Classifying Drinking Water Contamination for Regulatory Consideration
Bruce A. Macler
U.S. Environmental Protection Agency
The bottom line of the U.S. Environmental Protection Agency (EPA) Drinking Water Program is to provide safe water that everyone can drink and remain healthy. Technological applications are necessary to do this, but they must be as vigorous and solid as possible. The majority of the current water treatment needs are day-to-day operations at the local level, specifically at a small system level. Southern California Water and the Orange County Water District are among the top 400 utilities in the United States in size and sophistication. There are 55,000 public water systems in the country that are community water systems. There are more than 100,000 transient water systems. There may be 20,000 more systems that serve places such as schools or factories. These are not as sophisticated, and thus they need very vigorous and robust treatment technologies and approaches.
There is a general misunderstanding as to what regulations are about, what they are aimed to achieve, and how they are developed. The EPA does oversee assessment of wastewater; however, the primary area covered in this paper is drinking water treatment and distribution. Both problems and opportunities are considered.
The primary goal of regulation is protection of public health. The aim is to have water that is safe to drink, water that is swimmable, and water that provides fish that are safe to eat.
Clean Water Act
Ecosystems are considered particularly under the Clean Water Act. For instance, protecting fish is an issue because there need to be enough fish available that are safe for human consumption.
The Precautionary Principle and Safety
The biggest component is the precautionary principle, embedded in the water laws that lead to water regulations. Applying the precautionary principle means that you want to be as protective as possible. If the risk is unknown, there must be a conservative estimate, and the worst must be assumed. There is also a big consideration of cost-benefit and the appropriateness of matching costs against public health return.
How safe is safe? Perceptions of safety are politically, socially, and economically important.
The federal laws and the regulations are means to ends. The Clean Water Act, the Safe Drinking Water Act, and all of the environmental acts involving other media are like paints on a palette, used to paint a picture of improving public health. For water utilities, the regulations are an end in themselves, because most utilities are not proactive. The top 400 can be proactive because they have the resources and the mind-set. They will deal with contaminants at a much lower level than they actually have to. Many utilities, however, are just trying to meet standards, which mean that the standards have to provide the needed level of protection.
The Clean Water Act rules are fairly simple: water should be fishable and swimmable. It is fishable in the sense that there are enough fish to catch, and from a public health standpoint, they must be edible and nontoxic. Swimmable means that the water is nontoxic. You can swim in it and you can get a little water in your mouth and it will not hurt you.
Safe Drinking Water Act
Ambient water quality criteria have been set up by toxicologists and risk assessors in the EPA and give end points for what is safe. Those end points are developed strictly on a precautionary principle basis and without any regard for whether they are achievable. The health effects of a con-
taminant are evaluated based on viability and how solid the information basis is. If the information is based on adult exposure, the results are extrapolated to be protective of everyone, and a benchmark is set on a reference dose or cancer level. The value may be well below detection or treatability, but it will stand as the criterion for the contaminant.
Similarly, under the Safe Drinking Water Act, there is also a maximum contaminant level goal for oral ingestion set with respect to risk and conservativeness. There is no consideration of practicality. The goals are essentially zero risk, but zero risk is not achievable. There must be a consideration of what passes or what can be achieved.
From a risk assessment standpoint, this causes problems because the value cannot be detected epidemiologically. Some people have said that people are not getting sick from arsenic; therefore, there should not be a regulation, or the regulation should be set at a level where people do get sick. Epidemiology does not have that resolution. The bottom line is that these goals are below what can be detected in the real world.
As a result, it is not possible to backtrack to determine success. It is hoped that the risk assessments are legitimate.
Congress essentially said in the 1986 Safe Drinking Water Act Amendment that there would have to be more regulations. The EPA then had the task of trying to find additional chemicals to regulate that it had not already given Congress.
Now it is necessary to take a look from a systems standpoint. Hazard assessment via a critical control point process, such as that used in the food industry, could also be applied to drinking water. In that case, a system-wide examination from source to tap is done in order to identify problem areas and move forward.
Improved Analytical Methods
There is a need for straightforward and reliable analytical methods that are relatively cheap. For many of the small utilities, the cost of monitoring is the major cost, particularly with small groundwater systems for which they really do not have surface infrastructure or big treatment plants. They basically have a well, a storage tank, and the lines that go out from them. Methods exist for nearly every contaminant, but sometimes they are expensive. Working toward cheaper methods that serve our purpose is necessary.
Improved Treatment Technologies
Another important need is low-maintenance treatment technologies. It is very hard to find operators that are well trained to run sophisticated treatment hardware. We need stand-alone equipment that will run automatically, a necessity if we are to move forward.
Point-of-use devices and point of entry devices have been suggested as a means to beat the problem of highly treated water being used to water lawns or wash cars while a small percentage is used for drinking. The idea would be good if every homeowner could be trusted to run a device reliably to remove all contaminants on the list, and this is a daunting task.
WATER USE PROBLEMS: FROM SOURCE TO TAP
Contaminants removed from water have to be disposed. Clean Water Act activities focus primarily on contaminant control—point control through National Pollutant Discharge Elimination Systems (NPDES) permits, the permits on discharge from manufacturing or from point sources and nonpoint sources such as agricultural or water runoff. The parameters (e.g., pH, temperature) are very broad and are largely there to protect fish and other aquatic life from degradation.
There are controls on specific pollutants obviously, but in general, there are broad parameters. There is an increasing worry over the persistent organics or chemicals that have been created and are somehow getting into wastewater. There are some worries, but there is no way to know right now if they are dangerous, because too little is known.
Drinking Water Resources
With regard to drinking water source and resource issues, neither the Clean Water Act nor the Safe Drinking Water Act deals with resources directly. These acts were written that way specifically; Congress did not want to interfere with state’s rights in water resources. However, they have been used to deal with resources to some degree, and they certainly come to play in desalination, recycling, and reuse.
There are source water protection provisions in the Safe Drinking Water Act and the Clean Water Act that are aimed at prevention and control of contaminant releases. The Safe Drinking Water Act and the earlier versions really dealt with treatment plant operations.
One potential drinking water resource is obtained through desalination of salt water. In desalination, the main issue is brine disposal, particularly in inland areas. Desalination in coastal areas works well because the brines can be disposed by running them back into the ocean and, to a first approximation, they are relatively benign. To a certain extent this also works inland, as in the Los Angeles area where there are brine drains.
On the other hand, in agricultural areas in the Central Valley of Northern California, there are problems from agricultural drainage, which are essentially brines. The political and technical aspects of this prevent sending the brines down through the delta and into the ocean or creating a pipeline from the inland areas out to the ocean. In order to keep the lands in production there have to be better ways to dispose of the brines.
Reuse of wastewater is another potential source of drinking water. The current problem with reuse is that there is not much information about the health effects of persistent organics. There are enough persistent organics in the water for some concern, which may or may not be legitimate. It will be some time before there is adequate health information to go forward. Unfortunately, there may well be pressure to regulate before there is a good handle on the associated risks.
In terms of needs, wastewater and reuse treatment technologies are certainly needed to control these persistent organics. The Orange County Water District and others are doing some very interesting work on this. Ideally it would be good to get away from a chemical and biochemical approach to this problem and instead try to remove all organics regardless of their potential for harm.
Another problem with wastewater reuse is contamination by disinfection by-products. Water disinfection technologies are needed to minimize these by-products, especially with the growing concern about nitroso-containing chemicals. The work on membranes is impressive but could improve. The reliability of membranes is great, but lowering the energy costs will make a difference, particularly for applications to smaller systems.
Drinking Water Treatment
The main issue with drinking water treatment has always been dealing with microbial pathogens in a multiple-barrier approach. The laws from the Safe Drinking Water Act, the Surface Water Treatment Rule and its variations, prescribe a physical removal basis and a disinfection basis. Source water protection now presents an additional barrier, and physical removal has many steps: coagulation, flocculation, precipitation, and filtration.
Waste minimization of sludge is still an issue. Dealing with and minimizing sludge and finding other types of coagulants that do not produce as many tons of sludge would be great. Removal of the problematic chemical contaminants is also important; if they cannot be kept out of the sources, then they must be removed through treatment.
All surface water systems and any groundwater system will have multiple contaminants. Until regulating contaminants one by one stops, utilities are going to have to work with multiple contaminants that may be quite distinct chemically.
Water Treatment Needs
The immediate needs for water treatment include improvement in the membrane technology, such as pretreatment, resistance to fouling, durability, ease of use, and breadth of contaminant removal.
There are some oxidized metals and metalloids and other chemicals on the horizon. Of the chemicals dealt with in drinking water on a national basis, from a public health side, arsenic is the biggest issue and is worse than most contaminants except for the microbials and disinfection by-products. Treatment needs lie in the area of “-ate” control: arsenate, chromate, nitrate, perchlorate, phosphate, silicate, and others.
Utilities are looking at their distribution systems, many of which have been in the ground for a hundred years and are finding that the systems are starting to fall apart. The systems have to be replaced at a rate of more than a million dollars per mile. Many larger systems are thousands of miles in length.
Most utilities deal with physical failure on a when-it-happens basis. If a main blows causing a sinkhole in the middle of the street, they go out and fix it. Most utilities have a replacement rate of around 1 percent a year.
Options are needed that would minimize corrosion and damage. Infiltration and intrusion of contaminants, particularly microbials, have to be prevented. In the urban setting, drinking water transmission lines run in the same areas as sewer lines. Sewer lines leak and the area in soil around sewer lines has a lot of contaminants that are not wanted in drinking water. The problem is particularly bad in cities that use wells; as wells are kicked on and off, there are pressure fluxes through the distribution. There can be zones of zero or negative pressure moving through. Siphoning also occurs.
There is a cross-connection from different types of intrusions to worry about. About a third of waterborne disease outbreaks are associated with distribution system failures. There are going to be regulations on distribution system protection in the next four to six years.
There may be ways of lining the distribution system pipes with polymers or membranes that would help harden the system. Methods exist but more are needed. Furthermore, in-place pipeline rehabilitation materials and the methods for putting them in without digging up streets would be beneficial.
Systems must maintain disinfection and the disinfectant residual in distribution because of potential intrusions and to some extent because of the potential for bioterrorism. However, increased disinfection boosts disinfection by-products, but these are currently regulated. Additionally, there is a need to have residual and to control biofilms that can sequester materials.
What are the needs for distribution systems? A broad-spectrum disinfection agent that can persist in the distribution system and does not create problematic by-products is critical. Chlorine works well but its use has downsides. It creates halogenated by-products and a taste that people do not like.
Water systems are incredibly vulnerable to bioterrorism, from both the chemical side and the biological side. It is pretty remarkable that to date there is no evidence of attacks
on water systems, because it is essentially impossible to prevent an attack. There have been a couple of attempts that did not work, but there have been no thought-through attacks by individuals or groups such as state-supported bioterrorists or local militia types.
Multiple barriers to attack have limited effectiveness in the likelihood that an attack spread beyond a few people; the goals are to minimize the damage. If there was a bioterrorism attack, it would be a unique event, so water systems are not going to put a lot of effort into preventing an attack. Monitoring tools are being worked on as a means of response.
Not many people will have access to weapon contaminants, so there is not a need for very many labs to be able to analyze those. False positives are bad, because we do not want to scare the public, but a false negative is much worse. Current technologies for analysis are crude and include very broad parameters. It would be useful to be able to have some technologies that could be performed on a real-time basis.
Cleanup and Disposal
Technologies are becoming available now for cleanup and disposal, such as thin film-type technologies for specific contaminants, but more technologies would be better. They can start being distributed, but there is not a huge need for that. If contaminants get into a distribution system, there will be miles of system to clean up. A small consideration is that some of those agents will be fairly persistent at times.
NEEDS FOR THE FUTURE
Over the next five years, simple removal technologies utilizing disposable media are needed for small utilities. Control of arsenate, perchlorate, n-nitrosodimethylamine, and methyl tert-butyl ether is needed. Ion exchange is great if it can be done. There are large-scale applications in the perchlorate area. The Colorado River has a chlorate concentration of 4-5 parts per billion (ppb), and it is conceivable to have a future regulation in that range.
The best way to work on water quality is to go upstream to Henderson, Nevada, where the contamination originates and try to eliminate the problems there. However, that feat would take some work. As mentioned previously, a better distribution system disinfectant would be helpful.
For 10 or more years from now, it would be ideal to have the technologies ahead of the regulations. Control of persistent organics in the brine and sludge disposal technologies is needed. The long-term trends should be to control the organics and clean up the water to customer satisfaction because there is more to water quality that meeting standards. There is a persistent pressure to control things that people worry about, and the health information is not available for backup.
Limiting the Proliferation of Biocides
Don Phipps, of the Orange County Water District, stated that on the control side of biofilms, some interesting technologies have recently been getting the attention of the media. There are also new surface-active biocides, chemicals that can be covalently bound to surfaces and will limit the proliferation or attachment of bacteria. He mentioned that he thought it might be an interesting process to try to develop pipe liners that are coated with these compounds and see if someone can gauge a test to determine if this would help limit the proliferation of biocides in the pipes. Mr. Phipps said that limiting the proliferation of biocides has a couple of advantages, including the reduction of the disinfection load required to maintain a residual. If the proliferation could be limited, the efficiency would be increased for the standard liner, and the total microbial load distribution system could be lowered.
Dr. Macler said this would reduce the demand on the residual. However, he said the real issue would be the stability of the biocide on the surface. How would you get the residual on the surface to begin with?
Mr. Phipps thinks the way these biocides really operate is that they are bound to the surface. They are not an exchangeable biosynthetic.
Water Regulation of American Indian Casinos
Dick Carlson, of San Diego County, commented that in his county there are a number of American Indian casinos using recycled water because they need waste distribution as a result of the waste load. They actually have anecdotal evidence of recycled water being used inside one of the casinos for toilet flushing, which is not a problem. However, this has been connected to the drinking water system in the same casino. He wanted to know who regulates this kind of activity.
Dr. Macler explained that the EPA is responsible for regulating those activities. He said that under the Safe Drinking Water Act, EPA develops and promulgates these drinking water regulations and the states can take on the authority to implement them to gain primacy. American Indian tribes have some sovereignty, but EPA deals with their lands. The agency has direct implementation authority and handles the situation with the tribes. He said that more tribes are running these casinos in California and they are growing from small-
scale operations of about 200 people drinking from a system, to a large scale where thousands to tens of thousands of people are being exposed.
Environmental Management System for Local Municipal Government Entities
Dan Askenaizer, of Montgomery Watson and Harza Engineering Company, asked whether EPA regional offices or headquarters are involved in the effort to encourage the use of an environmental management system for local municipal government entities.
Dr. Macler answered that the programs are not really on the drinking water side. There are plans for hazardous materials. He thinks some of the utilities were looking to the effort to be relieved of regulatory burden.
The federal level in Washington, D.C., has not really embraced the drinking water program. Therefore, he is unsure what the likelihood of persistence for the program will be.
Mark Matsumoto, of the University of California at Riverside, acknowledged the discussions about the amount of dollars available for R&D going down in drinking water areas as well as the fact that few impacted agencies are able to provide the funds or have the funding to do the R&D necessary for their own research needs. He asked Dr. Macler what he sees as the near-term and long-term funding picture for government research.
Dr. Macler affirmed the interesting situation in R&D. Looking at just EPA’s budget and the money that goes to the laboratories in Cincinnati for drinking water R&D, he said it goes down. At the same time, there are congressional line items where a congressman in Glendale provides $900,000 to the city for R&D, $750,000 for more R&D, and $2 million for still more R&D; it becomes evident where a lot of R&D is getting done. Dr. Macler said that the Metropolitan Water District has received a federal line item of $3.5 million for desalination research that he is overseeing. The district has to match it. EPA has about $6 million in desalination membranes, brine disposal, and pretreatment. There is money in other places, but no one is trying to bring the pieces together to make sure everyone is trying to reach one main goal. Money is available. In looking at the congressional line item, three or four years ago it was a few million dollars and now it is hundreds of millions. He added that there are a lot of $0.5 million to $2 million R&D projects that are supposed to go out for bid.
Mr. Matsumoto next mentioned the perchlorate debate as an example of the role of science with regard to setting limits.
Dr. Macler referred to the Safe Drinking Water Act in which Congress said that water should essentially be without risk. This essentially means that if the risk is unknown, the worst case should be assumed. Often the data being used come from both human and animal studies.
For example, for perchlorate, the studies are at the 200- to 300-ppb exposure level. That is the lowest observed adverse effect level. He said the study on humans was very short term in adults. The study on rats used multiple generations of rats and looked at the pups, fetuses, and pregnant mice. The data then have to be extrapolated to make some judgment for humans. A sensitive subpopulation for humans is pregnant women and infants, but studies cannot be done on these populations. Therefore, a conservative extrapolation must be made. For cancer, a linear approach is used to extrapolate downwards. For noncancer end points, uncertainty factors are used.
Dr. Macler continued that if the limit is moved up an order of magnitude, the unknown risk is greater. The level might be legitimate, but the comfort level is reduced. He said that science can only go so far, and the risk assessment world has to make a determination. Risk assessors must be willing to take uncertainties into account. Eventually there is a management decision that takes into account the social aspect of the situation.
Dr. Macler said that a possible reason for a perchlorate drinking water regulation is the Colorado River and the 34 million people that drink its water. If the risk is one in a million, this is equivalent to 34 people, which is unacceptable from the Safe Drinking Water Act standpoint. It would have been best if cleanup of the Colorado River had started five years ago. He said that some methodologies are being developed for cleanup of such sites that are not too expensive.
Different Levels of Treatment for Different Uses
Debbie Elcock, of Argonne National Laboratory, raised questions about having different levels of treatment for different uses. Based on the statutes and legislation, we know what the level should be for drinking water. Do we know what the level would be for different industrial uses or for agriculture? Water for these purposes would not have to be as treated. It might be economical to actually treat for the different levels for different uses. Is there the legislative authority to do that?
Dr. Macler replied that treatment for industrial use would depend on the process and on whether the regulations were for environmental or occupational safety and health purposes. For agricultural purposes in California there are different watering requirements depending on the vegetation. Median strips along freeways are watered with different water than golf courses. If human exposure is more likely, the water quality must be better. Another consideration is whether plants can tolerate the water.
Bhaskar Davé, of Ondeo Nalco, added that he thinks the safety of people working in an industrial environment should be first and foremost. He said that according to Title 22 for California, there is a minimum residue of chlorine allowed to ensure that pathogens are killed before coming to an industrial site. As to different treatment levels for different uses, the industry at a given site would have to make a decision about where the water is being used, because different water could be used in a cooling tower than in the boiler. Dr. Davé said that the treatment would be very application specific and not very cost-effective.
Mr. Phipps added that for industrial use, the constraints of the process will probably be more stringent on the quality of the water than the regulatory environment.
Dr. Macler responded that the true value of water often is the cost. He likes to think of water having values in different ways; that is, industrial water is ultrapure and very valuable. There are quantifiable values for water that is used for certain applications in industry, and other applications do not require that high water value. Domestic drinking water is valuable because it is to be protective of public health, but water for lawns should not be valued as heavily.
The National Water Research Institute has had some projects to determine values of waters to be able to have a decent economic approach. California would benefit from giving some consideration to the value for different uses.
Fareed Salem, of ConocoPhillips, commented that in principle for maximizing reuse, water should be treated to the maximum quality affordable. The difficulty in assessing the value to each industry or each application is where the struggle is. He thinks most industries are beginning to think defensively by determining the minimum that can be done to get by the regulation. He asked how we can change the mindset or the culture in the business community to think proactively about water quality and to adopt an offensive rather than a defensive strategy. Fundamentally, if we were to really think of water sustainability and long-term solutions, we have to change the mind-set of the culture into one that thinks differently about this resource.
Dr. Macler did not know if that was feasible. His experience is that usually an economic situation is being attacked. If you can make these technologies in a way that is more profitable and get the benefit of risk reduction or better environmental qualities, then it will happen. Additionally, the small systems do not have a lot of money to be able to adopt a proactive outlook. They can barely be reactive.
Maximum Contaminant Level Goal
An unidentified speaker mentioned some studies on recyclable water where a risk level of 1 in 10,000 infections per year on the drinking water side was used. That value is then compared to the safety of the recycled water. The speaker wondered where such a value might have originated.
Dr. Macler stated that he believes whoever came up with the value “pulled it out of the air.” The concept of acceptable risk came up in the 1980s during the 1986 revision of the Safe Drinking Water Act, in which Congress set up the maximum contaminant level goal (MCLG) and a level of no known health consequence with an adequate margin of safety. He said that traditionally the MCLG has been set at zero for carcinogens, meaning there is no acceptable level. That was a decision not a regulation or a law. The real wording is that the risk is one in a million, an upper-bound determination of the worst that could possibly happen. The real risk is quite likely to be much less than that and could actually be zero. Dr. Macler continued that the percentage of deaths due to gastrointestinal problems from microbials is near 0.01 percent. The risk is then 1 in 10,000. He asked whether this was a reasonable number. He said that it turned out to be a feasible number with respect to the Surface Water Treatment Rule and the regulation for Giardia. The regulation manager investigating microbial risk assessments presented this number at a conference, the value got printed, and finally it showed up in the preamble to the regulation. The background rate of infection is about 1.5 gastrointestinal illnesses per person per year.