Evaluating Technologies for Small Systems
Before installing a new water treatment system, water utilities must obtain approval from state drinking water regulators. Prior to granting approval, regulators may require pilot tests, depending on the technology to be installed. Package water treatment plants often use innovative designs to fit the treatment processes into compact units, and therefore regulators are often hesitant to approve them without detailed pilot testing. Pilot tests can last for periods of time as short as several weeks or as long as 1 year or more. Long programs of pilot testing add substantially to the costs of installing package plants for small systems. For example, one equipment manufacturer reported that pilot testing increased the capital cost of a treatment system by 28 percent (McCarthy, 1995).
This chapter discusses the degree to which testing of treatment technologies appropriate for small communities can be standardized. It describes when preexisting pilot test data or plant operating data are adequate to ensure performance of the technology at a new location and when site-specific testing is necessary. The chapter also discusses the availability of data on performance of water treatment technologies for small systems.
Before making decisions about treatment processes to employ and the extent of pilot testing that will be necessary, water system engineers will need to obtain information on raw water quality and desired treated water quality. Safe Drinking Water Act (SDWA) regulations specify the basic requirements for finished water quality. Because of customer or management preferences, water systems may also decide to add additional treatment (for example, water softening) not required under the SDWA. In some situations, the source water quality may be so high that the water meets all SDWA requirements and customer demands
without treatment. Where treatment is needed, in some cases water system engineers can select a treatment system based on performance data from other locations with water of similar quality or based on relatively inexpensive bench-scale tests. In other situations, however, available information on source water quality may provide convincing evidence that pilot testing for a particular treatment process is needed before a full-scale plant is installed. Thus, while current requirements for pilot testing of water treatment technologies lead to some duplication of effort and can be reduced, for certain combinations of treatment technologies and source waters, some degree of site-specific pilot testing always will be necessary to ensure that the equipment will perform adequately.
Current Requirements for Pilot Testing
Regulators require pilot tests in part to ensure that the water treatment system, whether package or custom designed, will effectively treat the water at the particular location. They regard such testing as especially important for surface water systems, for which water quality can be highly variable not only from place to place but also from season to season. For example, filter-clogging algae can appear in surface waters intermittently.
Because uncontrollable factors such as nutrient matter in the water and sunlight strongly influence algae growth, prediction and control of algae blooms is difficult if not impossible in the context of small system operations. Similarly, turbidity in some reservoirs and in many rivers varies for reasons—such as heavy rainfall and runoff, flooding, and heavy runoff from melted snow—that water utilities cannot control. Ground water tends to have more consistent quality than surface water, so, theoretically, site-specific testing is less important for ground water systems when performance data are available from other locations. However, even where the quality of the source water is relatively high and consistent, regulators usually require that package plants be pilot tested because of concern about the legitimacy of performance data provided by the manufacturers, which the regulators may perceive as a "sales pitch" (GAO, 1994). Regulators have indicated that independent, third-party evaluations of package devices are lacking, so the performance of package plants usually must be verified at each new location even when the manufacturer claims to have used the technology on water of similar quality elsewhere (GAO, 1994).
Requirements for pilot tests may vary significantly from state to state and even within a given state (WMA, 1994). For example, Illinois regulators always require a pilot study, usually lasting three seasons, for systems that treat surface water because of the high variability of surface water quality; for ground water systems they almost always require a 3- to 4-week pilot test. Similarly, New York regulators almost always require site-specific pilot studies prior to approving package systems. In Minnesota, conversely, regulators will approve package plants without pilot testing if the plants have been proven effective on waters of
similar quality at other locations. In Pennsylvania, engineers in six regional offices decide on the extent of pilot testing, and pilot testing requirements therefore vary within the state.
Pilot data collected in one state may not be considered valid in other states. For example, one equipment manufacturer reported that several states have refused to approve a technology that has operated effectively at more than 100 sites nationwide because of their reluctance to use data from other states (GAO, 1994). Seven western states (Alaska, California, Colorado, Idaho, Montana, Oregon, and Washington) attempted to encourage information sharing and to streamline their testing requirements for filtration systems, including package technologies, by developing a guidance document known as the Western States Protocol (GAO, 1994; David Clark, Washington State Department of Health, personal communication, 1995). The protocol applies to technologies not defined as "conventional" in the Surface Water Treatment Rule (SWTR), the U.S. Environmental Protection Agency (EPA) regulation that specifies filtration requirements for surface water. Examples of package technologies that could be evaluated under the Western States Protocol are bag filters and cartridge filters. However, rather than uniformly implementing the protocol, individual states have modified it to meet their specific needs. Although modification of a testing protocol perhaps could be justified on the basis of individual needs that vary from state to state, the overall effect of such modifications is to make the transfer of accumulated testing data from state to state difficult. An analogous situation would be if modifications were permitted in a standard EPA method for testing water quality, resulting in each state using a slightly different analytical approach for measuring water quality parameters such as turbidity and free chlorine concentration.
Establishing a Third-Party Certification Program
State regulators would likely reduce requirements for extensive piloting of package technologies on a case-by-case basis if equipment manufacturers could receive credible third-party certification of their products. A third party is a technically and otherwise competent body other than one controlled by the producer or buyer. Certification would provide assurance to decision makers that the product is capable of performing as advertised. (Of course, certification would not release the system from operation and maintenance activities to keep the equipment performing properly.)
Under a third-party certification program, manufacturers would voluntarily submit their equipment or processes to a certification agency for approval. Certification would include three key elements:
- verification of the manufacturer's claims, especially claims of reductions in contaminant levels;
- testing of construction materials to ensure that they are safe for contact
NSF International has limited standards for point-of-use and point-of-entry devices, drinking water treatment chemicals, and drinking water system components but none (yet) that specifically apply to individual package treatment systems. Two of the standards cover the ability of point-of-use and point-of-entry devices to improve aesthetic properties of the water and eliminate compounds that cause adverse health effects (see Chapter 3). Four cover specific point-of-use and point-of-entry devices for individual home use: ultraviolet systems, reverse osmosis systems, distillation systems, and cation exchange systems (McClelland, 1994). NSF standards for water treatment chemical additives and drinking water system components are aimed at ensuring that the treatment process itself does not create health hazards in the drinking water.
NSF uses expert committees to develop its technology standards. The primary committee developing a standard, known as the "joint committee," includes representation from industry, government, and consumer groups. This committee also receives input from a council of public health consultants and a certification council that has expertise in test methods. Once an NSF committee develops a standard, the NSF applies to have it certified by the American National Standards Institute (ANSI). An ANSI designation means that only one standard exists for that type of product in the United States and that the standard follows all of ANSI's guidelines.
Point-of-use and point-of-entry devices undergoing NSF review must meet requirements in four basic areas: (1) the equipment must meet manufacturers' claims for the level of contaminant removal provided; (2) the materials used in the equipment must be disclosed and tested for leachability; (3) the equipment must be tested for structural soundness; and (4) the equipment must have an adequate installation manual and must be accurately labeled.
- with drinking water and are capable of handling operating conditions for the expected life of the equipment; and
- evaluation of operation and maintenance manuals to ensure that they provide accurate and complete information about the equipment.
Third-party certification is currently available through the National Sanitation Foundation (NSF) International for a limited number of point-of-use and point-of-entry treatment devices and for certain water treatment chemicals and system components (see Box 4-1). However, certification is not yet available for package plants. In late 1995, the EPA launched a program, the Environmental Technology Verification Program, to test a wide range of environmental technologies, including package water treatment plants. The EPA has provided funding for NSF International to develop equipment performance verification protocols and test plans for evaluation of water treatment package plants. A key aspect of the program is the involvement of state regulators in protocol development. After the EPA and NSF International develop the protocols, third parties will test the technologies in a manner intended to facilitate acceptance by state regulatory
agencies while reducing the burden of repeated testing now faced by equipment manufacturers and vendors. The EPA funded the program with the intention of continuing support for 3 years, with hopes that testing fees will sustain most of the costs after this period and that testing will continue as manufacturers and entrepreneurs develop new treatment technologies.
Protocols for Technology Testing: Principles
Standard protocols for testing water treatment equipment exist for a variety of technologies, but they have not been collected in a common location. As a result, designers of water treatment systems conduct bench and pilot studies using their own individual methods. Whereas one experimenter may test for treatment efficiency using turbidity measurements, another may use a particle counter. These data cannot be directly compared, so they are essentially lost to the drinking water field after their initial use. Testing protocols specifying tests to perform, analytical procedures to follow, and a standard method for data reporting are essential for allowing interpretation and comparison of testing results from various sources and for eliminating unnecessary duplication in pilot testing. Whether testing is performed by the manufacturer, an engineering firm, a utility, or an approved third party, use of a standard testing protocol will aid in interpretation and acceptance of the data.
Protocols for technology testing need to be sufficiently comprehensive to ensure that the total needs of small water systems are considered when testing is carried out. The obvious requirement for water treatment processes is to produce a water quality that meets SDWA requirements and customer preferences. In addition, water systems need to look beyond the general capabilities of processes for contaminant removal and also consider process efficiency, operating ease, and operation and maintenance expenses. Installation of technology that is too complex to operate or too expensive to maintain is likely to result in regulatory compliance problems in the long run. Process efficiency and operating ease are related to factors such as length of filter run, extent of on-site operator attention needed, and extent of pretreatment required. Operation and maintenance expenses are influenced by factors such as chemical dosages, volume of water treated before system components need replacement, and energy requirements. Data on such aspects of water treatment need to be included in testing protocols.
Testing should be performed in a laboratory or pilot facility owned by the manufacturer and certified by a qualified third party or owned by a third party and having a national laboratory certification. The certification should come from the EPA, NSF International, American National Standards Institute, or another nationally recognized organization. National licensing of the laboratory is necessary to ensure that state regulators, and ultimately the organization that certifies the technology, will accept the data.
Tests must be performed at least in duplicate for each operating condition.
All analytical methods should follow a nationally accepted format such as those outlined in Standard Methods for the Examination of Water and Wastewater (APHA, AWWA, and WEF, 1992). Testers should measure and report initial and final concentrations for every regulated contaminant for which the equipment manufacturer claims a concentration reduction.
Evaluators should test the systems at the low end of the expected water temperature range and also at the high end if a high temperature can have any adverse effect. They should also perform tests at the high and low end of the recommended operating pressure and flow range. The protocol should require reporting of the full range of data collected.
Information should be obtained on raw water quality or on water quality before some process modification and on treated water quality or the quality of the water after some process modification. During the study, information on operating parameters such as water flow rate, chemical feed rates, operating pressure, contact time, filter bed lead loss, mixer or flocculator speed, backwash cycle time, empty bed contact time, air flow rate, percent of water rejected by a membrane, and so on should also be collected. As a general principle, if an adjustment can be made to process equipment to change its operation, that aspect of the equipment operation should be noted, and it might become a variable in the testing program. If equipment operation changes but the tester fails to note such changes, performance might seem erratic without any obvious reason.
Numerous kinds of miscellaneous data may be of use. Weather observations, river flows, lake or reservoir levels, industrial or municipal waste discharges, changes in equipment at the treatment plant, or repairs and maintenance in the distribution system might have impacts on various sorts of pilot studies. A good rule of thumb is to write down any observation that might be of value later. This at first seems burdensome, but thoroughly collected data can be used later to explain results, and if no problems develop in testing then some observations may prove unnecessary and would not have to be archived. Equipment performance data, void of supporting background information, are not likely to be accepted for waters other than the source water involved in the test program.
Finally, the protocol should require that results be measured and reported directly in the units specified in the regulatory language. For example, a particle-counting measurement cannot substitute for a turbidity measurement. Thus, the common language of testing protocols must be the language of the drinking water regulations.
Technology-Specific Testing Requirements
The degree to which test results from one location can be applied elsewhere, and therefore the extent to which third-party certification can reduce piloting requirements, depends on the technology and source water. Some types of equipment, such as chemical feeders, can be designed and tested, and all chemical
feeders of identical size and design should perform the same under identical equipment operating conditions. Similarly, aeration systems, disinfection processes, bag filters, and cartridge filters rarely require pilot testing. Performance of other types of equipment, such as a package plant employing coagulation and filtration, adsorption, membrane filtration, or lime softening, depends on the quality of the source water. For such systems, some degree of site-specific evaluation may be essential, as explained in detail below. Site-specific testing can range from bench-scale evaluations to operation of a pilot plant for a given time period. The extent of site-specific testing required depends on the amount of data available for the same or similar source waters.
The performance of aeration systems can be predicted with design equations (for example, see Kavanaugh and Trussell, 1980). Therefore, pilot testing should not be required for a well-designed system. The design equations are based on data from hundreds of installations using a variety of source waters. When used with a safety factor, they allow engineers to determine without site-specific testing the size of the units needed for removal of volatile organic compounds, radon, carbon dioxide, and natural compounds that cause taste and odor problems. Designers must pay careful attention to possible foulants, such as reduced iron and manganese, commonly found in ground water.
Several approaches are possible for evaluating membrane processes. Which approach is appropriate depends on the nature of the source water and the principles by which the particular membrane removes contaminants from water.
For ground water systems using membranes, the only site-specific analysis that may be necessary is evaluation of source water characteristics to determine the potential for chemical scaling of the membranes. Scaling is especially a concern for brackish ground water.
Assessing the treatability of surface waters by membranes is somewhat more complex. The capability of membrane processes to cope with particulate matter in raw water is limited. Thus bench-scale testing may be necessary to determine whether pretreatment is required to remove a portion of the particulate matter.
Currently, researchers recommend a pilot study on a single membrane element for 1,000 hours in each of the wet and dry seasons when membranes are used to treat surface waters (Taylor and Mulford, 1995). As more documentation is developed on membrane process capabilities, the need for pilot testing should decline. For example, for microfiltration, certain types of microbiological contaminants are too large to pass through the membrane pores, and therefore, once the removal capability has been established it should theoretically be unnecessary
to demonstrate the technology at every new application site. Similarly, once tests have proven the ability of a particular reverse osmosis membrane to reject an inorganic contaminant such as uranium or radium, retesting the same membrane at many different locations to verify the uranium or radium removal capability should not be required. This same concept applies to the other membrane processes (ultrafiltration, nanofiltration, and electrodialysis/electrodialysis reversal) when used to remove contaminants for which their effectiveness has been documented, although some testing specific to the source water will be necessary to determine the potential for membrane fouling.
Pilot testing requirements for membranes used in surface water treatment are already being reduced in some states. For example, based on testing by the Metropolitan Water District of Southern California at one site on the Colorado River Aqueduct (Kostelecky et al., 1995), the California Department of Health Services approved use of the Memcor microfiltration process for meeting requirements of the SWTR at several other sites along the aqueduct. The department granted 3-log removal credit for Giardia cysts (meaning regulators will assume the membrane can remove 99.9 percent of these organisms) and 0.5-log (meaning 68 percent removal credit for viruses.
Bench-scale testing of the effectiveness of the particular adsorptive material (activated carbon, ion exchange, or activated alumina) for the target contaminant or contaminants in the raw water is the minimum level of evaluation necessary to design an adsorption system.
For granular activated carbon (GAC) systems, the purpose for which the system is employed (adsorption of dissolved organic compounds, biological stabilization of the water, or particle removal) will influence the type of site-specific testing necessary to design a full-scale system. The most accurate technique for predicting the performance of a full-scale GAC system is a pilot-scale system using the same source water. A pilot system treats the same specific flow rate, carbon size, and influent water as the planned full-scale column but uses a smaller diameter and thus less carbon. Contaminant breakthrough in pilot columns has been shown to closely model full-scale breakthrough (Oxenford and Lykins, 1991). Pilot testing is expensive and time-consuming, however. Although not effective for testing biological or particle filtration applications, small-scale columns can accurately predict the performance of a full-scale GAC system for many adsorption applications (Crittenden et al., 1991). In small-scale tests, the GAC is crushed until the grain-size:column-size ratio is equivalent to that of the full-scale system. The crushed carbon is installed in a column that may be only a couple of inches in length. This small-scale column is used to evaluate the performance of the carbon in treating the source water. Small-scale column tests
are rapid and inexpensive and greatly reduce the time and money required to size a full-scale column.
As with GAC systems, ion exchange and activated alumina bench- or pilot-scale testing can provide the information needed for full-scale design of a system. It is important that the tests, no matter what scale, are performed on the source water to take into account competitive adsorption from other ionic species present.
Powdered activated carbon (PAC) addition to a mixed tank is a relatively inexpensive method of reducing organic contaminant concentrations in finished water. Equilibrium isotherm models, however, do not reliably predict PAC performance in water because contaminant characteristics and the effectiveness of mixing have a strong effect on the amount of contaminant removed by PAC. In addition, as for GAC, background organic compounds affect PAC performance. For these reasons, testing must involve the actual source water. Bench-scale testing can be accurate if careful attention is given to reproducing the mixing characteristics of the full-scale system. However, since existing pilot data on the same source would likely not mimic the intended full-scale mixing at a new site, adsorption kinetics must be taken into account when using such data.
Coagulation/Filtration Systems (Conventional Filtration, Direct Filtration, and Dissolved Air Flotation)
Because the physical and chemical principles governing the performance of coagulation/filtration systems are so complex, some degree of site-specific testing will always be necessary for these technologies unless the technology has proven effective at a different installation using the same source water. In some cases, bench-scale tests using jars to determine appropriate coagulant doses will be adequate, but in other cases site-specific pilot tests will be necessary. The degree of testing required depends in part on the design of the coagulation/filtration system and in part on the characteristics of the raw water.
Effect of Filtration System Design on Testing Requirements
A considerable experience base exists on the ability conventional treatment trains (coagulation followed by flocculation, sedimentation, and filtration) to successfully treat a broad range of water quality. Therefore, site-specific testing requirements are less extensive for package plants employing conventional treatment trains than for those using newer technologies. When a package plant employs newer methods, such as upflow or downflow granular media beds to flocculate and remove particles before filtration, site-specific pilot testing is likely to be needed unless the range of raw water quality characteristics is well within the values for which the equipment has been demonstrated to the satisfaction of consultants and regulatory engineers. Direct filtration plants should always
be pilot tested unless one is already operating successfully to treat the same source of water because these systems are so highly sensitive to water quality.
Through a centralized pilot testing program, it may be possible to reduce site-specific pilot testing requirements for the various kinds of filtration technologies, especially those for which the experience base is not extensive. The range of turbidity that the filter can manage could be determined by testing a very muddy source water and dilutions of that water to provide a range of raw water turbidities for evaluation. Filtered water turbidity and rate of head loss development would be the key performance parameters to document in such testing. Similar tests could be carried out on waters having a wide range of color. The objective would be to treat water of higher and higher turbidity or color to the point of reaching either failure of the process or a very high upper limit, such as 2,000 nephelometric turbidity units or 400 to 500 color units. Finding a quality of water that was not treatable, or documenting the capability of a process train to treat raw water worse than would be encountered in an extreme case, would provide a sound basis for defining the appropriate water quality limits or for determining that nearly any expected raw water turbidity or color could be treated.
Performance limits for package filtration systems using unconventional technologies could also be established by evaluating the effectiveness of existing installations. Manufacturers could provide lists of each installation of their equipment to an appropriate neutral body, which could then review the data to assess the range of raw water quality characteristics that the filter can manage (see ''Centralizing Data Collection" later in this chapter). In particular, dissolved air flotation, although used extensively in Europe and South Africa, is rarely applied in the United States. Consultants and regulatory engineers would benefit from the availability of more performance data that delineates the ability of this process to handle raw water turbidity.
Effect of Raw Water Quality on Testing Requirements
Regardless of the type of coagulation/filtration technology, some level of site-specific testing will always be required, at a minimum to determine appropriate coagulant doses, unless the identical system is treating water from the same source at another facility. Whether bench-scale testing will be sufficient or more extensive pilot tests will be required depends in large part on the source water quality. Source waters with a single quality factor that needs to be treated present the simplest cases and require the least amount of prior testing and evaluation. Examples of these are waters that have no algae and either high turbidity but low color or low-turbidity but high color. When only a single problem needs to be evaluated, determining the appropriate coagulant dose is much less difficult than when multiple factors, such as high turbidity and high color, are involved. Conversely, for waters with various combinations of turbidity, color, and algae, site-specific testing is unavoidable for coagulation/filtration technologies. Tables 4-1
and 4-2 show the types of site-specific pilot data that might be collected during pilot testing of a conventional coagulation/filtration system for a source water with moderate turbidity, algae problems, color, and periodic iron and manganese. For all constituents except turbidity, water samples would be obtained for analysis during the steady-state portion of the filter run, after the initial hour or two of operation, when turbidity improves, but before the end of the run, when turbidity breakthrough (increase) might occur. Continuous measurement of filtered water turbidity provides a record of the complete filter run from beginning to end.
Testing for Removal of Turbidity. Bench-scale jar testing is sufficient for determining the performance of package plants employing coagulation and conventional filtration if the quality of water to be treated falls within the range of water quality for which the package plants have already proven effective. In jar tests, coagulant doses are determined by adding several different doses to laboratory jars containing samples of the source water, stirring the samples, and measuring the turbidity of the treated water after flocculation and sedimentation. As mentioned above, for the other types of coagulation/filtration systems some degree of pilot testing, in addition to jar testing, may be required, depending on the technology and the base of experience in using it.
Testing for Removal of Color. For removal of color with package plants using conventional filtration, jar testing followed by a brief program of pilot testing at cold temperatures and high color concentrations can sufficiently demonstrate process performance. Pilot testing of a broader range of conditions may be necessary for package coagulation/filtration plants using technologies other than conventional ones.
Testing for Removal of Turbidity and Color. Site-specific pilot testing is likely to be needed, even for conventional treatment systems, when both color and turbidity are high. Attaining effective removal of turbidity and color simultaneously can be difficult and usually requires trial-and-error testing to determine optimum coagulant doses, pH, and equipment operating parameters.
Testing for Removal of Algae. Pilot testing on-site is unavoidable for source waters with algae problems unless the algae-laden water is low in turbidity, in which case dissolved air flotation would be applicable because of its proven capabilities to treat such waters. Many types of algae are filter cloggers, causing severe head loss problems unless removed ahead of the filter. Therefore, pilot testing is required to determine the chemical doses and system adjustments needed to ensure removal of the algae prior to filtration.
TABLE 4-1 Water Quality Sampling and Analysis for Testing of a Conventional Coagulation/Filtration System
TABLE 4-2 Pilot Plant Operating Data and Operator Actions for Testing of a Conventional Coagulation/Filtration System
Diatomaceous Earth Filters
Diatomaceous earth (DE) filtration is well suited to small systems because coagulation is not needed for effective removal of Giardia and Cryptosporidium. However, because DE filtration is commonly used without a clarification step ahead of filtration, source water quality limitations are somewhat stringent. Various grades of diatomaceous earth are available, ranging from coarse grades with low rates of head loss build-up to fine grades with substantial rates of head loss build-up. The finer grades of diatomaceous earth are very effective for removing turbidity as well as protozoan contaminants, but the use of such grades causes filter runs to be shorter. Pilot testing may be warranted to demonstrate the effects of using different diatomaceous earth grades, both in the context of turbidity reduction and head loss build-up. Syrotinski and Stone (1975) reported on the use of microstrainers ahead of DE filters in New York as a means of removing algae ahead of DE filtration and thus prolonging the filter runs. Although a long and comprehensive pilot testing program probably would not be needed for DE filtration, a few weeks' testing is valuable for establishing the level of turbidity in filtered water that can be attained by different grades of diatomaceous earth and for indicating the length of filter runs that might be expected with a full-scale plant. The scale of pilot testing can be modest. A DE test filter having a filter area of 0.093 m2 (1.0 sq ft) and operated at a rate of 2.4 to 4.9 m/h (1 to 2 gpm/sq ft) is adequate to provide data for design purposes.
Slow Sand Filters
Pilot testing is always necessary for designing slow sand filters unless a slow
sand filter is already treating the source water in question. Understanding of slow sand filtration technology is insufficient to allow engineers to predict what filtered water turbidity an operating slow sand filter might attain based on chemical and physical analyses of a water to be treated. Construction of a slow sand filter without pilot plant testing and without prior slow sand filter operating experience on the water source in question could result in a small water system having a new filtration plant incapable of meeting one or more drinking water standards. The nature of slow sand filtration is such that after the design parameters of plant filtration rate, bed depth, and sand size have been set, there is little a plant operator can do to improve performance of a slow sand filter that does not produce water of a satisfactory quality.
Slow sand filter pilot plant testing does not have to be expensive. Pilot plant testing has been done using manhole segments and other prefabricated cylindrical products as filter vessels. Plans and a list of materials for such a pilot filter were presented by Leland and Logsdon (1991). Slow sand filter pilot facilities operate over long periods of time—up to a year—but the level of effort can be quite low, consisting of checking head loss, flow rate, water temperature, and turbidity on a daily basis and taking samples for coliform analysis once or twice per week. Leland and Logsdon (1991) provide a recommended schedule for pilot testing of slow sand filters.
Bag Filters and Cartridge Filters
The performance of bag and cartridge filters depends on careful manufacture and use of the equipment rather than on manipulation of the water or equipment during the treatment process. Therefore, testing done in advance is a good indicator of the performance potential of this filtration equipment, and site-specific pilot testing should be unnecessary. Bag filters and cartridge filters are proprietary process equipment designed and built to the specifications of their manufacturers. The filtration occurs as water passes through the bag or cartridge inside a filter housing (pressure vessel) built by the manufacturer. When the manufacturers fabricate bags and cartridges to their specifications, and when users of the bags or cartridges apply them in the proper filter housing, use of such filters should yield reproducible results for the removal of protozoan microorganisms (mainly Giardia and Cryptosporidium ). While not necessary for determining whether the filter can remove protozoan organisms, on-site testing may be useful for determining what water volume the filter can treat before becoming blinded.
Lime Softening Systems
Lime softening, as described previously, is not well suited to application in systems serving fewer than approximately 2,000 people. For small systems
serving more than 2,000 people, lime softening is best suited to ground water sources, which have relatively stable water quality, rather than surface water sources, which can have quality that varies rapidly over time.
For application of lime softening to ground water, bench-scale testing with a jar test apparatus is necessary to determine appropriate process pH and the necessary quantities of lime and perhaps soda ash. Doses of these chemicals should not change greatly over time unless the ground water is subject to periodic infiltration by surface water that changes in quality. For this reason, pilot plant testing is unnecessary for lime softening of ground water that is not influenced by surface water.
If lime softening of a surface water were undertaken by a small system, the requirements of the SWTR would have to be met. For a source water having stable quality, data from other lime softening plants treating source waters of the same or poorer quality, plus jar test data on the source in question, might suffice. For source waters of variable quality, pilot testing, on the water in question or operating data for a nearby full-scale plant using the same source would be preferred. Again, jar test data would be helpful for evaluating treatment options in conjunction with the other data.
Water systems need not conduct pilot plant tests of disinfection systems that use free chlorine, chloramine, chlorine dioxide, or ozone, although limited testing may be beneficial for establishing the disinfectant demand in the presence of organic compounds, iron, or manganese, especially when ozone is used. Studies of chemical disinfectants traditionally have been carried out in centralized laboratory facilities. Regulators consider the laboratory results to be applicable to all systems. Extensive laboratory tests yield information on the extent of microorganism kill that the disinfectant can attain over a range of conditions of temperature, pH, and exposure time. From these data, the EPA has developed tables that specify the concentration and time (CT) conditions needed for inactivation of Giardia cysts and viruses by free chlorine, chloramine, chlorine dioxide, and ozone. Water utilities use the CT data as a guide to managing chemical disinfection.
The basis for this approach to determining the effectiveness of disinfection technologies is the presumption that laboratory results with test organisms are indicative of results in actual water treatment plants under similar conditions of pH, temperature, disinfectant type and residual, and contact time. The approach appears to have developed in part because of the high level of skill needed for the testing, making universities and research laboratories the appropriate settings for carrying out studies, and in part because of the way EPA has approached disinfection regulation and management. The CT concept, while oversimplifying disinfection kinetics, offers water treatment plant operators a practical way of
assessing the adequacy of their disinfection practices, and this is a key factor in managing disinfection in treatment plants.
While the CT concept provides a means for evaluating the effectiveness of chemical disinfectants, no national regulatory guidance is available for ultraviolet disinfection of water by public water systems. UV systems would typically be used only by small systems with ground water sources, and ground water disinfection is not regulated by the EPA (as of June 1996).
Corrosion Control Systems
Current regulations allow small systems to install corrosion control systems without pilot tests. The Lead and Copper Rule allows small systems (but not large ones) to select corrosion control strategies based on desk-top reviews of documents and records of water quality because the cost for long-term corrosion control pilot studies would likely be prohibitive for small systems. Under this rule, small systems were to begin monitoring tap water for corrosion-related problems in July 1993; systems requiring corrosion control are to have treatment installed by January 1998.
One alternative to performing a corrosion control study for a small system is to have a state drinking water regulator or other knowledgeable authority review the quality of the water involved and recommend pH or alkalinity adjustments, use of a corrosion-inhibiting chemical, or a combination of these strategies. Another is to implement a corrosion control strategy in use at another nearby system if both use the same source water and treat it in similar ways. The latter option is especially appropriate for ground waters coming from the same aquifer.
Matching Operator Skill to Equipment Complexity
If any package treatment equipment or treatment process requires skilled operation in order to work effectively, certification of the equipment and approvals for its use should incorporate provisions for proper operation. The skill level of the operators needs to be commensurate with the skill level requirements imposed by the equipment being used. Small systems should never accept the contention that "this equipment runs itself and you don't need an operator." State regulatory agencies, consulting engineers, and equipment manufacturers need to discuss this issue and find alternative approaches to ensuring the level of operation needed for the successful application of treatment technology; Chapter 6 recommends ways to improve training of small system operators.
The skill level or type of understanding essential for successful operation of process equipment varies from process to process. Lime softening plants and those incorporating coagulation require some knowledge of the chemistry associated with the processes. Generation of ozone often involves not only the actual ozone generation step but also an air preparation step. A considerable amount of
mechanical and electrical equipment can thus be involved, as contrasted to a simple chemical feed pump if sodium hypochlorite is used as a disinfectant. If process equipment manufacturers make extensive use of sensors and automated analytical techniques, such as streaming current detectors, turbidimeters, pH sensors, and so forth, the small system using such technology will either need to have an operator who understands electronics and instrumentation and can keep everything in good repair, or it will need to have rapid-response service contracts so that the instrumentation can be repaired quickly if it malfunctions.
Centralizing Data Collection
Currently, there is no one centralized database or clearinghouse for information on the performance of drinking water treatment technologies. Several organizations have databases or other sources of information on treatment technologies for small systems (see Box 4-2), but the databases cover only a limited number of technologies, lack standard data reporting formats, and often are missing information on the full range of parameters (for example, raw water quality finished water quality, and operation and maintenance costs) necessary to evaluate technology performance.
One result of the lack of central data collection is that for all but conventional technologies that have a long record of data, considerable "reinvention of the wheel" occurs in testing. A second result is that, lacking accurate, current information, many engineers, utility managers, and local decision makers continue to select the best-known technology for their system, even if it is not the best choice. Information on alternative technologies and package plants must be made available to these decision makers in an organized and prompt fashion.
The EPA should establish a national clearinghouse to serve as a repository for data obtained on the operational efficiency of treatment technologies. Dissemination of available information on treatment technologies may assist in their acceptance and reduce the expense involved in their adoption. This is especially true for innovative or alternative treatment technologies and systems sold as package plants. Information in the clearinghouse should be made available to all interested parties, including state regulators, water utility managers, consulting engineers, and equipment manufacturers and suppliers.
This central clearinghouse could be established by expanding the Registry of Equipment Suppliers of Treatment Technologies for Small Systems (RESULTS) database at the National Drinking Water Clearinghouse (NDWC) (see Box 4-2). RESULTS could be expanded to include information on both raw and finished water quality, operational requirements, operation and maintenance costs, and useful life of the technology. Manufacturers could provide lists of installed equipment, and the performance of the equipment at each location could be included. Pilot-scale data, in addition to data from full-scale operations, could be entered into the database. Periodically, the NDWC or another organization
identified by the EPA could evaluate the data to assess the performance limits of different technologies.
For the information in the clearinghouse to be useful in comparing technologies, it must be reported in a standard format. Therefore, a standard testing protocol should be developed, and any test data entered into the database should follow the protocol.
State agencies responsible for regulating drinking water systems should assign an individual to serve as liaison to the central clearinghouse. This individual would be responsible for staying informed of performance data for the systems of importance in his or her state. The states could then continually update their requirements for site-specific testing of water treatment systems. For many systems, testing requirements can be decreased as more data become available on system performance under a range of water quality conditions.
Small systems can expend significant sums in pilot testing water treatment technologies prior to installation. While site-specific testing requirements cannot be entirely eliminated, they can be streamlined. In developing programs to reduce pilot testing requirements, regulators at the state and federal levels will need to consider the following issues:
- Failure to share water treatment performance data from state to state leads to site-specific pilot testing requirements for package plants that are in some cases unnecessary. Pilot data collected in one state may not be accepted in another state. In some cases, small water systems must spend money to prove elements of technology performance that have already been demonstrated elsewhere.
- A nationally accepted program of treatment technology testing and verification could help reduce repetitious pilot testing requirements. Technology certification would provide assurance to state regulators that a product will perform as advertised and that manufacturers' data are not just a "sales pitch."
- Even with a national water treatment technology certification program in place, some site-specific testing will be needed before a treatment system is designed and installed. The testing can be as simple as laboratory tests of water quality parameters or as complex and expensive as multiseason pilot plant investigations. The extent of testing required depends on the type of technology under consideration and the quality of the source water.
- Surface water treatment technologies for which performance is linked to source water quality generally have more complex testing requirements than technologies whose performance is largely independent of water quality. For example, coagulation and filtration systems are more likely to need site-by-site evaluation than membrane filtration systems.
The following sources provide a starting point for obtaining information about technologies for small drinking water systems. In addition to these organizations, a regional phone book may have a listing of engineering consulting firms that specialize in water treatment plant design and firms that provide certified treatment plant operation and maintenance services on a contract basis.
- The EPA should continue the technology verification program, to be implemented by the National Sanitation Foundation, for water treatment technologies for small systems. Equipment evaluation should include verification of the manufacturer's claims, especially claims of reductions in contaminant levels, and evaluation of operation and maintenance manuals to ensure that they provide accurate and complete information about the equipment. The verification report should indicate the level of operator oversight required for proper technology performance. After successful testing, technologies should carry a stamp or marking to identify that their performance has been verified.
NDWC provides RESULTS, which runs on any IBM-compatible computer, for a nominal charge, approximately the price of a computer diskette. RESULTS can sort information by contaminant type, technology, plant location, equipment supplier, or cost. Multiple criteria can be linked to derive information on a system that, for example, removes Giardia cysts and costs less than $50,000. RESULTS is limited in that some small system technologies are not included in the database, information is not reported in a standard format, and important parameters for technology evaluation (such as operating costs and raw or finished water quality) are missing from some of the entries.
The NDWC is housed at West Virginia University along with its ''sister" organizations: the National Small Flows Clearinghouse for wastewater treatment technologies and the National Environmental Training Center for Small Communities, a group that provides training resources for drinking water, wastewater, and solid waste treatment. In addition to RESULTS, NDWC offers a quarterly newsletter, a toll-free technical assistance line, a toll-free electronic bulletin board called the Drinking Water Information Exchange, and many free or low-cost educational products.
- The EPA should establish a standardized national database for water treatment technology information by expanding the existing RESULTS database at the National Drinking Water Clearinghouse. All database entries should include quality assurance information. The EPA should permit anonymous data entries to allow those providing data to include all reliable data, not just data that comply with regulations. As with the current RESULTS database, the information should be made available at a nominal fee and should be configured for use on a desk-top computer. It should also be made available electronically, via the Internet. The availability of the database should be advertised to regulators, water utilities, and consulting engineers.
- The EPA should oversee the development of standard protocols for pilot testing of water treatment technologies. These protocols will have to be developed for each treatment technology separately but should allow information gained from pilot tests to be entered into the standardized database and shared with other potential technology users. Pilot plant data should be made available in the national database to allow rapid dissemination of this information for use by utility decision makers and state regulatory agencies.
- The language of certification and testing should be standardized and should be the language of the Safe Drinking Water Act regulations. Data on raw and finished water quality should be collected in common units corresponding to the requirements of the SDWA.
- State regulatory agencies responsible for overseeing drinking water systems should establish a mechanism for reviewing and updating their requirements for the testing required before a drinking water plant can be installed. As more experience is developed on treating types of water important in the state, the amount of site-specific testing required can be decreased.
APHA, AWWA, and WEF (American Public Health Association, American Water Works Association, and Water Environment Federation). 1992. Standard Methods for the Examination of Water and Wastewater, Eighteenth Edition. Washington, D.C.: APHA.
Crittenden, J. C., P. S. Reddy, H. Arora, J. Trynoski, D. W. Hand, D. L. Perram, and R. S. Summers. 1991. Predicting GAC performance with rapid small-scale column tests. Journal of the American Water Works Association 83(1):77–87.
EPA (Environmental Protection Agency). 1993. The Safe Drinking Water Act: A Pocket Guide to the Requirements for Operators of Small Water Systems. San Francisco: EPA, Region 9.
EPA. 1994. Office of Ground Water and Drinking Water Publications. EPA 810-B-94-001. Washington, D.C.: EPA, Office of Ground Water and Drinking Water.
GAO (U.S. General Accounting Office). 1994. Drinking Water: Stronger Efforts Essential for Small Communities to Comply with Standards. Washington, D.C.: GAO.
Kavanaugh, M. C., and R. R. Trussell. 1980. Design of aeration towers to strip volatile contaminants from drinking water. Journal of the American Water Works Association 72(12):684–692.
Kostelecky, J. D., M. C. Ellersick, W. W. Trask, Jr., B. M. Coffey, and D. A. Foust. 1995. Implementation of microfiltration for metropolitan's small domestic water systems. Presented at 1995 AWWA Membrane Technology Conference, Reno, Nevada, August 13–16, 1995.
Leland, D. D., and G. S. Logsdon. 1991. Pilot plants for slow sand filters. Pp. 191–227 in Slow Sand Filtration, G. S. Logsdon, ed. New York: American Society of Civil Engineers.
McCarthy, R. 1995. Presentation to the National Research Council's Committee on Small Water Supply Systems, Washington, D.C., March 2, 1995.
McClelland, N. I. 1994. NSF International: programs and services offered internationally. Presented at U.S. Russia Business Development Committee Standards Working Group, Moscow, Russia, May 24–25, 1994.
Oxenford, J.L., and B. W. Lykins, Jr. 1991. Conference summary: practical aspects of the design and use of GAC. Journal of the American Water Works Association 83(1):58–64.
Syrotynski, S., and D. Stone. 1975. Microscreening and diatomite filtration. Journal of the American Water Works Association 67(10):545–548.
Taylor, J. S., and L. A. Mulford. 1995. Membrane protocol to meet the ICR. In Proceedings of the 1995 Membrane Technology Conference. Denver: American Water Works Association.
WMA, Inc. 1994. Small Systems BAT Task Force: Interviews with State Officials Regarding the Application of the Recommended Standards for Water Works in Reviewing Small System Technologies. Report prepared for the EPA. Alexandria, Va.: WMA, Inc.