Final Report: Arsenic Removal System for Residential and Point-of-Use ApplicationsEPA Contract Number: 68D03062
Title: Arsenic Removal System for Residential and Point-of-Use Applications
Investigators: Turchi, Craig S.
Small Business: ADA Technologies Inc.
EPA Contact: Manager, SBIR Program
Project Period: October 1, 2003 through December 31, 2004
Project Amount: $225,000
RFA: Small Business Innovation Research (SBIR) - Phase II (2003) Recipients Lists
Research Category: Small Business Innovation Research (SBIR) , Water and Watersheds , SBIR - Water and Wastewater
Arsenic contamination in groundwater represents a severe health risk to populations throughout the world. In the United States, the problem is most pronounced in the West, parts of the Midwest, and the Northeast. In response, the U.S. Environmental Protection Agency (EPA) announced a tougher drinking water standard for arsenic, lowering the standard from 50 ppb to 10 ppb. This change is expected to impact 10 percent of the nation’s community drinking water systems. Although several technologies are readily amenable to incorporation in large water treatment processes, fewer options are available for small water systems, particularly those serving less than 500 users. Point-of-entry/point-of-use (POE/POU) devices are approved small system compliance technologies for meeting the revised arsenic maximum contaminant level (MCL). The primary advantage of POU treatment in a small system is reduced capitol and treatment costs, relative to centralized treatment.
The goal of this research project was to demonstrate a complete POU/POE drinking water system. The unit combined a highly effective arsenic sorbent with a metered cartridge that tracks cumulative flow and automatically stops flow at a predetermined volume limit. ADA Technologies, Inc. (ADA) also investigated low-cost arsenic sensor designs with the goal of providing a device capable of monitoring system performance. The system provided easy-to-maintain hardware for individual home use or deployment in small, centrally managed water systems.
During the Phase II portion of this research project, ADA evaluated multiple sensors consistent with the proposed objective of online monitoring of the performance of POU/POE drinking water systems. One approach was an audible alarm designed to mate with the POU hardware sold by project partner Kinetico Incorporated (Newbury, OH). ADA developed and tested a sensor that alarms at a predetermined flow volume (set in conjunction with the unit’s shut-off volume), effectively warning the user that a new filter is needed soon. Although the tested unit only provided a local alert, remote notification could be incorporated.
ADA also investigated the potential of two electroanalytical arsenic sensors. The first involved anodic stripping square wave voltammetry (ASSWV). ADA demonstrated that both common forms of aqueous arsenic, arsenite (As3+), and arsenate (As5+) can be detected to ppb limits using ASSWV. As3+ was more easily detected, with a lower detection limit of 1 ppb compared to a detection limit of 60 ppb for As5+. This was believed to be attributable to the fact that the prerequisite reduction step of As5+ to As0 is kinetically slower than the As3+ to As0 process. A longer deposition time will be needed to enhance As5+ deposition and detection.
The second sensor technology was tested by team member Brims Ness Corporation (South Portland, ME) and utilized their quartz crystal microbalance (QCM) technology. Phase II activities centered on finalizing the sensor’s hardware, the selection of the arsenic-receptor coating for the QCM, and testing of the sensor with As5+ and phosphate in various challenge water compositions. Brims Ness made significant developments improving the functionality of the QCM sensor, leading to a contract between Brims Ness and the University of Maine-Orono to proceed with fabrication of commercial prototypes.
Despite progress made with both arsenic detection methods, neither sensor was sufficiently developed for use in the Phase II field testing. Both entities (ADA and Brims Ness) are continuing their development efforts to convert these promising technologies into viable field analyzers.
Sorbent Selection and POU System Testing
Early work was focused on determining if ADA’s Amended Silicate™ arsenic sorbent was compatible with a commercially available POU water treatment system developed by team member Kinetico, thus providing a direct path into the POU market. The testing showed that use of the Amended Silicate™ sorbent would require a redesign of Kinetico’s existing MACguard™ (MAC) POU product line. Although technically feasible, such an effort could not be economically justified because it would require modification of Kinetico’s production and inventory systems. Therefore, the iron-based sorbent UltrAsorb-F was evaluated in the POU systems. UltrAsorb-F is a commercially available iron hydroxide media produced by Engelhard (Iselin, NJ) as ARM-200. The media behaves similarly to other iron-based sorbents and is American National Standards Institute (ANSI)/National Sanitation Federation (NSF) Standard 61 approved.
ADA carried out standardized testing to evaluate the performance of the MAC system. Three POU units were exposed to different formulations of standardized arsenic challenge water. The water composition was based on the mean values for common water constituents found in regions that had arsenic concentrations above the MCL of 0.010 mg/L. This challenge water is in ANSI/NSF Standard 53 for adsorption-based water treatment devices. The prototypes were tested at pH levels of 6.5 and 8.5 using 0.050 mg/L As5+, and at a pH of 8.5 using 0.050 mg/L As3+. This test strategy provided performance data at the most common test conditions and gave information about the ability to capture the less common but more toxic As3+.
After testing was complete, the cartridges were frozen, cut into 1-inch sections, and sent to Acme Analytical Laboratories (Vancouver, British Columbia, Canada) for arsenic loading analysis. These data were used to determine the mass-transfer front within the sorbent column. Analysis of the front indicated the capacity of the sorbent (in the saturated zone), the length of the mass-transfer zone, and the corresponding column capacity to breakthrough. Breakthrough was defined as the number of bed volumes treated until the effluent arsenic concentration exceeded the 10 ppb MCL. The column breakthrough curves were calculated using the empirical Freundlich Equation, q = KCl/n, coupled with isotherm data and spent sorbent digestion data. In the equation, q is sorbent capacity (obtained by arsenic analysis of the media), C is the equilibrium/effluent arsenic concentration (calculated), and K and n are sorbent-dependent parameters (obtained through isotherm testing at ADA).
Figure 1 depicts the calculated breakthrough curves for the standardized testing. These data trends are typical of iron-based sorbents; that is, the media performed better when exposed to As5+ compared to As3+, and the capacity of the media decreased as the pH increased. Using the calculated breakthrough curves, it was estimated that the MAC system loaded with 500 g of sorbent can treat 1,750 gallons of water when exposed to challenge water at a pH of 6.5 and spiked with As5+, 1,130 gallons at a pH of 8.5, and 470 gallons when spiked with As3+ at a pH of 8.5.
Figure 1. POU breakthrough curves calculated using the empirical Freundlich Equation coupled with spent sorbent digestion data and isotherm equilibrium values. Data points represent actual data obtained during standardized column testing. A MAC Cartridge holds approximately 500 g of sorbent.
Standardized testing determined the capacity of the MAC system under controlled, standard conditions, and these data are important to benchmark the technology versus other treatment processes. Performance in the real world, however, is the ultimate measure of a system’s technical and commercial success.
ADA examined the POU system at two locations in Colorado where the arsenic concentration was above the MCL of 10 ppb. ADA selected a ranch southwest of Gunnison, Colorado (Rock at Ute Trail River Ranch) and Adams State College in Alamosa, Colorado, for field testing. Analysis of the tap water at Ute Trail River Ranch indicated that the total arsenic concentration was 15.2 ppb, with 14 mg/L silica at a pH of 7.1. The water composition at Adams State had an arsenic concentration of 38 ppb, with 112 mg/L silica at a pH of 7.4.
Five hundred gallons were run through the cartridge at Ute Trail River Ranch prior to its automatic cut off. Water samples showed that the effluent remained well below the 10 ppb limit during this time. The recovered cartridge was cut into 1-inch sections and each section was assayed for total arsenic. Again, these data were used to determine the mass-transfer front within the sorbent column (see Figure 2). Analysis of the front indicated the capacity of the sorbent (in the saturated zone), the length of the mass-transfer zone, and the corresponding column capacity to breakthrough. It was estimated, using a technique similar to the standardized tests, that the POU system could treat 1,100 gallons of groundwater at this site before filter replacement was required.
Figure 2. Rock at Ute Trail River Ranch arsenic loading (µg-As/g-sorbent, ppm) as a function of column section (left) and corresponding diagram of column sections (right).
An identical POU system installed at Adams State treated only 200 gallons of water before breakthrough (see Figure 3). The silicate concentration in the tap water at this site (112 mg/L) ranks with the highest in the country, a level that presents challenges for any adsorption system. Silicate is a known interferent for arsenic adsorbents and was thought to contribute to the low capacity of the POU filter at this location.
Figure 3. Breakthrough curve generated during field testing at Adams State College. The relatively low arsenic capacity is thought to be because of the extremely high silica concentration (112 mg/L) in the tap water.
The results from the two field tests clearly illustrate the difficulty of presetting the shut-off point for a POU adsorbent cartridge. Although the approach has obvious merit, differing water quality necessitates a conservative set point, thereby wasting media and increasing maintenance labor in most applications. Two methods to alleviate this limitation are providing arsenic effluent detection or creating a better sorbent performance model based on water composition. Both of these approaches have their own issues. Although the former increases system cost, the latter assumes a constant water quality.
The disposal of spent sorbents and the ultimate fate of contaminants is a key concern when evaluating the life cycle costs of treatment technologies. For adsorbent media, such as those used in this study, the hazard level is determined by standardized leaching tests. The two relevant protocols for arsenic sorbents are EPA’s Toxicity Characteristic Leaching Procedure (TCLP) and the similar California Waste Extraction Test (WET). Both methods consist of tumbling the arsenic-laden sorbent in an acidic solution for a fixed time period. The sorbent then is filtered from the solution and the filtrate is analyzed for arsenic.
Sorbent samples from the standardized testing performed at ADA’s laboratory were subjected to both the TCLP and WET leaching tests. Leachates from the TCLP tests were considerably lower than the leachates from the WET test; however, the leachates of both tests were at least 7.5 times lower than the current allowable soluble threshold limit of 5,000 ppb.
The goal of this research project was to demonstrate a POU/POE arsenic removal system that combined an effective arsenic sorbent with a metered cartridge that totalized and automatically stopped flow at a predetermined volume limit. User alert functions and arsenic sensor technologies were explored to augment the cumulative flow cut-off system.
The work used Kinetico’s existing MAC cartridge system as the basis for the POU system. Early work in the project was focused on determining if ADA’s Amended Silicate™ arsenic sorbent was compatible with this POU water treatment system. This work concluded that the unique physical and hydraulic properties of the Amended Silicate™ sorbent, in particular the relatively low bulk density of Amended Silicate™, were not amenable for use with Kinetico’s MAC cartridge. Consequently, the commercially available, iron-based UltrAsorb-F was used for field testing the POU systems.
During Phase II, the ADA team evaluated multiple sensors consistent with the proposed objective of online performance monitoring. One method ADA explored was an audible alarm that mates with the MAC’s flow totalizer. ADA’s sensor would alarm before filter replacement is required, effectively warning the user that a new filter is needed. The unit provides a local alert with the option for remote notification. Although successfully tested, Kinetico did not think the POU market would accept the additional cost of this alarm.
ADA also investigated the application of two forms of arsenic sensors. ADA demonstrated that both As3+ and As5+ can be detected to ppb limits using ASSWV. As3+ is more easily detected, with a lower detection limit than As5+ . This work is continuing at ADA, as the team strives to produce a robust sensor for field use.
The second approach, championed by project team member Brims Ness, utilized an inexpensive QCM technology. In Phase II, Brims Ness was able to advance the design of the sensor hardware, identify a superior arsenic-receptor coating for the QCM, and test the sensor with As5+ and phosphate in various challenge water compositions. Brims Ness made significant developments improving the functionality of the QCM sensor, leading to a contract between Brims Ness and the University of Maine-Orono to proceed with fabrication of commercial prototypes.
Phase II testing demonstrated that the MAC system was capable of removing arsenic below the MCL without mechanical troubles. The novel automatic shut-off feature provides low-cost, maintenance-free insurance that the cartridge will not be overused. The system, however, still relies on a predetermined set point (generally conservative) for system shut-off. This limitation could be resolved by an effective online arsenic monitor, but it appears that the size and price sensitivity of the POU market will not justify the additional expense. Still, POU arsenic removal systems could serve as an alternative to centralized water treatment plants for individual private wells and small water systems. The primary advantage of employing POU treatment in a small system is reduced capital and treatment costs. For example, it was estimated that the equipment costs for POU units installed at 100 taps with water quality similar to the Ute Trail River Ranch would be approximately $150 for hardware, plus a filter replacement cost of approximately $50 every 2 years. At present, however, these advantages are offset by the reporting and tracking requirements placed on the water provider. Given the reluctance of utilities to undertake the tracking necessary for deployment of POU systems, it appears the POU market will be relegated to individual well owners, and centralized systems will continue to be the technology of choice for small system operators.