 30 gfd 15 gfd 7.3 -0.4/-0.5 psi -1.1/-1.3 psi -1.8/-2.0 psi 8.3 -0.6/-0.8 psi -1.2/-1.5 psi -1.9/-2.2 psi 9.4 -0.6/-0.8 psi -1.3/-1.8 psi -2.2/-2.6 psi 10.6 -0.7/-1.0 psi -1.6/-2.2 psi -2.4/-2.8 psi 11.2 -0.6/-1.0 psi -1.7/-2.3 psi -2.6/-3.4 psi Note: After BF/Before BF Table 6 Flux vs. Permeability at Various Operating pH Flux pH 15 gfd 30 gfd 45 gfd 7.3 29.6 26.29 26.12 8.3 25.81 25.31 23.95 9.4 25.14 23.03 20.43 10.6 23.0 19.32 19.21 11.2 21.4 19.05 18.17 Note: Permeability is gfd/psi
36 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.000100200300400500600700800900Operating Hoursgfd/psi Figure 8 Permeability Profile at Various Operating Conditions
37 0 0.5 1 1.5 2 2.5 3 0100200300400500600700800900Operating HoursTMP psi Figure 9 TMP Profile at Various Operating Conditions Composite raw water and filtered water samples were analyzed for Ca, Mg, Alkalinity, pH, turbidity, and TOC. Grab water sample from the membrane reactor was also taken for total suspended solids analysis. The results of the water analysis are shown in Table 7. An analysis of water sample was also conducted by the CSD Bottler and shown in Table 8. Note that there is difference between the operating pH value and the pH of the Filtrate analyzed in the laboratory. The pH of th e filtrate was expected to be lower due to the effect of aeration in the UF tank which tends to strip the CO2 or add CO2 from the air. Aeration has stabilizing effect on the filtrate. Du ring the test the amount of lime in each run was not monitored, however every time a batch was prepared, the quantity was recorded.
38 Table 7 Analysis of Water Samples at Various Operating Conditions Operating pH 7.3 7.3 8.3 9.4 10.6 11.2 pH of the Sample 7.31 7.31 8.06 9.2 10.3 10.8 Type of Water Raw Filtrate Filtrate Filtrate Filtrate Filtrate Alkalinity, mg/L CaCO3 76 76 62 30 36 36 TOC, mg/L 3.8 3.8 3.7 3.5 3.3 3.2 Ca, mg/L as CaCO3 65 65 57 38 41 56 Mg, mg/L as CaCO3 4.6 4.6 4.5 4.3 3.9 2.4 Turbidity NTU 0.1 0.05 0.05 0.05 0.05 0.05 From the tables above, it can noted that there is a significant reduction of alkalinity and hardness, whereas at pH 10.6, the alkalinity and hardness increased. At pH 10.6 and Above, the increase in alkalinity and calcium was due to the lime addition. The magnesium concentration continues to drop as the pH went up as expected. The dilute sludge that accumulates at the bottom of the membrane reactor tank is manually drained, when the unit is stopped. During backwashing, the water in the membrane reactor overflows to lime reactor tank. The concentration of the suspended solids in the membrane reactor is shown in Table 9. It was observed that there was slight change in the sludge concentration wh en operating pH changed as shown in Table 9.
39 The concentrated sludge that accumulated at the bottom of the lime reactor tank was pumped out using another rotary flexible impeller pump rated at 0.25 gpm. Usually, 1/3 of the sludge in the conical section of the lime reactor tank was drained when the volume of sludge reaches the top of the conical section. 4.3 Operation at CSD Bottler Plant Conditions The next phase of the test was to simulate the operation in an actual CSD bottler plant condition. Operation at pH 9.4 to 9.8 was chosen because the results in the previous tests satisfied the water quality requirement of the CSD bottler using the same source water as used in this test, although their actual operating pH was slightly higher. The resulting alkalinity level was favorable to their operation. The flux selection of 30 gfd (or flow of 3.7 gpm) was based on the following factors: economics, competing UF membranes operating flux, test results from the second phase of the test, and guideline of the membrane manufacturer. The test also predicted the in tervals between cleaning and estimated the consumption of lime. Water samples were ta ken and analyzed. The amount of lime used was also monitored. The test lasted for over 30 days. Figure 10 shows the permeability profile and Figure 11 shows the TMP profile throughout the duration of the te st period. On the 18th day of test the TMP has almost doubled and the permeability dropped to down to 50% from the first day value. Based on expe rience, when this condition occurs, it is necessary to chemically clean the UF membra ne. The cleaning was made as per the CIP procedure described Section 3.2.3. After clean ing the TMP and permeability values were
40 restored to the first day values. The operati on of the pilot unit was continued for another 10 days after cleaning. The TMP and permeability pr ofile after cleaning is similar to the initial profile. The analysis of the filtrate by STLI and the CSD bottler are shown in Tables 7 and 8. Table 8 Analysis of the Filtrate by CSD Bottler Operating pH 7.3 9.45 9.6 9.8 pH (Lab) 7.3 8.49 9.65 9.14 Phenolphthalei n Alkalinity, mg/L CaCO3 4.2 9.6 23.3 14.8 Methyl Orange Alkalinity, mg/L CaCO3 88.1 35.8 33.7 26.8 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 02004006008001000 Operating HoursPermeability Permeability gfd/psi Figure 10 Permeability Profile at CSD Bottler Operating Conditions
41 0 0.5 1 1.5 2 2.5 02004006008001000 Operating HoursTMP psi Figure 11 TMP Profile at CSD Bottler Operating Conditions The concentration of suspended solids in the membrane reactor tank was maintained at 600 to 700 mg/L range. Backflus hing seemed to maintain constant solids concentration in the membrane reactor. During b ackflushing, the excess water flowed back to the lime reactor tank, carrying suspended so lids, and the backwash water diluted the water in membrane reactor. The sludge from the membrane and lime reactor tanks were drained as described in Section 4.2. Table 9 Average Suspended Solids Concentrations in the Membrane Reactor Operating pH 7.3 8.3 9.4 10.6 11.2 Suspended Solids conc., mg/L 10 580 600 600 680
42 Chapter Five Summary and Conclusions 5.1 Alkalinity Reduction Alkalinity reduction to less than 50 mg/L or to the preferred level of 20 to 30 mg/L and maintenance of the desired Phenolphthalein Alkalinity and Methyl Orange Alkalinity (2*P alk MO alk = 2 to 7) can be achieved continuously in the lime softening UF unit with relatively simpler control, operation and maintenance compared to conventional lime softening process. The lime softening UF unit can be started in a matter of minutes, unlike the conventional lime softening which requires hours or days to build up of the sludge blanket before stable operation is achieved. The lime dosage during the third phase of test (operating pH=9.8) was 70 mg/L, based on raw water alkalinity concentration of 76 mg/L and pH of 7.3 and the filtrate alkalinity and pH are 26.8 mg/L and 9.18 respectively. The theoretical or calculated dosage using the Rothberg, Tamburini, and Windsor model was 65 mg/L. The lime dosage of the CSD bottler was in the range of 120 to 130 mg/L operating at pH of 9.8 to 10.2 with ferric chloride addition. 5.2 UF Filtrate Turbidity The turbidity of the filtrate was consistently observed to be in the range of 0.04 to 0.05 NTU throughout the duration of the test. The filtrate turbidity was not affected by
43 the incoming feed water turbidity. When the pilot unit was operated without the lime addition, the feed water and filtrate turbidity were 0.1 NTU and 0.05 NTU, respectively. The suspended solids concentration in the membrane reactor tank throughout the test was in the range of 580 to 650 mg/L. Table 9 shows the average suspended solids concentration in the membrane reactor. 5.3 Trans-membrane Pressure (TMP) vs. pH and Flux The increase in flux results to corresponding increase in TMP, however as the operating pH increases, the rate of TMP increases as shown in Figures 13 and 15. 5.4 Permeability The operating the pH vs. permeability profile shown in Figure12 indicates, the permeability decreases with increasing operating pH. The TMP vs. flux profile shown in Figure 14 indicate permeability decrease with increasing flux. The decline in permeability during the second phase of the test was due to the increase in operating pH. The starting and ending average permeability values were 31.25 gfd/psi and 17.53 gfd/psi. The prolonged operation without CIP had not impacted the permeability, because when the third phase of the test started, the starting average permeability during the first 2 days of operation was 26.93 gfd/psi, which is comparable to 25.5 gfd/psi when the operation started in second phase of the test at pH 9.4.
44 0 10 20 30 40 50 60 70 80 90 220.127.116.110.611.2Operating pHPermeability gfd/psi 45 gfd 30 gfd 15 gfd Figure 12 Permeability vs. Operating pH at Various Flux Rates 0 0.5 1 1.5 2 2.5 3 18.104.22.1680.611.2Operating pHTMP psi 15 gfd 30 gfd 45 gfd Figure 13 TMP vs. Operating pH at Various Flux Rates
45 0 5 10 15 20 25 30 35 153045Flux gfdPermeability gfd/psi pH 7.3 pH 8.3 pH 9.4 pH 10.6 pH 11.2 Figure 14 Permeability vs. Flux at Various Operating pH 0 0.5 1 1.5 2 2.5 3 153045Flux gfdTMP psi pH 7.3 pH 8.3 pH 9.4 pH 10.6 pH 11.2 Figure 15 TMP vs. Flux Various Operating pH 5.5 Total Organic Carbon (TOC) The data in Table 7 indicate that there was no reduction in TOC when the pilot unit was operated without lime addition. With the addition of lime, there was a slight reduction of TOC. The reduction in TOC ranged from 2.6% to 15.8%, when the pilot unit was operated at various pH values.
46 5.6 Hardness Reduction Table 7 indicates the reduction in Ca and Mg hardness which was expected as a result of the increase in operating pH. The reduction of hardness is secondary concern in CSD bottling operations. It is assumed that alkalinity reduction will reduce hardness. 5.7 Operating Flux The operating flux of 30 gfd was initially selected because most of the ultrafiltration membranes used in treating municipal operate at this flux value, although Trisep recommendation is 25 gfd for treating municipal water supply, when dosing coagulants (such as ferric chloride or sulfate, alum and polyaluminum chloride). It was assumed that lime will behave like the other coagulants although there were concerns of excessive fouling and scaling. The results of this research confirmed that the immersed SpiraSep UF membrane can achieve the treatment objectives when operated at flux of 30 gfd, and fed with lime treated water at pH 9.8, with suspended solids concentration of 600 mg/L. The cleaning of the membrane or CIP was initiated when the TMP value was doubled, which correspond to about 50% of clean membrane permeability. The CIP was conducted after 19 days of operation, noting that the pilot unit has been in operation for over 30 days in the first and second phases before the third phase started. The third phase of the test also confirmed the following: the cleaning procedures and chemicals mentioned in Section 3.2.3 effectively restored the membrane to its starting TMP and permeability; by extrapolating the permeability and TMP profiles the expected next cleaning will be after 48 days. This corresponds to 30 days cleaning interval.
47 5.8 Chlorination During the entire duration of test, chlorine was not added to the back flush water or in the PFE. The residual chlorine in the feed water ranged from 0.2 to 0.7 mg/L. Chlorine was dosed only during CIP and when the unit was stopped longer than 24 hours. 5.9 Benefits of the Lime Softening Ultraf iltration (LSUF) Process to CSD Bottler The benefits of the Lime Softening Ultrafiltration Process to CSD bottler, based on the results of this study can be summarized in the following: There is considerable economic benefit when the conventional treatment processes comprising of chlorination, lime softening, clarification, and filtration, is replaced with LSUF comprising of a single equipment with smaller footprint. With less equipment, operation and maintenance will be simpler. The LSUF process requires shorter time for start-up, unlike conventional lime softening which requires time to build up sludge, stabilize the flow and attain the desired treated water quality. The LSUF process produces less sludge and dirty backwash water. It can be operated at relatively lo wer pH and with no addition of ferric chloride which significantly reduced the volume of sludge. The water during backflush operation can be returned back to the system. The water wasted is the water that goes with the waste sludge, which is minimal.
48 Continuous chlorination of raw water can be eliminated, reducing the formation of the THMs. Process control in LSUF reduced to adjustment of pH and flows. The process is less sensitive to temperature. In LSUF process, the sludge removal is simplified because there is no sludge blanket to maintain. The ultrafiltration process provides physical barrier for microorganism and particles, minimizing the contamination in the down stream processes. Existing lime softening plants can be retrofitted and their rated capacity can be increased with just the addition of the UF system processes.
49 References Bachelor, B. and M. McDevitt ,1984. An Innovative Process for Treating Recycled Cooling Water . Journal WPCF 56 10, 1110 1117. Bachelor, B., M. Lasala, M. McDevitt, and E. Peacock, 1991. Technical and Economic Feasibility of Ultra-High Lime Treatment of Recycled Cooling Water. Research Journal WPCF 63, 7, 982 990. Collins, M.R., G.L. Amy, and P.H. King, 1985 Removal of Organic Matter in Water Treatment. J.Env. Eng. 111:6: 850 864. Gould, B., 2004.,Personal Communication, Trisep Corporation, Santa Barbara, CA, Gould, B., 2003 Utilizing Spiral Wound Membrane Efficiency for Ultrafiltration by Enabling Backwash Configuration , Paper presented at 12th ACS Anniversary. Li, C., J. Jian and J. Liao, 2004. Integrating Membrane Filtration and a Fluidized-bed Pellet Reactor for Hardness Removal. J. AWWA, 96:8:151158. Liao, M.Y. and S.J. Randke, 1985. Removing Fulvic Acid by Lime Softening. J. AWWA 77:8:78 88. Humenick, M..J, 1977. Water and Wastewater Treatment. New York: Marcel: Decker, Inc. Morelli, Cliff D., 1994. Water Manual 3rd ed., New York, World Beverage. Permutit, 1961, Water and Waste Treatment Data Book, New Jersey: The Permutit Company. Scuras, S, M.R. Rothberg, S.D. Jones, and D.A. Alami, 1999. The Rothberg, Tamburini, & Windsor Model for Water Process and Corrosion Chemistry Version 4 Users Guide Denver, Colorado: Rothberg, Tamburini, & Windsor, Inc. Shachman, M., 2004. Soft Drink Companion: A Technical Handbook for Beverage Industry, New York: CRC Press.
50 Trisep, 2003. SpiraSep Ultrafiltration Membrane Technology Transfer Manual Trisep Corporation, Goleta, CA. US EPA, 1999. Enhanced Coagulation and Enhanced Precipitative Softening Guidance Manual EPA 815-R-99-012. Washington, DC.
52 Appendix A Pilot Unit Equipment Description 1.0 Pilot Plant Systems Parameters The pilot plant consist of the lime reactor and the UF system. The lime reactor was designed to suit the requirement of this research. The UF system is a full scale commercial unit with one (1) UF element. 1.1 Lime Reactor Retention time: 30 minutes Flash Mixing Chamber Retention: 30 60 seconds 1.2 Pilot Plant Process Flows Plant Capacity (Effluent): 0 7 gpm Membrane Flux Range: 0 60 gfd System Recovery: 90% 1.3 Aeration Aeration Flow Rate: 6.2 scfm 1.4 Membrane Backwash Frequency: 15 minutes Duration: 30 seconds Back Flush Flow Rate: 7.5 gpm Back Flush Water Volume Used per Backwash: 5 6 gallons Back Flush NaOCl Dosage Concentration: 10 mg/L 1.5 Periodic Flux Enhancement (PFE) PFE Back Flush Flow Rate: 1.8 gpm PFE Water Volume Used per PFE: 35 gallons NaOCl PFE Frequency: 24 48 hours Citric Acid PFE Frequency: 3 days PFE Back Flush Length: 10minutes PFE Static Soak Length: 10 minutes
53 Appendix A: (Continued) NaOCl PFE Dosage Concentration: 100 mg/L Citric Acid PFE Dosage Concentration: 0.1% 1.6 Clean-In-Place (CIP) CIP Backwash Flow Rate: 1.8 gpm CIP Tank Volume per Manifold: 75 gallons NaOCl CIP Cleaning Frequency: 3 months CIP Duration: 4 8 hours NaOCl CIP Dosage Concentration: 2,000 mg/L NaOCl CIP Concentration: 0.1% Citric Acid CIP Cleaning Frequency: 3 months Citric Acid CIP Concentration: 1.0% 2.0 Equipment Specifications 2.1 Lime Reactor Type: Cylindrical with conical bottom Capacity: 200 gallons Materials of Construction: PE Mixer: 1/20 hp 2.2 Ultrafiltration Membrane Model: SpiraSep 900 Chemistry: PES Quantity: One (1) Element Diameter: 9.38 inches Element Length: 42 inches 2.3 Aeration Manifold Material: Schedule 40 PVC Size: 1 inch Schedule 40 PVC
54 Appendix A: (Continued) 2.4 Membrane Tank Quantity: One (1) Material: PVC Height: 60 inches Water Level: 54 inches Diameter: 18 inches Effective Volume: 60 gallons 2.5 Filtrate Storage Tank Quantity: One (1) Material: PE Volume: 75 gallons 2.6 Filtrate Pump Quantity: One (1) Pump Type: Self-priming centrifugal Model: Flotec FP5162 Construction: Noryl wetted parts Process Piping: SCH 40 PVC Control: Manual Throttle Valve Capacity: 10 gpm @ 20 feet suction lift Pump Power: 0.75 hp Power: 115/230 VAC, 1 phase, 60Hz 2.7 Backwash Pump Quantity: One (1) Pump Type: Centrifugal Model: American Stainless SSPC1 Construction: 316 SS wetted parts Process Piping: SCH 80 PVC, 316 Control: VFD Capacity: 10 gpm @ 10.0 psi discharge Pump Power: 1.0 hp Power: 230/460 VAC, 3 phase, 60Hz
55 Appendix A: (Continued) 2.8 Blower Quantity: One (1) Quantity per Train: One (1) Blower Type: Regenerative (oil-less) Model: Ghast or equivalent Construction: Carbon Steel Process Piping: Galvanized Steel Control: Manual Throttle Valve Capacity: 10 scfm @ 2.5 psi discharge pressure Blower Power: 1.0 hp Power: 230/460 VAC, 3 phase, 60Hz 2.9 Low Capacity Chlorine Metering Pump Quantity: One (1) Pump Type: Positive Displacement Model: LMI Wetted Ends: Polypropylene with PVC Diaphragm: PTFE Balls: Ceramic Capacity: 0.2 gpd Controller: Manual Power: 115/230 VAC, 1 phase, 60Hz 2.10 High Capacity Chlorine Metering Pump Quantity: One (1) Pump Type: Positive Displacement Model: LMI Wetted Ends: Polypropylene and PVC Diaphragm: PTFE Balls: Ceramic Capacity: 14 gpd Controller: Manual Power: 115/230 VAC, 1 phase, 60Hz
56 Appendix A: (Continued) 2.11 Sodium Hydroxide Metering Pump Quantity: One (1) Pump Type: Positive Displacement Model: LMI Wetted Ends: Polypropylene and PVC Diaphragm: PTFE Balls: Ceramic Capacity: 10 gpd Controller: Manual Power: 115/230 VAC, 1 phase, 60Hz 2.12 Citric Acid Metering Pump Quantity: One (1) Pump Type: Positive Displacement Model: LMI Wetted Ends: Polypropylene and PVC Diaphragm: PTFE Balls: Ceramic Capacity: 10 gpd Controller: Manual Power: 115/230 VAC, 1 phase, 60Hz 2.13 PLC/Control Panel Quantity: One (1) Model: Automation Direct DL 06 Power Input: 230/460 VAC, 3 phase, 60 Hz Enclosure: NEMA 12, Carbon Steel Operator Interface: LCD with push buttons 2.13 Instrumentation Level Switches: 4, NOC Float Switches Pressure Gauges: 2, Ashcroft or equivalent Quantity Rotameters: 4, Blue-White or equivalent Air Rotameters: 1, Blue-White or equivalent Temperature Gauge: 1, Cole Palmer or equivalent
57 Appendix A: (Continued) pH Meter : 1, SIGNET Turbidimeter 1, HACH 2.14 Piping and Automated Valves Automated Ball Valves: Four (4) Manual Globe Valve: Four (4) Manual Ball Valves: Four (4) Piping Material: SCH 40 PVC
58 Appendix B SpiraSep Trans-membrane Pressure (TMP) Measurements 1. 0 TMP Measurement The TMP of the SpiraSep system can be calculated by the following equation: TMP = Pva c + Htank Hct (1.1) where Pvac = vacuum pressure Htank = hydrostatic pressure in tank Hct = hydrostatic pressure in membrane core tube Since Htank = Hct, the TMP is equal to the vacuum pressure. Equation (1.1) now becomes: TMP = Pvac (1.2) 2. 0 Pressure Gauge Location The height of the pressure gauge location should even with the water level inside the membrane tank, as this will indicate the true trans-membrane pressure. It is important to account for any hydrostatic pressure losses/gains in the suction pipe when measuring TMP. Although the hydrostatic pressures inside the membrane tank and element core tube cancel each other out, the hydrostatic pressures in the suction line leaving the element must be accounted for. Below are several different scenarios on TMP measurement based on gauge/sensor location.
59 E-1 E-2 PPressure Gauge Permeate Pump H Appendix B: (Continued) Scenario 1 TMP = Pvac H, where Pvac is the pressure measured by the gauge/sensor. For example, if the pressure measured in scenario 1 by the pressure sensor is -1.5 psi and H is equal to 24 inches, then the TMP is equal to -2.37 psi (-1.5 minus 0.87). Scenario 2
60 Appendix B: (Continued) TMP = Pvac + H, where Pvac is the pressure measured by the gauge/sensor. For example, if the pressure measured by the pressure sensor is -2.0 psi and H is equal to 12 inches, then the TMP is -1.57 psi (-2.0 plus .43).