Role of Zeta (ζ) Potential in the Optimization of Water Treatment

Joint Task Force Centers for Disease Control and Prevention and U.S. Environmental Protection Agency (EPA), Environmental Health Needs and Habitabilit...
0 downloads 0 Views 774KB Size
Ind. Eng. Chem. Res. 2009, 48, 2305–2308

2305

Role of Zeta (ζ) Potential in the Optimization of Water Treatment Facility Operations Ana Morfesis,*,†,‡ Annette M. Jacobson,‡ Rosemary Frollini,‡ Matthew Helgeson,‡ Judy Billica,§ and Kevin R. Gertig§ MalVern Instruments, Inc., 117 Flanders Road, Westborough, Massachusetts 01581, Department of Chemical Engineering, Colloids, Polymers and Surfaces Program, Carnegie Mellon UniVersity, Pittsburgh, PennsylVania 15213, and Water Treatment Facility, City of Fort Collins, 4316 LaPorte AVenue, Fort Collins, Colorado 80521

Zeta (ζ) potential provides a measurable value to monitor optimal water clarification capabilities. Its value indicates the repulsive interaction between particles; a zero ζ potential means that the conditions for the aggregation of contaminants are maximized. Results from model experiments and data from a full-scale water treatment process at the Fort Collins Water Treatment Facility (FCWTF) will be reviewed. Optimal conditions for water production controls can be maintained by monitoring ζ potential, turbidity, and flocculating agent concentration, which is especially important during seasonal fluctuations of raw water input. Introduction Drinking water in the U.S. and developed nations of the world is treated to remove contamination of foreign materials, both mineral and organic. Contaminants can be present in almost any water source and can be observed in water supplies obtained from mountains, rivers, and soil with even greater contamination potentials coming from domestic sewage and industrial waste. Natural disasters have highlighted the need for the improved monitoring of water clarification procedures. During the aftermath of Hurricane Katrina in 2005 the Centers for Disease Control (CDC) and the Environmental Protection Agency (EPA) reported abnormally high concentrations of Escherichia coli (E. coli), coliform bacteria in the water supply.1 The World Health Organization reported a broad spectrum of microbiological contamination as well as additional problems from chemical and increased salinity contamination after the Indian Ocean Tsunami in 2004.2 Particulate and microbiological contamination issues in water supplies around the world are not just specific to natural disasters; surface water is always most vulnerable, while poorly constructed and shallow wells are also at risk.3 One of the requirements for removing or inactivating particulate contaminants as well as biological pathogens is to enhance the coagulation capabilities in water treatment facilities. Impurities in wastewater are primarily anionic, negatively charged species. Cationic (positively charged) additives have been developed to neutralize these charged anionic contaminants. By introducing optimal quantities of cationic additives that will cause the system to reach a neutral charge, it is possible to improve the efficiency of removing contaminants via sedimentation or filtration. ζ potential measurements provide a tool to quantify the optimal concentration of the oppositely charged additives required to enhance coagulation of contaminants from a water supply.3-7 In the case of water treatment applications, charged particles should always be considered to be stable particles and therefore undesirable. The objective is to reduce the surface charge of * To whom correspondence should be addressed. Tel.: 412-268-1725. Fax: 412-268-7139. E-mail: [email protected]. † Malvern Instruments, Inc. ‡ Carnegie Mellon University. § Water Treatment Facility, City of Fort Collins.

the contaminants by treating them with oppositely charged additives to the point where the charge of the contaminating material is close to zero or neutral. These oppositely charged additives (cationic flocculating agents in this case) will adsorb to the surface of the (anionic) particles. When the particle/ additive complexes have little or no charge, the ζ potential approaches zero (termed point of zero charge) and there will then be no repulsive forces maintaining the stability of these particles in the dispersion leading to their aggregation and sedimentation.4-7 ζ potential measurements, used in conjunction with turbidity measurements, provide a method to (1) monitor the particle stability which varies over time and (2) identify maximum aggregation and flocculation conditions.7 Since raw (input) water quality can vary significantly with time due to many factors previously discussed, these measurements provide a method to continuously assess process parameters that indicate water quality so that a quick response can be made when high raw water variability occurs. It should be noted that the case study below presents results from a full-scale water treatment facility (FCWTF) and provides a year’s worth of data monitoring the ζ potential, turbidity measurements, and alum dosage in the water treatment process. These results indicate the importance of using ζ potential and turbidity to control the process and highlight the potential for reducing or eliminating the need of 2 hour jar test results. Theory ζ potential is a measure of the electrical potential that causes interparticle repulsion. It has been used in the water treatment industry for many years to help determine coagulant dosages.8,9 Dentel10 provides references to work by others that indicate a variety of ζ potential ranges for optimum turbidity removal, including coagulated waters with ranges of -4 to +3, -5 to +5, and -13 to +13 mV. Bench and pilot studies have shown that operating within the ζ potential range of -10 to +5 mV can minimize contaminants measured as total organic carbon (TOC) residuals.10-14 Generally, compounds contributing to TOC levels are comprised of humic substances, which are naturally occurring polyelectrolyte substances of low to moderate molecular weight. These substances are derived from soil

10.1021/ie800524x CCC: $40.75  2009 American Chemical Society Published on Web 12/01/2008

2306 Ind. Eng. Chem. Res., Vol. 48, No. 5, 2009

Figure 1. ζ potential (mv) and turbidity (NTU) of bentonite/water vs dosage of flocculant (ppm).

Figure 2. Expanded scale of a water treatment control chart during typical seasonal input water variations. ζ potential (mV) and settled turbidity (NTU) in a single train/settling tank monitored over time.

and produced within water and sediments by processes occurring during the natural decomposition of vegetation.4-6 High TOC and particulate levels exert a very strong demand on flocculating agents such as alum. Therefore, narrowing the operating ranges of any water production facility (in general) must be site specific and must be determined empirically.5,15 Experimental Procedures To demonstrate the usefulness of turbidity and ζ potential measurements to determine optimal flocculant dosage requirements, a model experiment was conducted. Simulated contaminated water was prepared by dispersing bentonite clay (Fisher Scientific, laboratory grade) particles into tap water at a

concentration of 200 ppm using a high-speed mixer. Bentonite is an aluminum-silicate clay that when dispersed in water forms a negatively charged colloidal dispersion. In the flocculation experiment, 250 mL volumes of the bentonite dispersion were dosed with a polyelectrolyte flocculating agent, Cat Floc T (Calgon Corp.) in doses ranging from 0.01 to 150 ppm. The dispersions were mixed on a gang stirrer at 100 rpm for 3 min, 50 rpm for 5 min, and 20 rpm for 5 min and then left to settle for 10 min. A 20 mL sample was removed from the top half of the settled sample; turbidity and ζ potential of this sample were determined. Turbidity (in NTU) was monitored using a Hach Model 2100A turbidimeter. ζ potential measurements were carried out using either a Zetasizer Nano

Ind. Eng. Chem. Res., Vol. 48, No. 5, 2009 2307

Figure 3. Water treatment production control chart in 2002 displaying ζ potential (mV) and alum dosage (mg/L) across four trains (settling tanks) during a 1 year period at FCWTF.

ZS instrument with MPT-2 automated titrator or a Zetasizer 2000 HSA instrument both manufactured by Malvern Instruments, Ltd., Worcestershire, U.K. Fort Collins Water Treatment Facility (FCWTF) uses a conventional treatment process, including alum coagulation, flocculation, settling, and filtration. Flocculation and settling take place in four parallel trains with design flows to each train of about 20 million gallons per day (MGD). The flocculation portion of each train is characterized by serpentine plug flow with variable-speed, horizontal paddle flocculators. A high molecular weight cationic organic polymer is added during flocculation to improve the size, strength, and density of the floc. Settling in one of the trains is aided by tube settlers, while settling in the other three trains is aided by lamella plates. A polymer filter is added to the settling basin effluent, prior to filtration by 23 dual media, constant-rate filters.7 Alum (Al2(SO4)3 · 14.3H2O, 48.5% by weight solution strength) feed uses a flow paced control. Alum is added at in-line mechanical mixers downstream of a CO2 feed point. The inline mechanical mixers consist of vertically oriented, constantspeed impeller type mixers operating within a pipe spool. Each mixer is equipped with dual axial flow impellers to generate a high mixing intensity. The design velocity gradient is 1000 s-1. In-line mechanical mixers are used at the FCWTF to provide for the near instantaneous mixing of alum that is necessary for the coagulation reaction.7 The alum dose has historically been selected by the FCWTF operators based on a number of indicators including jar test (sedimentation) results, ζ potential data, streaming current monitors, settled and filter effluent turbidities, settling characteristics of the floc, filter run times, residual aluminum and raw water TOC, color, and turbidity.7 However, during specific times of the year, because of natural weather conditions, water quality can fluctuate quickly, and at these times rapid decisions about coagulant dosages must be

made. Specifically, 2 hour jar tests conducted in the morning may not be representative of the water that is coming into the plant by afternoon. Therefore it is the frequent use of ζ potential measurements that provide critical results for determining the alum dosage at the FCWTF particularly during times of rapidly changing water conditions.7 Water production data were measured in the same manner as the laboratory experiments; however, grab samples were obtained during each 8 h shift and each train of the FCWTF throughout 2002. Results By studying the concentration of flocculant additive dosage vs ζ potential and turbidity model experiments, it can be seen that when the ζ potential of the dispersion results in a nearzero ζ potential, the turbidity, or cloudiness, of the mixture is also nearly zero. Figure 1 indicates that at low dosages of flocculating agents, the ζ potential is approximately -16 mV and the turbidity is high. At the point where the ζ potential value becomes +2 mV, the turbidity is at an absolute minimum. After this point in nearly obtaining a zero charge and turbidity, further addition of flocculating agent now reverses and increases the charge of the contaminating material and therefore restabilizes these particles in the water. The results shown in Figure 1 verify that there is an optimal concentration (10 ppm) for addition of flocculating materials, and the measurement that is critical to the control of the additive concentration is the ζ potential. Analysis of ζ potential vs coagulant dosage results is used to evaluate the effectiveness of various chemicals (typical examples are alum and ferric sulfate, etc.) and a number of commercial brands of cationic polymers (these usually contain a quaternary ammonium surfactant that imparts the cationic surface charge). Knowledge of ζ potential is important to adjustment of coagulant dosage levels periodically in order to minimize the cost of chemicals for water purification.

2308 Ind. Eng. Chem. Res., Vol. 48, No. 5, 2009

In the case of the FCWTF, the concentration of the flocculating agents (alum in this case), the particle ζ potential, and the turbidity are monitored and adjusted during each work shift, each day. Figures 2 and 3 provide a view of the water treatment facility’s control charts obtained over a 1 year period in 2002. Figure 2 has an expanded scale for turbidity and ζ potential to demonstrate how these values fluctuate in a typical train/settling tank during a time of seasonal input water fluctuations. It is important to note the range of values on the axes of Figure 2: ζ potential ranges from -6 to +6mV and turbidity from 0 to 1.2 NTU. Water clarity requirements, i.e., turbidity below 1.0 NTU, for this treatment process are achieved when particle ζ potential values range from -5 to +5 mV. By careful monitoring of the particle ζ potential and turbidity values over time, adjustments to the flocculating agent dosage can be made in a rational and optimal manner even during peak variations in raw water input quality. Figure 3 shows data for ζ potential and alum concentration during a 1 year time period in the four settling tanks across the FCWTF. Further analysis of the data shown in Figure 3 indicates that seasonal changes do affect the process control. The goal of FCWTF is to maintain ζ potential results of their water production at approximately zero, specifically between the limiting values of +5 and -5 mV (the dashed red lines represent the upper and lower process control limits). Because of requirements for thorough mixing of the flocculating agent with the incoming water supply, Figure 3 indicates a shift in the operational ζ potential range based on the changing water quality during what are seasonal (April through May) changes that occur naturally due to the raw incoming water supply at this location. Figure 3 indicates that during the April to May time frame water production operated more on the negative side of the ζ potential control range and this produced the best floc and best water quality during that time period. The alum dose was also increased during this time because TOC levels were also increasing. By the first week in June, TOC levels began decreasing, but as shown in Figure 3 the alum dose was kept constant during the month of June (about 23 mg/L). A decreasing TOC concentration during June combined with a constant alum dose resulted in an increase in the ζ potential (increased TOC levels exert a very strong demand on alum and vice versa; when TOC levels decrease, less alum is required). The alum dose could have been decreased in June, but, due to an operational decision, in this specific case, it was not. Therefore, the ζ potential stayed on the plus side. By July, TOC concentrations in the water supply had dropped significantly and were no longer a factor, so the alum dose (concentration) was decreased. During the late summer, FCWTF was operating on the negative side of the -5 to +5 mV ζ potential range; this produced the highest quality water at this time. However, by late fall and early winter, the operation was optimized on the positive side of the -5 to +5 mV range. In both events, this water production case study indicates that FCWTF remained within specification and in control of production efficiencies. Conclusions Particle ζ potential measurements can be successfully used in conjunction with turbidity measurements by plant operators as a key tool to help optimize the coagulation process and meet plant effluent goals. They are used as indicators of optimal

dosage for flocculating agents: ζ potential as an indicator of particle stability and charge and turbidity as an indicator of optical clarity and particle presence. Together they are excellent tools for determining the effectiveness of agents as well as optimizing the efficiency and economy of the operation of a water treatment facility. Frequent measurements of ζ potential have replaced the need to conduct jar tests to determine alum dose. ζ potential results are key to monitoring and maintaining optimal conditions in water treatment facilities, especially during upsets or naturally occurring seasonal changes in water production. Future challenges to consider include real-time ζ potential monitoring devices that could potentially improve future optimization of the water treatment process. In summary, ζ potential provides a process control parameter that indicates changes in raw water (input) chemistry which is essential for process optimization. Literature Cited (1) Joint Task Force Centers for Disease Control and Prevention and U.S. Environmental Protection Agency (EPA), Environmental Health Needs and Habitability Assessment, September 17, 2005. (2) Clasen, T.; Smith, L. The Drinking Water Response to the Indian Ocean Tsunami Including the Role of Household Water Treatment. Protection of the Human EnVironment: Water, Sanitation and Health; World Health Organization: Geneva, Switzerland, 2005. (3) AWWA Fact Sheet Cryptosporidium and Giardia in Water, Catalog No. 70107. (4) Amirtharajah, A.; O’Melia, C. R. Coagulation Processes, Destabilization, Mixing and Flocculation. In Water Quality & Treatment, A Handbook of Community Water Supplies, 4th ed.; McGraw-Hill: New York, 1990; Vol. 26, Chapter 6, pp 9-365. (5) Stumm, W.; Morgan, J. J. Aquatic Chemistry, 2nd ed.; WileyInterscience: New York, 1981. (6) Schroeder, E. E. Water and Waste Treatment; McGraw-Hill: New York, 1977. (7) Billica, J. A.; Gertig, K. R. Use of Zeta Potential to Optimize FullScale Treatment of High TOC Water, AWWA WQTC Conference Proceedings, 2006. (8) EVerything you wanted to know about coagulation and flocculation, 4th ed.; Zeta-Meter: Staunton, VA, 1993 (Apr); p 20. (9) Letterman, R. D.; Amirtharajah, A.; O’Melia,C. R. Coagulation and Flocculation. Water Quality & Treatment, A Handbook of Community Water Supplies, 5th ed.; McGraw-Hill: New York, 1999; Chapter 6, pp 6-58. (10) Dentel, S. K. Coagulation Control in Water Treatment. Crit. ReV. EnViron. Control 1991, 2 (1), 41–135. (11) Sharp, E. L.; Banks, J.; Billica, J. A.; Gertig, K. R.; Henderson, R.; Parsons, S. A.; Wilson, D.; Jefferson, B. The Application of Zeta Potential Measurements for Coagulation Control: Pilot Plant Experiences from UK and US Waters with Elevated Organics. Water Sci. Technol.: Water Supply 2005, 5 (5), 49–56. (12) Sharp, E. L.; Parsons, S. A.; Jefferson, B. Coagulation of NOM: Linking character to treatment, IWA Particle Separation Conference, Seoul, Korea, June 1-3, 2005. (13) Randtke, S. J. Organic Contaminant Removal by Coagulation and Related Process Combinations. J.-Am. Water Works Assoc. 1997, 89 (5), 64–77. (14) Owen, D. M.; Amy, G. L.; Chowdhury, Z. Characterization of Natural Organic Matter and its Relationship to Treatability, AWWA and AWWA Research Foundation, 1993. (15) Gregory, D. Enhanced Coagulation for Treating Spring Runoff Water. Opflow 1998 (Feb).

ReceiVed for reView April 2, 2008 ReVised manuscript receiVed August 12, 2008 Accepted August 28, 2008 IE800524X