Quality control of bottled and vended water in California: a review and comparison to tap water.

Consumption of bottled and vended water in the U.S. is increasing annually by 12% to 15% and is highest in California; residents consume over 36% of the bottled water sold, resulting in an annual consumption of 22 gallons per capita (1). A comparison of the major sources of drinking water in the U.S. and California is presented in Table 1. Thirty-three percent of Californians use bottled water as a primary source of drinking water (4). Despite increasingly stringent regulations to ensure high quality tap water, consumers are willing to pay on the average 700 times more for alternative sources. Californians have placed a higher value on bottled water primarily for reasons of taste and perceived health concerns (4). As consumption of bottled water increases, so does concern over whether this water is an adequate source for mass consumption.

This research was conducted to compare the current regulations and compliance for quality control of bottled and vended water in California with that of the tap water industry. The tap water industry, as referred to herein, consists of public water systems that have at least 15 service connections or regularly serve at least 25 individuals. Non-carbonated, non-mineral water, comprising 90% of the bottled and vended water market, was the focus.

Methodology

Data were obtained through interviews with federal, state, and local regulators and industry representatives; published studies; California and federal statutes and regulations; and compliance records. Both site visits and telephone interviews were used. Changing regulations and market forces result in continual changes in the bottled water industry. Although the literature review included earlier studies, only data published after 1985 were considered representative of current conditions. There is considerable literature published on the bottled water industry, but the vast majority addresses non-regulatory, non-technical aspects (e.g., consumer behavior and marketing aspects). No published reviews of vended water were found.

Background

The first large scale assessments of bottled water quality conducted in the early 1970s revealed serious problems with quality control and catalyzed significant changes in federal and state regulations (5,6). Since the early 1970's numerous independent researchers, consumer advocacy groups, and government agencies have assessed various aspects of the industry. Generalization of results must be made with caution because such assessments are often limited in scope (e.g., limited number of samples taken) and cannot be extrapolated over time. However, on occasion these analyses have identified gaps in existing regulations or problems in the industry. A review of the more recent studies highlights two potential public health concerns: inconsistent water quality due to inadequate process control at bottling facilities and deterioration of water quality after storage as described below.

Periodic analyses of the chemical composition of bottled water indicated that a majority of bottled water met applicable standards (7,8,9,10). However, in some samples, volatile organic compounds (VOCs) including toluene and benzene were identified. This contamination, thought to be introduced during processing, was attributed to inadequate process control. The equipment and handling intensive processing of bottled water provides many opportunities for the introduction of contaminants. For instance, operation and maintenance of equipment (e.g., ozonators, pumps, deionizers, bottle fillers) requires the use of lubricants and cleaning solvents which if not adequately controlled will contact the water product. The presence of VOCs indicates the need for better quality control at the bottling plants, and continuous monitoring and surveillance by the oversight agency. Although current regulations establish requirements for monitoring and recording of process control activities, the underlying causes for poor process control are untrained employees and poor management policies which are, in general, unaffected by regulations. Adequate process control is more likely to be better ensured through stricter enforcement of existing requirements than through new requirements.

Table 1. Provision of drinking water and population served in 1990

Source                                 Population Served
                             United States            California

Tap Water (2,3)              242 x |10.sup.6~         29.6 x |10.sup.6~
Private Wells (2,3)          10.6 x |10.sup.6~        0.2 x |10.sup.6~
Bottled Water (1,4)          27.5 x |10.sup.6~        9.9 x |10.sup.6~

Limited research has been conducted to evaluate the effects of storage on the bacteriological quality of bottled water (11,12). Bottled water is often stored, frequently at room or higher temperatures prior to consumption. Overall, the studies indicated that while bottled water met existing bacteriological standards based on total coliform organisms, changes in the bacteriological quality indicated by heterotrophic plate counts were strongly dependent on the initial water quality, container quality, length of storage, and temperature. Scarpino, et al. (12) found heterotrophic populations in excess of 500 colony-forming units (CFU) per mL in 27% of bottled products analyzed after 48 hours at 35 |degrees~ C. Further increases occurred after seven days of incubation at room temperature. Existing microbiological standards based on total coliform counts do not indicate all risks posed to the public; under adverse storage conditions, certain pathogenic bacteria may flourish, endangering immuno-compromised individuals. Some research has been conducted to assess the health implications of the microbiological quality of bottled water (13,14); results were inconclusive.

It seems likely that the effects of storage are most directly related to the initial quality of both the finished water and the container. In theory, excellence in process control would assure quality water reaching the consumer. However, the tap water industry has learned over decades of service that setting up multiple barriers to contamination, rather than relying on design and operation perfection at a single stage, is a more responsible position to take when dealing with a commodity as vital as drinking water. Multiple barriers in drinking water include selection and protection of the best possible source water, removal of contaminants with both physical and chemical processes, disinfection, finished water monitoring, and having a residual disinfectant in place to guard against regrowth or re-entry of microorganisms. Although many of these barriers are common to both the tap and bottled water industry, the use of a residual disinfectant is not common in bottled water. In fact the absence of residual chlorine (and its attendant taste) is one of the marketing tools used by the industry. Given the absence of a residual disinfectant; the wide range of storage conditions encountered; the documented cases of poor process control as related to chemical contamination; and research results indicating high counts of non-coliform organisms in water meeting current microbiological standards, it is prudent to set up alternate barriers. One barrier would be the expansion of microbiological standards beyond coliform organisms alone, such as instituting a limit of 500 CFU per mL, as recommended by Scarpino, et al. (12). Another would be to include a date of bottling on water products so that at a minimum consumers would have additional information on which to make decisions.

A historical overview of the regulations for bottled and vended water in California was conducted by Richardson (15). Overall, there exists very little documentation on the current status of regulations and compliance for quality control of bottled and vended water in California in comparison to tap water. In 1985, Umino (16) of the California Assembly Office of Research compared the quality standards for bottled and vended water and tap water, and reviewed inspection reports and consumer complaints for several bottlers for the time period of 1979 to 1984. Umino concluded that under the quality standards applicable at that time, bottled water was not guaranteed to be safe. The report cited 121 specific complaints made against 12 bottling companies. Most of the complaints were referenced as being made by consumers to state regulatory agencies or as being lodged during regulatory inspections. The four types of complaints made serve to illustrate potential problems (not uniformly well-referenced) with quality control of bottled water. First, unsanitary conditions (e.g., unclean equipment, personnel improperly attired, low temperature bottle wash water, questionable quality rinse water, and rust in storage tanks) at processing plants were noted. Second, contamination of finished water (e.g., chemicals, microorganisms, insects, and plant material) was noted. Third, potential contamination (e.g., underground gasoline storage tank, a cesspool, improper well sealing, and a cross connection between a spring and a well) of source waters was cited. Fourth, poor recordkeeping, including falsified reports, was noted. State legislation was passed to address several of the regulatory inadequacies identified.

In 1989, based on a very limited comparison of federal regulations for bottled and tap water, the Environmental Policy Institute concluded that regulations for bottled water were not on par with those for tap water, and that bottled water was not necessarily any safer than tap water (17).

Description of Industries

Although both industries produce and supply drinking water to the public, their operating environments differ markedly and have direct implications for applicable regulations and possible public health concerns. Tap water, an integral part of a community's infrastructure is available for consumption upon demand without transport or storage by the consumer. Bottled water must be purchased, transported, and stored prior to consumption. In California, the cost of tap water ranges from $0.45 to $2.85 per 1000 gallons with an average cost of $1.60 per 1000 gallons (18). In 1990, the average cost of bottled and vended water in the U.S. was $0.90 per gallon (1).

Both industries are fragmented. Sixty-three percent of the tap water systems in the U.S. are small and serve less than 3% of the population. The vast majority of the population in the U.S. (94%) and California (97%) are served by large community water systems with more than 200 service connections (2,3). There are approximately 475 bottled water plants in the U.S., producing 600 brands of bottled water (19). The four largest manufacturers produce approximately 40% of the bottled water (17). Most bottled water companies produce small volumes for a particular region. In California, 98 facilities produce bottled water, as well as 30 out-of-state producers and 11 producers from outside the U.S (20).

In the U.S. tap water industry, groundwater supplies 81% of the systems and 33% of the population, whereas surface water supplies 19% of the systems and 67% of the population (23). Approximately 70% of the source water for the bottled water industry is from groundwater(19). The remaining 30% of bottled water is produced from tap water. Almost all vended water is produced from tap water. The treatment that tap or bottled water receives prior to distribution is dependent upon the quality of the source water, applicable regulations, community resources, and consumer tastes. For instance, treatment of tap water from surface sources often entails chemical coagulation, sedimentation, and mechanical and activated carbon filtration followed by chlorine disinfection. Groundwater supplies, on the other hand, are often used by municipalities with no or minimal (chlorination) treatment. Treatment of bottled water also spans a wide range of options. A spring water may undergo only filtration and ozonation to produce a "natural spring" water, whereas a chlorinated tap water may be subject to activated carbon filtration, reverse osmosis or deionization, and ozonation to produce a "purified" water. Treatment of tap water is primarily to produce a safe drinking water supply and secondarily to improve water aesthetics. Conversely, most treatment of bottled water is to improve aesthetics.

Regulatory Climate 

Although regulations for both industries are intended to ensure acceptable water quality, historically, the development of those regulations has taken independent paths, and enforcement has involved different state and federal agencies. Regulations for the industries differ due to the nature of the industries, the respective oversight agencies, and applicable statutes.

Early regulation of both the bottled and tap water industries involved federal and state government, each having separate jurisdictions and regulations. However, tap water now has uniform national regulations, whereas bottled water is still subject to federal regulation with limited applicability and inconsistent state regulation. The federal government has not been successful in setting a regulatory precedent or in maintaining uniform regulations for bottled water, and states have established regulations with varying levels of stringency.

Until 1974, each state established its own standards for intrastate tap water, although the federal Public Health Service (PHS) standards (applicable to interstate water systems) frequently were recommended as minimum requirements. In the 1960's, the increased mobility of Americans led to expectations for a consistently high quality tap water throughout the country. Inconsistent and sometimes ineffective state regulations were circumvented in 1974 by passage of the federal Safe Drinking Water Act (SDWA) which established uniform national drinking water regulations applicable to all public water systems (21). The U.S. Environmental Protection Agency (EPA) was given the responsibility to establish and oversee national regulations. State governments are granted primacy for implementing and enforcing provisions of the SDWA if they adopt and enforce regulations no less stringent than the federal regulations. After passage of the SDWA, California adopted its own Safe Drinking Water Act and established primacy. Specific regulations are delineated in the California Domestic Water Quality and Monitoring Regulations (22).

Federal regulation of bottled water is mandated through the 1938 Food, Drug, and Cosmetic Act (FDCA) and administered by the U.S. Food and Drug Administration (FDA) (23). In response to studies in the early 1970s that revealed quality control and sanitation problems at bottled water plants (5,6), the FDA established regulations specific to bottled water in 1973. These regulations included quality standards equivalent to the PHS standards of 1962 and Good Manufacturing Practice regulations. Furthermore, within 180 days after the EPA adopts new primary drinking water regulations, the FDA is required to propose similar standards for bottled water or provide justification for not doing so; the FDA has not always acted within this time constraint. FDA regulations are only applicable to bottled water manufactured in one state and marketed in a different state, comprising as little as 30% to 40% of the market (23). No federal regulations have been set for the intrastate market or vended water.

California regulates the bottled and vended water industry to a much greater degree than provided in the FDCA. California's Sherman Food, Drug, and Cosmetic Law is the basic statute that authorizes such regulation. Regulations consist of those established specifically for bottled and vended water; general state laws and regulations related to sanitation in food processing; and federal regulations related to bottled water quality standards and good manufacturing practices. Details of the regulations can be found in Allen and Darby (24).

Oversight Agencies

The tap and bottled water industries have different oversight agencies, but regulations for both imply considerable self-regulation. Industry representatives have the primary responsibility for ensuring that potable water is supplied to the consumer. Oversight agencies conduct inspections and review analyses to assess compliance and use enforcement tools to bring about compliance.

A single focus agency, the Office of Drinking Water (ODW) in the California Department of Health Services (DHS), is responsible for oversight activities for the 1150 large tap water systems that serve 97% of Californians (2). Funding for oversight activities is obtained from the state general fund (80%) and the U.S. EPA (20%). County health authorities oversee approximately 9700 small community, noncommunity, and state small water systems serving 2.5% of the population as well as oversee other public health activities (2). Funding is obtained from fees charged to water systems and general county support. The primary oversight agency for the bottled and vended water industry is the Food and Drug Branch (FDB) in the California DHS. The FDB oversees the 100 bottled water TABULAR DATA OMITTED producers in the state as well as 12,400 other firms involved in the manufacture and distribution of food, drug, and cosmetic products. Funding is obtained from licensing and inspection fees and the state general fund. In principle it is likely that a single focus agency is better able to enforce regulations.

Standards; Monitoring and Reporting Requirements; and Compliance

Standards for bacteriological, inorganic, organic, and radiological contaminants that have potentially adverse health effects have been set for both industries. A detailed description of the requirements in California can be found elsewhere (22,24). In general, water quality standards for both industries are comparable in terms of identifying potential risks to public health. However, the majority of organic standards for bottled and vended water are only applicable to the source water. As described previously, studies of bottled water quality indicate that organic contaminants have been introduced into finished waters during bottling. Without an applicable standard for the finished water, this contamination may not be detected.

The bottled water industry, as profit generating businesses, is expected to be able to comply with all applicable regulations as a condition of staying in business, whereas tap water, as a sole provider, can petition to be exempt from certain non-health related quality standards under special conditions. For example, the Twenty-nine Palms Water District has been granted a 30-year variance to the maximum contaminant limit (MCL) for fluoride due to unusually high natural concentrations of fluoride in the source water.

Currently, bottled water is required to be labeled according to an established set of definitions and include the name and location of the water producer or distributor. Although not currently required in California, stating the origin and treatment of the water on the label would be useful to the consumer, particularly because current requirements produce labels that are meaningless without a copy of the definitions. Such information regarding tap water is public information and readily available.

Monitoring requirements for both industries specify the frequency, number, and type of water analyses to be conducted and are intended to provide continuous assurance that the water meets the standards. Due to the inherent differences between the industries, actual monitoring requirements vary, and a strict comparison between the two is not possible. However, the assessment conducted for this research indicated that monitoring requirements for each component of the bottled water industry are likely to identify if a standard has been exceeded.

Reporting requirements for both industries facilitate surveillance activities conducted by oversight agencies. Requirements for reporting of water quality analyses and violations to oversight agencies are significantly more stringent and comprehensive for the tap water industry. Foreknowledge of identified risks and ability to avert risks differs markedly between the two industries due to the nature of the product and the product delivery. The less stringent reporting requirements for the bottled and vended water industry seems to be based on the assumption that the water producer is both well-informed of risks (i.e., water quality is well-established prior to product consumption) and will take appropriate action without guidance by oversight agencies. Additional reporting for the bottled water industry may be warranted because under current requirements the oversight agency is not likely to be informed of identified risks to public health.

Table 3. Summary of violation rates for small water systems (2)

Standard           number of    monitoring       MCL      monitoring and MCL
                  systems(a)   violation(b)   violation      violations(c)

Bacteriological      2847          48%           16%              57%

Inorganic            1025          42%            6%              44%

Organic               87           45%            -                -

Turbidity             254          83%            7%              85%

Radiological         1025          62%         0.39%              62%

a Number of systems that must meet applicable standard, systems include small
community, noncommunity, and state small water systems; compliance with all
standards is not required by all systems.

b Monitoring violation rates are based on failures to monitor only; reporting
violations are not included.

c Percent of water systems violating either the MCL or monitoring requirements
or both requirements.

In theory, regulatory compliance can be evaluated by the frequency and severity of reported violations of applicable regulations or assessed by independent laboratory analyses. In practice, regulatory compliance for both industries is extremely difficult to ascertain. First, care must be used in reviewing compliance statistics; numbers alone are not necessarily indicative of risk to public health. As an example, there is a difference in risk to public health posed by a system that does not monitor for bacteriological quality; by one that monitors and complies with the MCL yet has taken an incorrect number of samples; and by one that complies with MCLs yet has reported analyses results late. Yet, all three systems would be categorized as violating the M/R for bacteriological quality. Second, records of violations are incomplete and not directly comparably for the two industries.

Although more complete records are available for the tap water industry, there is no routine comprehensive compilation of compliance statistics at the state level. For this research, recent compliance data for both small and large systems were obtained from an ODW district office and a county health department overseeing public water systems located in northern California. Data was also obtained from a 1988 DHS assessment of compliance among 3000 small water systems. Less than 5% of the large public water systems in California have been found to be in violation of the regulations recently. Violation rates for monitoring and reporting requirements are greater than those for MCLs. As an example, in 1988, 0.46% of large water systems in California violated the MCL for the bacteriological standard, whereas 0.74% violated the M/Rs for the same standard (2). Records of monthly violations for 160 large systems in northern California in 1990 indicated that the bacteriological MCL was violated only once whereas the M/R was violated 48 times. Violation rates for small water systems, shown in Table 3, are much greater than for the large systems. Breaking the total number of violations down by system size, a majority of the violations occurred in very small water systems, as illustrated in Figure 1. Continuing violations of monitoring requirements among the small systems is attributed to lack of sufficient resources and training for system operators to comply with complex and demanding regulations, as well as limited resources for county health authorities charged with enforcement. Enforcement actions taken by the DHS and county health authorities include administrative, civil, and criminal actions. Administrative actions (e.g., telephone calls, letters, citations, or compliance orders) are used for the majority of violations.

Assessment of compliance for the bottled water industry is particularly problematic. Not only are records at the FDB incomplete due to minimal reporting requirements, but there is no compilation of violation statistics for those records that do exist. For this study, the general level of compliance was assessed from communications with eight of the 15 FDB district offices and the FDB headquarters, and from a detailed review of industry files at one district office. For the majority of private water sources, water haulers, vending machines, and retail water facilities there were no significant violations reported. A wide range of violations were identified for water bottling facilities; the majority were related to recordkeeping (e.g., incomplete information on maintenance records, laboratory analyses, and labels; and poor documentation of process control). Operational violations (violations that create conditions that compromise the quality of the water) were noted at two small bottling plants. At one plant, inadequate monitoring and operation of ozonation equipment had occurred periodically. At another, improper control of the sanitization process for cleaning refillable five gallon bottles was noted. One recall was initiated due to an improperly functioning filling machine that resulted in glass slivers in produced bottled water. Similarly to ODW, enforcement actions taken by the FDB include administrative, civil, or criminal actions. One important difference in enforcement by the FDB is there are no penalties associated with any of the tools outside of lawsuits.

Conclusions

Tap water quality is mandated by uniform national regulations, whereas bottled and vended water is subject to federal regulation with limited applicability and inconsistent state regulation. California is one of several states that has sought to establish equivalency between quality control of bottled and vended water and tap water. Based on available statistics, regulatory compliance for the bottled and vended water industry is no worse than that of the tap water industry. The majority of violations posed no significant health risks. For both industries, small systems comprised the majority of the violations, whereas large systems generally had excellent compliance rates. On the basis of this study, three recommendations are made:

1. Water quality standards for the two industries are comparable except that many of the organic standards for bottled water are applicable only to the source water, a concern due to documented cases of organic contamination during processing. Standards for chemicals that have the potential to be introduced during processing should be applicable to the final as well as to the source water.

2. Reporting requirements are significantly less stringent for bottled water facilities and allow the industry much latitude in assessing risks as well as make assessment of compliance difficult. A protocol should be set in place to periodically collect, compile, and assess compliance statistics.

3. Although stricter enforcement of existing regulations for process control would be a positive step in assuring a high quality product reaches the consumer, more extensive microbiological standards (beyond total coliform bacteria) should also be considered. Further research is warranted on how deterioration during storage is affected by finished water and container quality and storage conditions and whether the deterioration is significant in terms of human health.

References

1. International Bottled Water Assoc. (1990), "20 Questions About the Bottled Water Industry," Alexandria, VA.

2. Calif. Dept. of Health Services (1990). "A Report to the Joint Legislative Budget Committee."

3. USEPA, Office of Drinking Water (1991), The National Public Water System Program, FY 1990 Compliance Report (draft).

4. The Field Institute (1990), "Californians' Views on Water." "A Survey of Californians' Opinions about Issues of Water Supply, Development, Quality, and Policy," San Francisco, CA, 13-28.

5. Ballatine, C.L. and M.L. Herndon (1973), "The Water that Goes into Bottles," FDA Consumer, 7:4:5-7.

6. USEPA, Water Supply Division (1972), "Bottled Water Report, A Pilot Survey of Water Bottlers and Bottled Water.

7. Suffolk Co. Dept. of Health Services (1990), "1990 Bottled Water Survey and Addendum," New York.

8. N.Y. State Dept. of Health, Bureau of Public Water Supply Protection (1987), "Survey of Volatile Organic Chemical Compounds in Bottled Water Products Distributed in New York State.

9. Consumers Report (1987), "Water, Water Everywhere," Consumers Union, 52:1:42-48.

10. Eitzer, B.D. and H.M. Pylypiw, Jr. (1990), "Survey of Bottled Waters Sold in Connecticut," Dairy, Food, and Environ. Sanitation, 607-609.

11. Geldreich, E.E., H.D. Nash, D.J. Reasoner, R.H. Taylor (1975), "The Necessity of Controlling Bacterial Populations in Potable Waters-Bottled Water and Emergency Water Supplies," J. Am. Water Works Assoc., 67:117-124.

12. Scarpino, P.V., G.R. Kellner, and H.C. Cook (1987), "Bacterial Quality of Bottled Water, J. Environ. Science and Health, A22:4:357-367.

13. Duquino, H.H. and F.A. Rosenberg (1987), "Antibiotic-resistant Pseudomonas in Bottled Drinking Water," Can. J. Microbiology, 33:286-289.

14. Edberg, S.C. (1991), "Technical Assessment of the Microbiological Health Effects of Bottled Water," Yale U. School of Med., New Haven, CT, 1-30.

15. Calif. Dept. of Health Services (1990), Statement by Stuart E. Richardson, Chief. "Proceedings of the Bottled Water Workshop," Committee on Energy Commerce, U.S. House of Rep.

16. Umino, W. (1985), "Bottled and Vended Water: Are Consumers Getting Their Money's Worth?" Assembly Office of Research, Calif. Legislature.

17. Marquardt, S., V. Smith, J. Bell, and J. Dinne (1989), "Bottled Water: Sparling Hype at the Premium Price," Environ. Policy Institute, Washington, DC.

18. Am. Water Works Assoc. (1986), "1984 Water Utility Operating Data Water Billing," Report 3, June 26.

19. U.S. Food and Drug Admin. brochure (1990), "New Bottled Water Standards."

20. Calif. Dept. of Health Services, Food and Drug Branch (1990), List of licensed bottled water plants and distributors in California, Bottled Water Licensing System, District Listing, (10-15-90).

21. Kawata, K. (1986), "Evolution of Drinking Water Regulations in the United States," J. Environ. Health 48:4:206-209.

22. Title 22 California Code of Regulations, Ch. 15, Register 77, No. 45, Sec. 64401 et seq.

23. Troxell, T. (1990), "Role of the FDA in Regulating Bottled Water," Proceedings of the Bottled Water Workshop, Committee on Energy and Commerce, U.S. House of Rep.

24. Allen, L. and J.L. Darby (1991), "Quality Control of Bottled and Vended Water in California: A review and comparison to tap water," Dept. of Civil Eng., U. of Calif. at Davis.

25. Calif. Health and Safety Code, Article 6.5, Sec. 26594.1.

Jeannie L. Darby, Ph.D., Asst. Professor, Dept. of Civil and Environmental Engineering, University of California at Davis, Davis, CA 95616