EPA is in the process of developing requirements for disinfection of groundwater-based drinking water systems (Groundwater Disinfection Rule, GWDR). The approach now being taken for this is somewhat different from that taken for other drinking water regulations. This paper will discuss the context, issues and status of this rule as of December 1995.

The regulatory work group, made up of representatives from States, EPA and other interested parties, has considered the goals for the GWDR to be that it make a positive impact on public health by ensuring that those drinking groundwater are safe from waterborne microbial disease, that it be feasible and that it can be successfully implemented.

Public health concern and the mandate for a GWDR
The health risks from microbial contamination of drinking water are well known. Many diseases have been identified: Hepatitis, Legionnaires's Disease, heart and liver diseases, diabetes, diarrhea and other gastrointestinal illnesses are just a few. The number of annual waterborne microbial illnesses in U.S. is estimated between 1-15 million cases. The majority of these are gastrointestinal diseases. These are generally acute and short-lasting, but for some may be chronic or yield chronic symptoms. These can be severe or fatal to some people, including the elderly and especially the immunocompromised and immunosuppressed. Estimated annual waterborne disease deaths in U.S. are between 900-1800.

It is not known how many of the 110 million people served by groundwater-based (versus surface water-based) public water systems become sick each year. Many disease outbreaks in the U.S. have been associated with groundwater systems, but we believe that this substantially underestimates the actual endemic disease rates, since the majority of waterborne disease cases go unrecognized and unreported. Because of the public health and regulatory implications of this, EPA has initiated research to better estimate these rates and determine their public health impacts.

Available occurrence data indicate at least 60-70% of groundwater sources have been contaminated with lecal viruses or bacteria. However, the health significance of these findings are unclear. This contamination is wide spread and results from human and animal fecal material migrating from surface deposition or from septic systems and sewers. Several occurrence studies are currently underway to provide a more detailed national picture of the problem, particularly to identify site conditions, such as hydrogeology and land use patterns, that may affect contamination of wells.

Analysis of EPA enforcement data show about 40,000 of 180,000 groundwater systems have had violations indicating bacterial contamination of groundwater, well or distribution system. Contamination of distribution systems results from crosscontamination events and from bacterial regrowth within a system. Larger community systems appear to have more distribution system violations. Smaller community and non-community systems tend towards source water contamination.

All these lines of evidence indicate that microbial problems exist in many groundwater systems. Fortunately, disinfection of pathogenic microorganisms in drinking water has long been acknowledged to yield significant improvements in public health. Few would question the general advisability of such treatment. Because of this, Congress directed EPA in the 1986 Safe Drinking Water Act Amendments to develop national requirements for disinfection of all public water systems. Under this legislation, these systems must disinfect their water, unless criteria are met to avoid disinfection. The Surface Water Treatment Rule (SWTR), promulgated in 1989, addressed microbial contamination of drinking water from surlace sources. The obvious vulnerability of surface waters to microbial contamination led to stringent provisions for filtration and disinfection and limited avoidance criteria. The analogous rule for groundwater systems is now in development. Since groundwaters may be less vulnerable to microbial contamination and contaminant organisms may be filtered out or inactivated during passage through the subsurface, these regulations will include well-developed and realistic avoidance criteria.

Current status of GWDR development
The formal EPA work group reconvened in January 1995 and meets regularly. It includes representatives from all 50 State drinking water programs, as well as EPA drinking water and groundwater staff from headquarters, laboratories and regions. All interested parties are encouraged to provide input to the rule development process. The intention of this open process is to ensure adequate discussion of issues and approaches prior to the development of regulatory language.

EPA is currently under a court-ordered deadline of December.

1995 to begin negotiating a schedule for proposing and promulgating a GWDR. The realistic time requirements for the necessary research, public discussion, regulation development and review suggest a proposal towards the end of 1997 or early 1998.

A variety of issues must be considered in the development of an appropriate regulation. 1) Public health goals will largely determine regulatory requirements and will influence avoidance criteria. Should this regulation focus on good operations and maintenance to enhance system reliability and thus ensure protection, or set specific treatment requirements to achieve acceptable numerical risk goals? 2) As noted above, the impacts of microbial contamination are poorly known. If these are large and widespread, groundwater systems would be considered "guilty unless proven innocent" and GWDR requirements far reaching. Treatment might be mandatory. Avoidance criteria might be stringent. If impacts are small, systems could be "innocent unless proven guilty" and the GWDR limited. Reliance on monitoring and programs to protect sources and systems might be emphasized. 3) EPA is concerned for consistency of the GWDR with other regulations. A wide variety of State regulations currently exist addressing microbial contamination of groundwater systems. The Total Coliform Rule has provisions that address several of the microbial problems of groundwater systems. Other drinking water regulations may affect choices for treatment or avoidance criteria. Any GWDR must be consistent with these. It will be important that the GWDR reflect and support those proven to be protective of human health. In this sense, the GWDR aims to provide national standards flexible enough for States now providing adequate regulation to continue without substantial new impacts, while ensuring that protection is given to individuals in States that do not. 4) At least some systems plausibly should not require disinfection. Means to accurately determine system vulnerability to contamination and set appropriate avoidance criteria must be available and field tested. These criteria should reflect credit for source water protection and well-head protection programs, cross-connection and back flow prevention programs, as well as existing State well codes and well siting requirements. They must also consider the hydrogeological factors affecting pathogen die-off rates and travel times and distances in the ground.

Possible easts and benefits
A major concern of the regulated community is the potential costs of this regulation. Costs and benefits will be determined quantitatively when the necessary information is available and potential elements of the rule developed, but are likely to be positive. The benefits of this rule will depend on the extent of the existing national problem, the public health goals selected and the rigor of implementation and enforcement. It is not yet known how many annual cases of waterborne disease could be prevented, but they are likely to be substantial. Approximately half the community groundwater systems currently do not treat their water. Substantially fewer noncommunity systems (schools, factories, rest stops, restaurants, etc.) disinfect. The work group has recognized that costs for even a minor case of diarrhea resulting in lost workdays and a trip to the physician may exceed $500. A recent report by the Congressional Budget Office on the costs of the SDWA concluded that the benefits from the SWTR substantially outweighed its costs. We expect a similar result for a GWDR.

As with the SWTR, a "treatment technique" approach will be developed, specifying operational requirements rather than achievement of MCLs for specific microorganisms. This is because methods to detect pathogens belbre they reach the customer are not available. Utilities' concerns for high costs Irom a GWDR are reflective of the possibility that such treatment could be required at every well. While this is possible, technologies for inactivating microoganisms in groundwater are well understood, practical and relatively inexpensive. About half the community systems currently have some type of disinfection and may not require substantial additional treatment. A major unknown being addressed is any potential need for storage capacity to achieve chlorine CT values. Storage tanks pose problems for siting, distribution system maintenance and costs. The use of chlorine itself may be a problem for some systems for compliance with the upcoming Disinfectant/ Disinfection Byproducts Rule. Besides chlorine, ultraviolet light and ozone are being considered as possible candidate disinfectants.

Alternative approaches to avoiding treatment generally require siting wells in locations not vulnerable to microbial contamination, providing adequate operations and maintenance and ensuring protection of the distribution system. These include conducting sanitary surveys, correcting any inadequacies, and compliance with groundwater source protection elements, well construction codes, cross-connection control programs and back flow prevention programs. Vulnerability to contamination varies greatly with hydrogeological setting, such that certain regions may not be significantly impacted by a GWDR, while other regions may require significant resources to comply. The magnitude of this is unknown at this time.

Current, highly speculative estimates for annual household costs suggest that these would be less than $100 for the smallest systems and in the range of $3-5 for the largest systems.

Conclusion
There are substantial reasons fbr EPA to move ahead with development of a GWDR, to determine health issues and goals, initiate necessary research and examine regulatory approaches.

The health hazards are well established. Occurrence may be widespread. Treatment technology is reliable and affordable. Approaches to avoiding contamination and treatment are practical and applicable to many contaminants. Finally, groundwater systems are not required to disinfect at this time; thus some populations may be inadequately protected from waterborne disease organisms. This does not mean that the elements of a GWDR have been determined or are obvious. The opportunity is here to fully consider a GWDR prior to proposal.