Food and beverage manufacturers and processors in the U.S. operate on the premise that incoming municipal water is safe, and they will always receive notice of any anomaly in time to protect their processes and products in case of contamination.
EPA drinking water regulations are cited as the rationale to give municipal water users a safe-harbor-like exemption to bypass water in their written food safety plans and not to evaluate their water integrity risks in a formal, analytical, or scientific way that facilitates planning, prevention, and risk mitigation.
At the same time, final Food Safety Modernization Act, or FSMA, rules make the food and beverage manufacturers responsible for ensuring that the water they use does not contaminate their product. But how sure can they really be that their incoming water is safe enough to mitigate the need for additional water safety precautions?
Not Structured for Food Safety
The U.S. EPA rules that implement the Safe Drinking Water Act use a level of protection based on water consumption data for a household of four, the viability and virulence of the organisms of concern, and the effective dose considering stomach acid as an effective barrier and developed immunity based on repeated exposure to the same water source.
However, food manufacturing facilities don’t have “stomach acid.” A food manufacturing or packing plant “drinks” almost 90 percent of the water it takes, so its risk profile is significantly different than that of an individual. Most of the water a plant “drinks”—90 percent compared to the 5 percent for the individual—is used for product, or used for cleaning, cooling, making ice, or various processing tasks like pushing product through pipes. These uses give microbes an opportunity to find a niche and grow and thrive, in essence contaminating or infecting the facility. Sometimes the municipal water is used to wash the outside of closures, containers for distribution, and cases—and this municipal water can transmit microbes through the food supply system. The 2010 Oregon dairy Salmonella outbreak illustrated this risk.
The EPA drinking water risk assessment yields rules for both prevention (treatment techniques) and performance assessment.
EPA preventive rules require water suppliers to meet the treatment standards (regardless of technique) for at least 95 percent of the water they distribute to the public. This provides about 5 percent flexibility from an operational standpoint (calculated by time or by volume). A system that does not achieve the 95 percent standard would have a treatment technique violation and would have to notify the public either in their annual Consumer Confidence Report or more immediately, depending on the circumstances.
Of course, no consumer—even if their flow rate is 500,000 gallons a day—would necessarily know if they were getting some of the allowable 5 percent off-spec water.
The theory is that 5 percent off-spec water would be dispersed enough that no individual intake would get enough to create serious contamination or illness. So a household is unlikely to be endangered by the small amount of off-spec water it would receive in the 300 gallons it would use during a day. But a food manufacturing facility could be significantly impacted based on its much higher consumption of 10,000 to 500,000 gallons a day. In food safety terms, this 5 percent threshold would be the equivalent of allowing 5 of every 100 gallons produced at a milk plant to be at risk of not having been treated adequately to inactivate the pathogens inside, clearly not an appropriate food safety standard.
On the performance assessment side, many issues surface in using rules designed to make safe water for individuals and applying them to food manufacturing facilities. For example, will the sampling regime catch contamination? Positive total coliform samples on the weekly or monthly monitoring could require repeat sampling, trigger sampling or E. coli testing, or other assessment. But positive samples would be reportable as a violation only if the system takes less than 40 tests a month or if more that 5 percent of the samples are positive. A system that has a positive sample can take more samples and stay under the threshold for public notice or violation, so that food production facilities would be unaware of the contamination, masking potential food safety risk.
In addition, where public notification is required, samples reflect water already used and the announcements (boil water advisories included) serve to protect residents prospectively. But food manufacturers who had used contaminated water may have manufactured the product and even shipped it. They would rarely have enough information to understand the risk of contamination of their facility’s piping and infrastructure. A food manufacturer could conceivably inoculate their pipes and internal water systems with difficult to trace or treat microbes and probably go about business as usual.
These rules are designed to protect household drinking water and not food manufacturers, processors, or packagers who use water provided by the most conscientious and compliant of suppliers.
Municipal Water Drawn from Deep Groundwater Wells
Many public water systems do not disinfect the water they provide to the public. As part of the research to evaluate the public health impact of this practice, the EPA chartered and funded the “WATHER” study, which took an epidemiological perspective on testing non-disinfected ground water supplies that met all the health standards. The multi-year study looked at public drinking water systems in Wisconsin. The state, like many others, does not necessarily require public drinking water systems disinfect groundwater supplies on the theory that the deep groundwater wells they draw from are safe and do not test as coliform positive.
The WATHER team announced in 2012 that they had found infective viruses in drinking water samples in 14 Wisconsin communities that did not disinfect their water, with up to 25 percent of the samples positive for infective viruses. The study, published in a peer-reviewed journal, reported that up to 22 percent of the acute gastro intestinal illness experienced in the communities was directly associated with the drinking water viruses.
Wisconsin is probably typical of the situation in other states, and while the WATHER project studied only a small fraction of U.S. groundwater systems that do not disinfect the water they distribute to customers, there is no reason to assume that a WATHER-type study would find different results in any other state.
A similar study in Minnesota found comparable results. In fact, the joint Health Department/Department of Agriculture study found 22 of 245 samples positive for Salmonella out of the 14 percent of 567 un-disinfected municipal water supplies that were tested.
But whether or not the WATHER or Minnesota data is echoed in all the other states, for the next decade at least, drinking water rules, requirements, and enforcement actions will only protect municipal water to the extent needed to achieve safe drinking water for individual households and not necessarily safe food manufacturing water.
Total Coliform Rule Complications
In February 2013, the EPA announced the final Revised Total Coliform Rule (RTCR), after a multi-year process to consider public water suppliers’ concerns. Municipalities, authorities, water districts, and private water companies campaigned for reform, claiming the following.
- Public notice rules are not useful; in a resource-constrained environment, money would be better spent fixing issues rather than notifying the public about them. The public doesn’t really understand the data; giving them unnecessary information reduces their confidence in the public water supply and impacts their ability to get money to fix issues in the rate setting process.
- Testing for indicators instead of targets is not conclusive. Coliforms are not necessarily pathogenic and do not always indicate fecal contamination. To the extent that a coliform-positive demonstrates a pathway, focus should be on “find and fix.”
- Most waterborne disease comes from viruses, which are not tested for or measured. As far back as 2002, data published by the EPA highlighted the presence of viruses in chlorinated water systems, and specifically in biofilms in pipes that connected the treatment plant to the distribution system. Research showed that viruses survived in groundwater and were infective, especially embedded in biofilms in chlorinated systems, where pseudomonas biofilms are resistant to chemical disinfectants.
In response to these and other concerns from its constituencies, the EPA reformed the TCR by:
- Reducing monitoring frequency for many systems,
- Changing from coliforms to E. coli as the key indicator on the notion that it actually indicated fecal contamination,
- Changing most public notice requirements, and
- Moving away from boil water advisories except in the most onerous cases of clear and present danger in the immediate future, shifting the regulatory scheme towards corrective action.
Dubbed “Find and Fix,” the new regulatory scheme requires water suppliers to do an assessment of each significant safety problem and develop a plan to correct it. Note that a corrective action can be to study the problem further. But as long as documentation of the problem and its corrective action is provided to the state within a 30-day period, both utility and regulator need take no further action. If the problem reoccurs within a specific time frame, the rules provide for a Level 2 Assessment and other measures, but not necessarily public notice.
Thus the RTCR will substantially reduce transparency, especially regarding information about safety problems that is available on a real-time basis to the public in general and food manufacturers in particular.
Food Not on EPA’s Radar
The bottom line is that once RTCR is fully implemented and rolled out nationwide in 2016, despite the anticipated challenges of aging infrastructure, reduced treatment budgets, lower repair budgets, and lower enforcement staff availability that will make water safety issues more rampant and more difficult to solve, there will be even less information and less protection for food manufacturers and processors who use municipal water.
The “Identifying the Gaps in Understanding the Benefits and Costs of Boil Water Advisories” report showed that less than 10 percent of over 508 notices reviewed for the first six months of 2010 would have been required to boil their water under the new rules. The study looked at the impact and costs and benefits of public notices and boil water advisories. It used data about the impact on local immediate water illness but did not consider outbreaks contributed by food manufactured in the communities subject to the boil water advisories. When questioned on this point, the study’s primary author explained that food manufacturers were not specified constituents by the 42 states or the 12 water utilities that participated.
Know Your Water
It is advisable to know your incoming water source, understand if your supplier buys finished water or treats it, and determine what percent of their output you receive. Institute a system to evaluate, monitor, and verify issues and changes in your specific supply such as checking on your state database for recent detections and violations, and whether you were notified of such; and finding out where you are located with respect to the treatment plant—are you at the end of the line or are there sampling points nearby that will provide information.
Posy is vice president, strategic services and regulatory affairs, for Atlantium Technologies. She works with governments and private entities to promote effective UV treatment standards and improve water safety through non-chemical disinfection. Reach her at email@example.com.